perm filename F85[JNK,JMC] blob sn#806988 filedate 1985-12-31 generic text, type C, neo UTF8
COMMENT ⊗   VALID 00680 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00091 00002	
C00092 00003	∂06-Aug-85  0931	NISSENBAUM@SU-CSLI.ARPA 	CSLI Talk 
C00093 00004	∂06-Aug-85  1056	TRUDY@SU-CSLI.ARPA 	Julia Robinson 
C00097 00005	∂06-Aug-85  1102	TRUDY@SU-CSLI.ARPA 	Julia Robinson 
C00098 00006	∂06-Aug-85  1316	EMMA@SU-CSLI.ARPA 	Newsletter 
C00099 00007	∂06-Aug-85  1702	SBARNES@SUMEX-AIM.ARPA 	SIGLUNCH  Friday  August 9, 1985    
C00107 00008	∂06-Aug-85  1831	DAVIES@SUMEX-AIM.ARPA 	No meeting this week  
C00108 00009	∂06-Aug-85  1900	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #9   
C00114 00010	∂07-Aug-85  1133	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	NEXT MONDAY'S PLANLUNCH   
C00117 00011	∂07-Aug-85  1532	ullman@diablo 	paper received 
C00119 00012	∂07-Aug-85  1718	EMMA@SU-CSLI.ARPA 	Newsletter August 8, No. 40    
C00128 00013	∂08-Aug-85  1244	YAMARONE@SU-CSLI.ARPA 	SANDWICH ORDERS <9:30 
C00130 00014	∂08-Aug-85  1320	SELLS@SU-CSLI.ARPA 	using Imprint  
C00132 00015	∂08-Aug-85  1527	PIERRE@SU-CSLI.ARPA 	Printing DVI files 
C00133 00016	∂09-Aug-85  0944	BETSY@SU-CSLI.ARPA 	SDF Visit 
C00135 00017	∂09-Aug-85  1340	DAVIES@SUMEX-AIM.ARPA 	Advanced Architecture Meeting Wednesday August 14!! 
C00136 00018	∂11-Aug-85  1837	ULLMAN@SU-SCORE.ARPA 	["Prof. David Maier" <maier%oregon-grad.csnet@csnet-relay.arpa>: Complexity of LP evaluation]
C00140 00019	∂11-Aug-85  2059	DAVIES@SUMEX-AIM.ARPA 	New mailing list: CARE-Users    
C00142 00020	∂11-Aug-85  2138	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #10  
C00153 00021	∂12-Aug-85  1244	YAMARONE@SU-CSLI.ARPA 	Extra Turkey Sandwich 
C00154 00022	∂12-Aug-85  1303	@SU-CSLI.ARPA:GEORGEFF@SRI-AI.ARPA 	Monday Planlunch   
C00157 00023	∂13-Aug-85  1027	ullman@diablo 	tomorrow's meeting  
C00162 00024	∂13-Aug-85  1040	ullman@diablo 	another talk   
C00165 00025	∂13-Aug-85  1225	YAMARONE@SU-CSLI.ARPA 	Two Turkey Sands.Available 
C00166 00026	∂13-Aug-85  1548	ullman@diablo 	Porter paper   
C00167 00027	∂13-Aug-85  1611	vardi@diablo 	Re:  Porter paper    
C00168 00028	∂13-Aug-85  1937	YM@SU-AI.ARPA 	Seminar on Control in Prolog  
C00171 00029	∂13-Aug-85  2205	@MIT-MC.ARPA:HEWITT@MIT-MC.ARPA 	Prolog will fail as the foundation for AI so will LOGIC as a PROGRAMMING Language
C00174 00030	∂13-Aug-85  2332	DAVIES@SUMEX-AIM.ARPA 	QLISP on CARE    
C00187 00031	∂14-Aug-85  0115	@MIT-MC.ARPA:Wayne@MIT-OZ 	Prolog will fail as the foundation for AI so will LOGIC as a PROGRAMMING Language 
C00197 00032	∂14-Aug-85  0243	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #11  
C00231 00033	∂14-Aug-85  0922	@MIT-MC.ARPA:DAM@MIT-OZ 	Symbolics Prolog    
C00233 00034	∂14-Aug-85  0936	@SU-CSLI.ARPA:barwise.pa@Xerox.ARPA 	Robin Cooper's Title   
C00235 00035	∂14-Aug-85  0958	@MIT-MC.ARPA:Vijay.Saraswat@CMU-CS-C.ARPA 	On logic programming as a foundation for AI.   
C00253 00036	∂14-Aug-85  1107	@MIT-MC.ARPA:Laws@SRI-AI.ARPA 	Challenge Problem  
C00258 00037	∂14-Aug-85  1258	NET-ORIGIN@MIT-MC.ARPA 	Re:  Prolog will fail as the foundation for AI so will LOGIC as a PROGRAMMING Language    
C00265 00038	∂14-Aug-85  1614	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH break -- for next three weeks  
C00267 00039	∂14-Aug-85  1743	EMMA@SU-CSLI.ARPA 	Newsletter August 15, No. 41   
C00279 00040	∂15-Aug-85  0819	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	THE SEVENTH THEORY DAY at Columbia University
C00282 00041	∂15-Aug-85  0831	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	my new address 
C00284 00042	∂15-Aug-85  1425	ullman@diablo 	papers received
C00286 00043	∂16-Aug-85  0125	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #12  
C00294 00044	∂16-Aug-85  0925	ullman@diablo 	a random thought before I forget it
C00296 00045	∂16-Aug-85  1010	PAPA@SU-SCORE.ARPA 	Re: a random thought before I forget it 
C00298 00046	∂17-Aug-85  1306	BRAD@SU-CSLI.ARPA 	System clock    
C00299 00047	∂18-Aug-85  1216	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	birth announcement  
C00301 00048	∂18-Aug-85  1505	@SU-SUSHI.ARPA:ANDERSON@SU-SCORE.ARPA 	Thesis Orals, August 22   
C00303 00049	∂18-Aug-85  1535	ACUFF@SUMEX-AIM.ARPA 	Local BUG-LISPM   
C00305 00050	∂19-Aug-85  0948	@MIT-MC.ARPA:Hewitt@MIT-MC.ARPA 	Prolog will fail as the foundation for AI so will LOGIC as a PROGRAMMING Language
C00318 00051	∂19-Aug-85  1005	ullman@diablo 	events this week    
C00321 00052	∂19-Aug-85  1516	avg@diablo 	linear approximations  
C00327 00053	∂19-Aug-85  1620	@MIT-MC.ARPA:august@JPL-VLSI.ARPA 	Got to keep trying! Sorry about all the copies.   
C00330 00054	∂19-Aug-85  2011	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #36
C00362 00055	∂20-Aug-85  1240	@SU-SUSHI.ARPA:ANDERSON@SU-SCORE.ARPA 	Complexity Year info 
C00365 00056	∂20-Aug-85  2251	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	"Official" FOCS airline  
C00369 00057	∂20-Aug-85  2302	ullman@diablo  
C00372 00058	∂20-Aug-85  2302	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Roommates for FOCS  
C00376 00059	∂21-Aug-85  0912	EMMA@SU-CSLI.ARPA 	Ventura Hall and air conditioning   
C00378 00060	∂21-Aug-85  0925	ullman@diablo 	don't forget   
C00379 00061	∂21-Aug-85  1158	STUCKY@SU-CSLI.ARPA 	D-lion Users  
C00382 00062	∂21-Aug-85  1658	ullman@diablo 	paper received 
C00383 00063	∂21-Aug-85  1736	EMMA@SU-CSLI.ARPA 	Newsletter August 22, No. 42   
C00394 00064	∂21-Aug-85  2018	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:Robert←S.←Printis.osbunorth@Xerox.ARPA 	Re: Complexity Year info 
C00396 00065	∂22-Aug-85  1609	MARJORIE@SU-CSLI.ARPA 	phone lines 
C00397 00066	∂22-Aug-85  1651	JAMIE@SU-CSLI.ARPA 	security  
C00398 00067	∂22-Aug-85  2229	JAMIE@SU-CSLI.ARPA 	security  
C00400 00068	∂23-Aug-85  0905	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Messages to SIAM    
C00402 00069	∂25-Aug-85  0130	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #13  
C00408 00070	∂25-Aug-85  1031	DAVIES@SUMEX-AIM.ARPA 	Vacation    
C00409 00071	∂26-Aug-85  0059	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #36
C00441 00072	∂26-Aug-85  0845	EMMA@SU-CSLI.ARPA 	Mailing lists   
C00442 00073	∂26-Aug-85  1033	ACUFF@SUMEX-AIM.ARPA 	Explorer
C00444 00074	∂26-Aug-85  1141	STUCKY@SU-CSLI.ARPA 	D-Lion Users  
C00446 00075	∂26-Aug-85  1434	ullman@diablo 	surprise visit 
C00448 00076	∂30-Aug-85  1301	PATASHNIK@SU-SUSHI.ARPA 	[karp%ucbernie@Berkeley (Richard Karp):]
C00450 00077	∂30-Aug-85  1324	ullman@diablo 	meeting today  
C00451 00078	∂30-Aug-85  1327	chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--September 3rd   
C00458 00079	∂30-Aug-85  1342	HANS@SU-CSLI.ARPA 	talk by Uwe Reyle    
C00459 00080	∂30-Aug-85  1348	EMMA@SU-CSLI.ARPA 	parking stickers
C00461 00081	∂30-Aug-85  1355	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--September 3rd    
C00468 00082	∂30-Aug-85  1407	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	moving    
C00471 00083	∂30-Aug-85  1409	EMMA@SU-CSLI.ARPA 	Re: parking stickers 
C00473 00084	∂30-Aug-85  1427	EMMA@SU-CSLI.ARPA 	Newsletter August 29, No. 43   
C00483 00085	∂30-Aug-85  1626	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.3, 30 Aug 85  
C00495 00086	∂01-Sep-85  2356	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	IJPP: new journal on parallel programming    
C00504 00087	∂02-Sep-85  1406	@SU-SUSHI.ARPA:PAPA@SU-SCORE.ARPA 	Ride for Berkeley   
C00505 00088	∂02-Sep-85  2235	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.4, 02 Sep 85  
C00520 00089	∂03-Sep-85  1004	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	call for papers - MIT conference on VLSI
C00528 00090	∂03-Sep-85  1530	ullman@diablo 	paper received 
C00530 00091	∂03-Sep-85  1654	RICE@SUMEX-AIM.ARPA 	Wednesday's Meeting
C00532 00092	∂03-Sep-85  1815	BRAD@SU-CSLI.ARPA 	vacation   
C00533 00093	∂04-Sep-85  1038	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
C00536 00094	∂04-Sep-85  1346	chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept 10    
C00541 00095	∂04-Sep-85  1354	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept 10
C00546 00096	∂04-Sep-85  1408	STUCKY@SU-CSLI.ARPA 	Dlion-users   
C00548 00097	∂04-Sep-85  1736	EMMA@SU-CSLI.ARPA 	Newsletter September 5, No. 44 
C00563 00098	∂04-Sep-85  2123	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.5, 04 Sep 85  
C00587 00099	∂05-Sep-85  1640	EMMA@SU-CSLI.ARPA 	New project mailing lists 
C00591 00100	∂06-Sep-85  0048	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.6, 06 Sep 85  
C00613 00101	∂08-Sep-85  0149	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.7, 08 Sep 85  
C00659 00102	∂08-Sep-85  1840	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.8, 08 Sep 85, 2nd issue today!    
C00692 00103	∂09-Sep-85  1257	YAMARONE@SU-CSLI.ARPA 	ONE EXTRA SANDWICH:TURKEY/LT.RYE
C00693 00104	∂09-Sep-85  1337	SUSI@SU-CSLI.ARPA   
C00694 00105	∂09-Sep-85  1417	CAROL@SU-CSLI.ARPA 	DANDELION ASSISTANCE
C00697 00106	∂09-Sep-85  1517	@SU-CSLI.ARPA:GEORGEFF@SRI-AI.ARPA 	PLANLUNCH resumes next week  
C00699 00107	∂09-Sep-85  2134	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.9, 09 Sep 85  
C00712 00108	∂09-Sep-85  2347	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #14  
C00736 00109	∂10-Sep-85  1014	RICE@SUMEX-AIM.ARPA 	Tomorrow's Meeting.
C00737 00110	∂10-Sep-85  1402	RICHARDSON@SU-SCORE.ARPA 	Sr. Faculty Meeting
C00738 00111	∂10-Sep-85  2130	ullman@diablo 	weird message  
C00739 00112	∂11-Sep-85  1239	BETSY@SU-CSLI.ARPA 	CSLI Activities in 1985-86    
C00748 00113	∂11-Sep-85  1741	EMMA@SU-CSLI.ARPA 	Newsletter September 12, No. 45
C00757 00114	∂11-Sep-85  1742	BRESNAN@SU-CSLI.ARPA 	Valentina Zavarin 
C00759 00115	∂12-Sep-85  0323	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:klawe.sjrlvm1%ibm-sj.csnet@csnet-relay.arpa 	talk on razborov's theorem    
C00761 00116	∂12-Sep-85  0905	chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept. 17   
C00766 00117	∂12-Sep-85  0908	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept. 17    
C00772 00118	∂12-Sep-85  0932	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	FOCS reminders 
C00776 00119	∂12-Sep-85  1208	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Next week's PLANLUNCH -- notice change in date, place   
C00780 00120	∂12-Sep-85  1353	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp%ucbernie@Berkeley   
C00783 00121	∂12-Sep-85  1422	BARWISE@SU-CSLI.ARPA 	TGBBQ   
C00784 00122	∂12-Sep-85  1424	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.10  
C00785 00123	∂12-Sep-85  1540	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.10  
C00801 00124	∂12-Sep-85  1628	NEUMANN@SRI-CSLA.ARPA 	The foregoing mailings of RISKS-1.10 
C00804 00125	∂12-Sep-85  1750	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa  
C00807 00126	∂13-Sep-85  0151	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.11  
C00832 00127	∂13-Sep-85  0307	NET-ORIGIN@MIT-MC.ARPA 	Enter on BBoard 
C00835 00128	∂13-Sep-85  0855	ullman@su-aimvax.arpa 	meeting
C00836 00129	∂13-Sep-85  1117	BETSY@SU-CSLI.ARPA 	New Charge for Specially Processed Checks    
C00838 00130	∂13-Sep-85  1209	@MIT-MC.ARPA:Ghenis.pasa@XEROX.ARPA 	Is the Phi of Sci list alive?    
C00840 00131	∂13-Sep-85  1218	YAMARONE@SU-CSLI.ARPA 	TGIF TODAY @ 4:00
C00845 00132	∂13-Sep-85  1302	YAMARONE@SU-CSLI.ARPA 	ONE EXTRA TURKEY SANDWICH:2.50  
C00846 00133	∂13-Sep-85  1505	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.12  
C00862 00134	∂13-Sep-85  1636	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.12  
C00878 00135	∂15-Sep-85  0339	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.13  
C00898 00136	∂15-Sep-85  1200	CS.AVI@R20.UTEXAS.EDU 	PODS call for paper   
C00904 00137	∂15-Sep-85  1251	@MIT-MC.ARPA:7thSon@SCRC-STONY-BROOK.ARPA 	philosophy of science 
C00906 00138	∂15-Sep-85  1307	@MIT-MC.ARPA:7thSon@SCRC-STONY-BROOK.ARPA 	philosophy of science 
C00908 00139	∂16-Sep-85  0135	@SUMEX-AIM.ARPA:SAMUEL@SU-SUSHI.ARPA 	SIGBIG 
C00912 00140	∂16-Sep-85  1004	EMMA@SU-CSLI.ARPA 	CSLI talk  
C00915 00141	∂16-Sep-85  2057	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp%ucbernie@Berkeley   
C00917 00142	∂16-Sep-85  2307	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.14  
C00939 00143	∂16-Sep-85  2355	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #15  
C00946 00144	∂17-Sep-85  0027	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #16  
C00957 00145	∂17-Sep-85  0911	ullman@su-aimvax.arpa 	meeting
C00959 00146	∂17-Sep-85  0918	ullman@su-aimvax.arpa 	paper recieved   
C00960 00147	∂17-Sep-85  0922	RICHARDSON@SU-SCORE.ARPA 	Sr. Faculty Meeting September 24  
C00962 00148	∂17-Sep-85  1140	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH REMINDER -- Wednesday, 11AM    
C00965 00149	∂17-Sep-85  1523	ullman@su-aimvax.arpa 	papers received  
C00966 00150	∂17-Sep-85  1622	MARJORIE@SU-CSLI.ARPA 	Books  
C00967 00151	∂17-Sep-85  2008	PIERRE@SU-CSLI.ARPA 	Books    
C00968 00152	∂18-Sep-85  0133	ACUFF@SUMEX-AIM.ARPA 	Explorer Mailing Lists 
C00970 00153	∂18-Sep-85  1536	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH   
C00974 00154	∂18-Sep-85  1615	SUSI@SU-CSLI.ARPA 	opportunity to share 
C00975 00155	∂18-Sep-85  2111	davies%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar, Sept. 24, 1985   
C00981 00156	∂18-Sep-85  2112	@SU-CSLI.ARPA:davies%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar, Sept. 24, 1985    
C00987 00157	∂19-Sep-85  0850	EMMA@SU-CSLI.ARPA 	Newsletter September 19, No. 46
C01001 00158	∂20-Sep-85  0040	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	separating DL and NL using an oracle ?  
C01004 00159	∂20-Sep-85  1019	VSINGH@SUMEX-AIM.ARPA 	Comments    
C01006 00160	∂20-Sep-85  1211	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.15  
C01029 00161	∂20-Sep-85  1248	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
C01033 00162	∂20-Sep-85  1751	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Final FOCS Forewarning   
C01036 00163	∂20-Sep-85  2106	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	parallel machine bibliography 
C01039 00164	∂21-Sep-85  1550	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #17  
C01051 00165	∂22-Sep-85  2315	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH reminder   
C01055 00166	∂23-Sep-85  0342	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #18  
C01077 00167	∂23-Sep-85  0651	PATASHNIK@SU-SUSHI.ARPA 	AFLB resumes   
C01079 00168	∂23-Sep-85  0847	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Reminder, COming up this Friday    
C01082 00169	∂23-Sep-85  1005	RICHARDSON@SU-SCORE.ARPA 	Sr. Faculty Meeting
C01083 00170	∂23-Sep-85  2016	PATASHNIK@SU-SUSHI.ARPA 	[Andrew Yao <YAO@SU-SCORE.ARPA>: Lower Bound Course]   
C01085 00171	∂24-Sep-85  1119	HAUNGA@SUMEX-AIM.ARPA 	Title and abstract for the SIGLUNCH, September 27th.
C01089 00172	∂24-Sep-85  1144	REULING@SU-SCORE.ARPA 	LOTS Registration
C01091 00173	∂24-Sep-85  1144	@SU-CSLI.ARPA:KAELBLING@SRI-AI.ARPA 	Course Announcement    
C01094 00174	∂24-Sep-85  1209	@SU-SCORE.ARPA:TW@SU-AI.ARPA 	Program for this year's computer forum  
C01098 00175	∂24-Sep-85  1317	MARJORIE@SU-CSLI.ARPA 	CSLI Computing pamphlet    
C01100 00176	∂24-Sep-85  1346	LIBRARY@SU-SCORE.ARPA 	Socrates:  New Forms, What To Do If You Misplaced Your Old Account 
C01103 00177	∂24-Sep-85  1400	BSCOTT@SU-SCORE.ARPA 	Salaries
C01105 00178	∂24-Sep-85  1412	DAVIES@SUMEX-AIM.ARPA 	No architecture meeting this week    
C01106 00179	∂24-Sep-85  1544	LIBRARY@SU-SCORE.ARPA 	Math/CS Library:  New Books
C01108 00180	∂24-Sep-85  1722	@SU-CSLI.ARPA:KAELBLING@SRI-AI.ARPA 	Room Change  
C01109 00181	∂24-Sep-85  1941	@SU-SCORE.ARPA:fournier@su-navajo.arpa 	Re:  Socrates:  New Forms, What To Do If You Misplaced Your Old Account    
C01111 00182	∂24-Sep-85  1951	@SU-SCORE.ARPA:fournier@su-navajo.arpa 	Re:  Socrates:  New Forms, What To Do If You Misplaced Your Old Account    
C01113 00183	∂25-Sep-85  0104	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #37
C01145 00184	∂25-Sep-85  1133	CONTRERAS@SU-SCORE.ARPA 	Parking Stickers    
C01146 00185	∂25-Sep-85  1202	@SU-SCORE.ARPA:MDIXON@SU-SUSHI.ARPA 	maps to new student lunch   
C01148 00186	∂25-Sep-85  1353	CHEADLE@SU-SCORE.ARPA 	Reminder--CSD Reception tomorrow
C01150 00187	∂25-Sep-85  1408	REGES@SU-SCORE.ARPA 	Department Lecture Series    
C01152 00188	∂25-Sep-85  1412	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 1  
C01157 00189	∂25-Sep-85  1738	EMMA@SU-CSLI.ARPA 	Newsletter September 26, No. 47
C01179 00190	∂25-Sep-85  2035	JF@SU-SUSHI.ARPA 	BATS at Berkeley 10/11/85  
C01180 00191	∂26-Sep-85  0059	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #38
C01206 00192	∂26-Sep-85  0751	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #39
C01222 00193	∂26-Sep-85  1048	@SU-SUSHI.ARPA:MAYR@SU-SCORE.ARPA 	talk by Marc Snir   
C01225 00194	∂26-Sep-85  1539	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.16  
C01249 00195	∂26-Sep-85  1704	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Monday's PLANLUNCH   
C01253 00196	∂26-Sep-85  2229	sinha%gmr.csnet@CSNET-RELAY.ARPA 	add sinha@gmr to NAIL
C01254 00197	∂27-Sep-85  0153	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #40
C01276 00198	∂27-Sep-85  0902	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.17  
C01300 00199	∂27-Sep-85  0930	JOHN@SU-CSLI.ARPA 	Philosophy 287A 
C01302 00200	∂27-Sep-85  0934	BETSY@SU-CSLI.ARPA 	1986 Site Visit
C01304 00201	∂27-Sep-85  1554	BSCOTT@SU-SCORE.ARPA 	Lost Articles, Yesterday's Reception  
C01306 00202	∂27-Sep-85  1606	LIBRARY@SU-SCORE.ARPA 	Math/CS Library Fall Hours 
C01307 00203	∂27-Sep-85  1615	LIBRARY@SU-SCORE.ARPA 	Socrates: New Version To Be Activated On Sunday, September 29th    
C01310 00204	∂29-Sep-85  1439	@SU-SCORE.ARPA:lantz@su-gregorio.arpa 	Re: speaking at orientation day
C01313 00205	∂30-Sep-85  0157	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #41
C01337 00206	∂30-Sep-85  0221	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #19  
C01360 00207	∂30-Sep-85  0829	PATASHNIK@SU-SUSHI.ARPA 	Abstracts 
C01365 00208	∂30-Sep-85  0935	RICHARDSON@SU-SCORE.ARPA 	Tuesday Lunch Series    
C01366 00209	∂30-Sep-85  1359	DAVIES@SUMEX-AIM.ARPA 	No meeting this week  
C01368 00210	∂30-Sep-85  1417	NILSSON@SU-SCORE.ARPA 	Course Schedule Policy
C01372 00211	∂30-Sep-85  1430	MAYR@SU-SCORE.ARPA 	topics for prob. theory course
C01375 00212	∂30-Sep-85  1828	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
C01378 00213	∂30-Sep-85  2209	ACUFF@SUMEX-AIM.ARPA 	Explorer meets TCP
C01380 00214	∂01-Oct-85  0855	RICHARDSON@SU-SCORE.ARPA 	Tuesday Lunch Series    
C01381 00215	∂01-Oct-85  0903	JF@SU-SUSHI.ARPA 	Still need stanford speaker for 10/11/85 bats  
C01382 00216	∂01-Oct-85  0922	@SU-SCORE.ARPA:SHORTLIFFE@SUMEX-AIM.ARPA 	Re: topics for prob. theory course    
C01385 00217	∂01-Oct-85  1017	CLT  	Seminar on Logic and the Foundations of Mathematics   
C01387 00218	∂01-Oct-85  1039	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar on Logic and the Foundations of Mathematics    
C01389 00219	∂01-Oct-85  1150	ullman@su-aimvax.arpa 	meeting tomorrow 
C01390 00220	∂01-Oct-85  1227	avg@su-aimvax.arpa 	Re:  meeting tomorrow    
C01391 00221	∂01-Oct-85  1340	HAUNGA@SUMEX-AIM.ARPA 	Title for Siglunch - Friday Oct 4    
C01392 00222	∂01-Oct-85  1525	LIBRARY@SU-SCORE.ARPA 	Socrates: SELECT 9 for Technical Reports  
C01395 00223	∂01-Oct-85  1600	RICHARDSON@SU-SCORE.ARPA 	Pre-Colloquium Cookie Time   
C01396 00224	∂01-Oct-85  1636	MARJORIE@SU-CSLI.ARPA 	key    
C01397 00225	∂01-Oct-85  1723	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Monday Planlunch Reminder!
C01401 00226	∂01-Oct-85  1726	POLLARD@SU-CSLI.ARPA 	new entity   
C01402 00227	∂02-Oct-85  0920	JOHN@SU-CSLI.ARPA 	Wheels for Wheels    
C01405 00228	∂02-Oct-85  1030	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ucbernie.Berkeley.EDU    
C01408 00229	∂02-Oct-85  1248	admin@ucbcogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 8 (Terry Winograd, Stanford) 
C01414 00230	∂02-Oct-85  1248	@SU-CSLI.ARPA:admin@ucbcogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 8 (Terry Winograd, Stanford)  
C01420 00231	∂02-Oct-85  1305	CHEADLE@SU-SCORE.ARPA 	Hopeful Ph.D. Grads   
C01422 00232	∂02-Oct-85  1323	LIBRARY@SU-SCORE.ARPA 	Socrates:  Display to account--SUSHI problem   
C01424 00233	∂02-Oct-85  1352	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
C01428 00234	∂02-Oct-85  1410	LIBRARY@SU-SCORE.ARPA 	Math/CS Library: New Reference Books--Guide to CS literature, Industrial Robotics Handbook, Computer-Readable Databas
C01430 00235	∂02-Oct-85  1524	STUCKY@SU-CSLI.ARPA 	Bibliography Files 
C01433 00236	∂02-Oct-85  1713	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #20  
C01441 00237	∂02-Oct-85  1726	EMMA@SU-CSLI.ARPA 	addendum to the newsletter
C01442 00238	∂02-Oct-85  1744	EMMA@SU-CSLI.ARPA 	Newsletter October 3, No. 48   
C01451 00239	∂03-Oct-85  0750	HAUNGA@SUMEX-AIM.ARPA 	title and abstract for SIGLUNCH Friday 4th.    
C01454 00240	∂03-Oct-85  1038	NILSSON@SU-SCORE.ARPA 	5-year Plan 
C01457 00241	∂03-Oct-85  1236	WINOGRAD@SU-CSLI.ARPA 	Environments group - Monday 12:00pm  
C01462 00242	∂03-Oct-85  1354	@SU-SUSHI.ARPA:broder@decwrl.ARPA 	Plane ticket swap wanted 
C01464 00243	∂04-Oct-85  1014	ACUFF@SUMEX-AIM.ARPA 	Symbolics Machines available at some ACM Conference  
C01466 00244	∂04-Oct-85  1214	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Next week's PLANLUNCH --- WEDNESDAY (not monday) OCT.9  
C01471 00245	∂04-Oct-85  1358	JAMIE@SU-CSLI.ARPA 	graduate student offices 
C01473 00246	∂04-Oct-85  1508	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
C01475 00247	∂04-Oct-85  1554	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.18  
C01498 00248	∂05-Oct-85  1030	JF@SU-SUSHI.ARPA 	Abstracts and other Information about 10/11/85 BATS 
C01508 00249	∂05-Oct-85  1701	@SU-CSLI.ARPA:barwise@csli-whitehead 	Linguistics Positions 
C01510 00250	∂07-Oct-85  0032	DAVIES@SUMEX-AIM.ARPA 	This week's meeting   
C01512 00251	∂07-Oct-85  0632	PATASHNIK@SU-SUSHI.ARPA 	Next AFLB 
C01515 00252	∂07-Oct-85  0815	EMMA@SU-CSLI.ARPA 	Logic Lunch
C01517 00253	∂07-Oct-85  0819	HAUNGA@SUMEX-AIM.ARPA 	Title & abstract for the SIGLUNCH, Friday 11th.
C01520 00254	∂07-Oct-85  0921	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
C01521 00255	∂07-Oct-85  0949	ullman@su-aimvax.arpa 	meeting
C01523 00256	∂07-Oct-85  1026	ullman@su-aimvax.arpa 	paper just received   
C01525 00257	∂07-Oct-85  1444	WASHINGTON@SUMEX-AIM.ARPA 	Siglunch location 
C01527 00258	∂07-Oct-85  1841	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar on Logic and the Foundations of Mathematics    
C01530 00259	∂07-Oct-85  1850	CLT  	Seminar on COMMON SENSE AND NON-MONOTONIC REASONING   
C01536 00260	∂08-Oct-85  0011	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	ICDT 86 - Call For Papers 
C01543 00261	∂08-Oct-85  0111	NEUMANN@SRI-CSL.ARPA 	RISKS-1.19   
C01562 00262	∂08-Oct-85  0912	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
C01563 00263	∂08-Oct-85  0936	@SU-SCORE.ARPA:TW@SU-AI.ARPA 	Reminder about forum talks    
C01565 00264	∂08-Oct-85  1043	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
C01566 00265	∂08-Oct-85  1424	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH mailing list    
C01568 00266	∂08-Oct-85  1724	ullman@su-aimvax.arpa 	meeting
C01569 00267	∂08-Oct-85  1756	BSCOTT@SU-SCORE.ARPA 	No-Smoking Policy 
C01572 00268	∂08-Oct-85  1910	NEUMANN@SRI-CSL.ARPA 	RISKS-1.20   
C01594 00269	∂09-Oct-85  0008	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #21  
C01607 00270	∂09-Oct-85  0155	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #42
C01631 00271	∂09-Oct-85  0954	HANRAHAN@SU-SCORE.ARPA 	petty cash 
C01632 00272	∂09-Oct-85  1024	@SU-SCORE.ARPA:ejm@su-shasta.arpa 	Re:  topics for prob. theory course
C01633 00273	∂09-Oct-85  1113	@SU-SCORE.ARPA:ejm@su-shasta.arpa 	Re:  petty cash
C01634 00274	∂09-Oct-85  1349	DAVIES@SUMEX-AIM.ARPA 	Alliant -- Computer Systems Seminar Today!!    
C01636 00275	∂09-Oct-85  1648	admin@ucbcogsci.Berkeley.EDU 	Cognitive Science Seminar--Oct. 15 (Ron Kaplan, Xerox PARC & Stanford)
C01643 00276	∂09-Oct-85  1656	@SU-CSLI.ARPA:admin@ucbcogsci.Berkeley.EDU 	Cognitive Science Seminar--Oct. 15 (Ron Kaplan, Xerox PARC & Stanford) 
C01650 00277	∂09-Oct-85  1703	EMMA@SU-CSLI.ARPA 	Newsletter October 10, No. 49  
C01678 00278	∂09-Oct-85  1705	AAAI-OFFICE@SUMEX-AIM.ARPA 	Cog Science Society Proposal    
C01684 00279	∂09-Oct-85  1915	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	Call for Papers, Spring'86 CUG   
C01688 00280	∂09-Oct-85  2015	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ucbernie.Berkeley.EDU    
C01693 00281	∂10-Oct-85  0904	LIBRARY@SU-SCORE.ARPA 	Socrates: Are people having problems when they attempt to telnet to Socrates?
C01695 00282	∂10-Oct-85  1107	DELAGI@SUMEX-AIM.ARPA 	"Volunteers"
C01698 00283	∂10-Oct-85  1357	LIBRARY@SU-SCORE.ARPA 	[The Mailer Daemon <Mailer@SU-SCORE.ARPA>: Message of 10-Oct-85 13:49:31]    
C01702 00284	∂10-Oct-85  1408	CAROL@SU-CSLI.ARPA 	New office
C01703 00285	∂11-Oct-85  0042	NEUMANN@SRI-CSL.ARPA 	RISKS-1.21   
C01737 00286	∂11-Oct-85  0804	BRESNAN@SU-CSLI.ARPA 	announcement of MSDI presentations    
C01739 00287	∂11-Oct-85  1315	PATASHNIK@SU-SUSHI.ARPA 	Speaker for October 17   
C01740 00288	∂11-Oct-85  1502	LIBRARY@SU-SCORE.ARPA 	Socrates:  Workshops to be held in the Green Library
C01742 00289	∂11-Oct-85  2217	BRESNAN@SU-CSLI.ARPA 	presentation by Peter Sells 
C01743 00290	∂12-Oct-85  0800	PATASHNIK@SU-SUSHI.ARPA 	AFLB Abstract  
C01746 00291	∂14-Oct-85  0755	@SUMEX-AIM.ARPA:HAUNGA@SU-SCORE.ARPA 	Title & Abstract for the Siglunch Friday 18th. 
C01749 00292	∂14-Oct-85  0851	RICHARDSON@SU-SCORE.ARPA 	Tuesday CSD Lunch  
C01750 00293	∂14-Oct-85  0907	@SUMEX-AIM.ARPA:HAUNGA@SU-SCORE.ARPA 	SPEAKER for Friday 18th SIGLUNCH is David Heckerman.
C01751 00294	∂14-Oct-85  1231	CAROL@SU-CSLI.ARPA 	Phone number   
C01752 00295	∂14-Oct-85  1557	ACUFF@SUMEX-AIM.ARPA 	Explorer Update   
C01753 00296	∂14-Oct-85  1849	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #22  
C01781 00297	∂15-Oct-85  0820	DAVIES@SUMEX-AIM.ARPA 	Meeting at 9:30 Wednesday  
C01782 00298	∂15-Oct-85  0829	RICHARDSON@SU-SCORE.ARPA 	Tuesday Lunch Series    
C01783 00299	∂15-Oct-85  0932	ullman@su-aimvax.arpa 	meeting
C01785 00300	∂15-Oct-85  1103	RICHARDSON@SU-SCORE.ARPA 	Tuesday CSD Lunch Series
C01786 00301	∂22-Oct-85  1545	BARWISE@SU-CSLI.ARPA 	Logic Lunch  
C01787 00302	∂22-Oct-85  1828	HANRAHAN@SU-SCORE.ARPA 	petty cash 
C01788 00303	∂22-Oct-85  1859	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar on Logic and the Foundations of Mathematics    
C01790 00304	∂22-Oct-85  1908	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #43
C01817 00305	∂22-Oct-85  1911	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C01818 00306	∂22-Oct-85  1915	ullman@su-aimvax.arpa 	meeting this week
C01820 00307	∂22-Oct-85  1916	SELLS@SU-CSLI.ARPA 	Hinrichs dissertation    
C01821 00308	∂22-Oct-85  1919	BRESNAN@SU-CSLI.ARPA 	"Topic, Pronoun, and Agreement in Chichewa"
C01822 00309	∂22-Oct-85  1919	LIBRARY@SU-SCORE.ARPA 	Math/CS Library:  Electronic Services--If New To Stanford Please Read   
C01826 00310	∂22-Oct-85  1921	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.Berkeley.EDU  
C01828 00311	∂22-Oct-85  1923	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #24  
C01840 00312	∂22-Oct-85  2055	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
C01841 00313	∂22-Oct-85  2056	RICHARDSON@SU-SCORE.ARPA 	[The Mailer Daemon <Mailer@SU-SCORE.ARPA>: Message of 17-Oct-85 10:40:56] 
C01844 00314	∂22-Oct-85  2058	@SU-SCORE.ARPA:WINOGRAD@SU-CSLI.ARPA 	Postponement of PhD program meeting  
C01846 00315	∂23-Oct-85  0609	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
C01850 00316	∂23-Oct-85  0749	@SU-CSLI.ARPA:RPERRAULT@SRI-AI.ARPA 	Talk by Bill Rounds    
C01853 00317	∂23-Oct-85  1048	CONTRERAS@SU-SCORE.ARPA 	Class Lists    
C01854 00318	∂23-Oct-85  1103	MODICA@SU-SCORE.ARPA 	Class Lists - Correction    
C01855 00319	∂23-Oct-85  1133	CONTRERAS@SU-SCORE.ARPA 	Mailboxes 
C01856 00320	∂23-Oct-85  1303	YAMARONE@SU-CSLI.ARPA 	R/B + swiss Available Now at Front Desk   
C01857 00321	∂23-Oct-85  1432	ullman@su-aimvax.arpa 	financial detail 
C01859 00322	∂23-Oct-85  1611	admin@cogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 29, 1985    
C01864 00323	∂23-Oct-85  1633	@SU-CSLI.ARPA:admin@cogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 29, 1985
C01869 00324	∂23-Oct-85  1733	EMMA@SU-CSLI.ARPA 	Newsletter October 24, No. 51  
C01893 00325	∂24-Oct-85  0110	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #25  
C01900 00326	∂24-Oct-85  0819	EMMA@SU-CSLI.ARPA 	Today's CSLI seminar 
C01901 00327	∂24-Oct-85  0847	AAAI-OFFICE@SUMEX-AIM.ARPA 	New Tutorial Chair    
C01903 00328	∂24-Oct-85  1436	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
C01906 00329	∂24-Oct-85  1514	ullman@su-aimvax.arpa 	Dave Smith's conjecture    
C01913 00330	∂24-Oct-85  1602	BETSY@SU-CSLI.ARPA 	Visitor Policy 
C01919 00331	∂24-Oct-85  1724	NILSSON@SU-SCORE.ARPA 	[Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]
C01922 00332	∂24-Oct-85  2129	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #26  
C01931 00333	∂24-Oct-85  2231	BRESNAN@SU-CSLI.ARPA 	"Case-Assignment by Nominals in Japanese"  
C01934 00334	∂25-Oct-85  0030	DAVIES@SUMEX-AIM.ARPA 	Wednesday meeting -- 9 am  
C01935 00335	∂25-Oct-85  0935	BARWISE@SU-CSLI.ARPA 	Logic Seminar Cancelled today    
C01936 00336	∂25-Oct-85  0938	AAAI-OFFICE@SUMEX-AIM.ARPA 	Mark's confirmation   
C01937 00337	∂25-Oct-85  1018	INGRID@SU-CSLI.ARPA 	House for Rent
C01939 00338	∂25-Oct-85  1032	INGRID@SU-CSLI.ARPA 	House for Rent
C01941 00339	∂25-Oct-85  1036	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	todays logic seminar is cancelled  
C01942 00340	∂25-Oct-85  1036	RICHARDSON@SU-SCORE.ARPA 	November Sr. Faculty Meeting 
C01943 00341	∂25-Oct-85  1106	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]  
C01945 00342	∂25-Oct-85  1112	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]  
C01947 00343	∂25-Oct-85  1118	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]  
C01949 00344	∂25-Oct-85  1304	INGRID@SU-CSLI.ARPA 	Internal Faculty Fellowships 
C01951 00345	∂25-Oct-85  1433	ullman@su-aimvax.arpa 	papers received  
C01952 00346	∂25-Oct-85  1559	MODICA@SU-SCORE.ARPA 	Class Lists  
C01953 00347	∂26-Oct-85  1019	BRESNAN@SU-CSLI.ARPA 	"Case-Assignment by Nominals in Japanese"  
C01954 00348	∂27-Oct-85  1418	NILSSON@SU-SCORE.ARPA 	Student names    
C01956 00349	∂27-Oct-85  1447	NILSSON@SU-SCORE.ARPA 	Committees  
C01973 00350	∂27-Oct-85  1849	LANSKY@SRI-AI.ARPA 	PLANLUNCH REMINDER - Pat Hayes, Oct. 28 
C01975 00351	∂28-Oct-85  0143	@SU-SCORE.ARPA:manfred%ucsc.csnet@CSNET-RELAY.ARPA 	techreport request!    
C01977 00352	∂28-Oct-85  0834	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
C01978 00353	∂28-Oct-85  0836	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar in Logic and the Foundations of Mathematics    
C01980 00354	∂28-Oct-85  1533	IIDA@SU-CSLI.ARPA 	Talk this Thursday by Masayo Iida   
C01982 00355	∂28-Oct-85  1624	@SU-SCORE.ARPA:pratt@su-navajo.arpa 	Student names
C01984 00356	∂28-Oct-85  2233	@SU-SCORE.ARPA:HADDAD@SU-SUSHI.ARPA 	Re: Student names 
C01988 00357	∂29-Oct-85  0614	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
C01994 00358	∂29-Oct-85  0907	MODICA@SU-SCORE.ARPA 	Class Lists  
C01995 00359	∂29-Oct-85  1011	TREITEL@SUMEX-AIM.ARPA 	Conjunct ordering    
C02003 00360	∂29-Oct-85  1129	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C02004 00361	∂29-Oct-85  1219	ullman@su-aimvax.arpa 	meeting
C02005 00362	∂29-Oct-85  1537	CAROL@SU-CSLI.ARPA 	Demo of Dlion gloss package   
C02008 00363	∂29-Oct-85  1623	EMMA@SU-CSLI.ARPA 	Halloween  
C02011 00364	∂29-Oct-85  2334	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	FST&TCS 5th Conference    
C02022 00365	∂30-Oct-85  0615	PATASHNIK@SU-SUSHI.ARPA 	Special AFLB   
C02026 00366	∂30-Oct-85  0931	PHILOSOPHY@SU-CSLI.ARPA 	Josh Cohen
C02027 00367	∂30-Oct-85  0931	PHILOSOPHY@SU-CSLI.ARPA 	Josh Cohen
C02028 00368	∂30-Oct-85  1055	PHILOSOPHY@SU-CSLI.ARPA  
C02029 00369	∂30-Oct-85  1140	INGRID@SU-CSLI.ARPA 	Announcement  
C02034 00370	∂30-Oct-85  1155	@SU-SUSHI.ARPA:vemuri@ngp.UTEXAS.EDU 	Re:  Special AFLB
C02035 00371	∂30-Oct-85  1639	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH  
C02038 00372	∂30-Oct-85  1732	EMMA@SU-CSLI.ARPA 	Newsletter October 31, No. 52  
C02060 00373	∂31-Oct-85  0917	TAJNAI@SU-SCORE.ARPA 	1985 IBM Party    
C02062 00374	∂31-Oct-85  0926	EMMA@SU-CSLI.ARPA 	Halloween 3
C02065 00375	∂31-Oct-85  1323	EMMA@SU-CSLI.ARPA 	All Hallow's Eve
C02066 00376	∂31-Oct-85  1450	EMMA@SU-CSLI.ARPA 	Halloween tea   
C02067 00377	∂31-Oct-85  1509	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
C02072 00378	∂31-Oct-85  1952	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	VLSI workshop announcement
C02080 00379	∂01-Nov-85  0040	ACUFF@SUMEX-AIM.ARPA 	Explorers working at last   
C02082 00380	∂01-Nov-85  0631	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 5  
C02089 00381	∂01-Nov-85  1030	ullman@su-aimvax.arpa 	talk today  
C02090 00382	∂01-Nov-85  1130	WINOGRAD@SU-CSLI.ARPA 	Lost coat   
C02091 00383	∂01-Nov-85  1203	COLEMAN@SU-CSLI.ARPA 	help the Aussies  
C02093 00384	∂01-Nov-85  1328	@SU-SCORE.ARPA:LES@SU-AI.ARPA 	Computer Facility Needs 
C02100 00385	∂01-Nov-85  1459	CHEADLE@SU-SCORE.ARPA 	Lost glasses
C02102 00386	∂01-Nov-85  1604	RINDFLEISCH@SUMEX-AIM.ARPA 	IBM S.J. Research Computer Science seminars 4 - 8 NOV 85 
C02111 00387	∂01-Nov-85  1644	TAJNAI@SU-SCORE.ARPA 	honors  
C02113 00388	∂01-Nov-85  1656	INGRID@SU-CSLI.ARPA 	House for Rent
C02115 00389	∂01-Nov-85  1752	COLEMAN@SU-CSLI.ARPA
C02116 00390	∂01-Nov-85  1822	ACUFF@SUMEX-AIM.ARPA 	Leaving Explorers 
C02117 00391	∂01-Nov-85  1829	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Call for papers - PODC    
C02122 00392	∂02-Nov-85  1158	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.berkeley.edu  
C02125 00393	∂03-Nov-85  1923	LANSKY@SRI-AI.ARPA 	planlunch reminder -- Monday/11am/David Israel    
C02128 00394	∂03-Nov-85  2051	BARWISE@SU-CSLI.ARPA 	Visitor 
C02130 00395	∂04-Nov-85  0917	INGRID@SU-CSLI.ARPA 	Symposia on AI
C02135 00396	∂04-Nov-85  1002	LIBRARY@SU-SCORE.ARPA 	Applied Numerical Mathematics--New Journal In Math/CS Library 
C02137 00397	∂04-Nov-85  1036	BARWISE@SU-CSLI.ARPA 	Talk on semantics of types in computer languages
C02139 00398	∂04-Nov-85  1140	HOLDEN@SU-CSLI.ARPA 	Friday afternoon drinking sessions
C02140 00399	∂04-Nov-85  1229	ACUFF@SUMEX-AIM.ARPA 	What Explorer files to keep 
C02142 00400	∂04-Nov-85  1250	ACUFF@SUMEX-AIM.ARPA 	Booting Lisp Machines  
C02145 00401	∂04-Nov-85  1436	ULLMAN@SU-SCORE.ARPA 	new contract 
C02146 00402	∂04-Nov-85  1546	GINSBERG@SUMEX-AIM.ARPA 	change to SIGlunch mailing list    
C02162 00403	∂04-Nov-85  1607	ACUFF@SUMEX-AIM.ARPA 	Symbolics shuffling and installation  
C02164 00404	∂04-Nov-85  1618	SJG  	change to SIGlunch mailing list   
C02165 00405	∂04-Nov-85  1651	ullman@su-aimvax.arpa 	more papers received today 
C02166 00406	∂04-Nov-85  1702	ullman@su-aimvax.arpa 	paper received   
C02168 00407	∂04-Nov-85  1703	REGES@SU-SCORE.ARPA 	lunch reminder
C02170 00408	∂04-Nov-85  1720	@SU-SCORE.ARPA:WINOGRAD@SU-CSLI.ARPA 	Forum speakers   
C02174 00409	∂04-Nov-85  2019	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Talk on semantics of types in computer languages  
C02176 00410	∂04-Nov-85  2048	@SU-CSLI.ARPA:dirk@SU-PSYCH 	Psychology Department Friday Seminar.    
C02181 00411	∂04-Nov-85  2138	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	POPL 86    
C02199 00412	∂04-Nov-85  2213	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	BATS at DEC - November 22  
C02201 00413	∂05-Nov-85  0046	DAVIES@SUMEX-AIM.ARPA 	No AAP meeting Wednesday   
C02202 00414	∂05-Nov-85  1148	BERG@SU-SCORE.ARPA 	technical reports   
C02204 00415	∂05-Nov-85  1438	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
C02209 00416	∂05-Nov-85  1519	LIBRARY@SU-SCORE.ARPA 	Socrates:  Searching By Call Number In The Technical Reports File  
C02211 00417	∂05-Nov-85  1543	LIBRARY@SU-SCORE.ARPA 	Socrates: How to Telnet--A Reminder  
C02213 00418	∂05-Nov-85  1551	LIBRARY@SU-SCORE.ARPA 	Socrates: Telnet message correction  
C02214 00419	∂05-Nov-85  1620	@SU-SCORE.ARPA:EVAN@SU-CSLI.ARPA 	Re: Socrates:  Searching By Call Number In The Technical Reports File  
C02216 00420	∂05-Nov-85  1632	@SU-SCORE.ARPA:EVAN@SU-CSLI.ARPA 	Last Message (an apology) 
C02217 00421	∂05-Nov-85  1748	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #27  
C02230 00422	∂05-Nov-85  1914	ullman@su-aimvax.arpa 	meeting tomorrow 
C02231 00423	∂05-Nov-85  2305	ACUFF@SUMEX-AIM.ARPA 	3640 setup and ready for use
C02232 00424	∂06-Nov-85  0631	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
C02238 00425	∂06-Nov-85  1409	ullman@su-aimvax.arpa 	paper received   
C02239 00426	∂06-Nov-85  1742	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH  
C02243 00427	∂06-Nov-85  2010	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	MFCS '86 Call for Papers  
C02252 00428	∂08-Nov-85  1647	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 12 (R. Wilensky, UCB)  
C02258 00429	∂08-Nov-85  1649	@SU-CSLI.ARPA:barwise.pa@Xerox.ARPA 	"Machines and the mental"   
C02260 00430	∂08-Nov-85  1704	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #28  
C02272 00431	∂08-Nov-85  1826	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu   
C02278 00432	∂08-Nov-85  1829	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 12 (R. Wilensky, UCB) 
C02284 00433	∂09-Nov-85  0513	VONHENKE@SRI-CSL.ARPA 	RISKS-1.22  
C02321 00434	∂09-Nov-85  1516	ACUFF@SUMEX-AIM.ARPA 	More Explorer info
C02323 00435	∂09-Nov-85  1822	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #29  
C02329 00436	∂10-Nov-85  1804	@SU-SCORE.ARPA:JMC@SU-AI.ARPA 
C02330 00437	∂10-Nov-85  1916	LANSKY@SRI-AI.ARPA 	PLANLUNCH reminder -- 11AM  Moshe Vardi 
C02334 00438	∂11-Nov-85  0140	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #44
C02343 00439	∂11-Nov-85  0945	NILSSON@SU-SCORE.ARPA 	New Keys    
C02346 00440	∂11-Nov-85  1043	INGRID@SU-CSLI.ARPA 	Symposia on AI
C02350 00441	∂11-Nov-85  1040	NILSSON@SU-SCORE.ARPA 	More "Keys" 
C02352 00442	∂11-Nov-85  1201	INGRID@SU-CSLI.ARPA 	HOUSEMATE WANTED   
C02354 00443	∂11-Nov-85  1402	@SU-CSLI.ARPA:VAL@SU-AI.ARPA 	Non-Monotonic Reasoning Seminar    
C02356 00444	∂11-Nov-85  1455	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #30  
C02370 00445	∂11-Nov-85  1539	CAROL@SU-CSLI.ARPA 	Demo postponed 
C02372 00446	∂11-Nov-85  1541	TAJNAI@SU-SCORE.ARPA 	IBM/Forum Party 11/14/85    
C02374 00447	∂11-Nov-85  1644	PATASHNIK@SU-SUSHI.ARPA 	nonAFLB talk   
C02376 00448	∂11-Nov-85  2307	DAVIES@SUMEX-AIM.ARPA 	No meeting this week  
C02377 00449	∂12-Nov-85  0812	RICHARDSON@SU-SCORE.ARPA 	CSD Lunch
C02378 00450	∂12-Nov-85  0835	EMMA@SU-CSLI.ARPA 	TINLunch   
C02380 00451	∂12-Nov-85  1117	BERGMAN@SU-SCORE.ARPA 	RA support  
C02382 00452	∂12-Nov-85  1118	NILSSON@SU-SCORE.ARPA 	Student Names    
C02385 00453	∂12-Nov-85  1119	NILSSON@SU-SCORE.ARPA 	SOE Committees   
C02388 00454	∂12-Nov-85  1236	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C02389 00455	∂12-Nov-85  1236	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
C02393 00456	∂12-Nov-85  1237	RICHARDSON@SU-SCORE.ARPA 	CSD 500 Colloquium Series    
C02394 00457	∂12-Nov-85  1239	NILSSON@SU-SCORE.ARPA 	Master Keys 
C02396 00458	∂12-Nov-85  1254	ACUFF@SUMEX-AIM.ARPA 	Explorer 1.0.2: anyone need it?  
C02397 00459	∂12-Nov-85  1346	WASOW@SU-CSLI.ARPA 	consulting offer    
C02399 00460	∂12-Nov-85  1448	ullman@su-aimvax.arpa 	paper received   
C02400 00461	∂12-Nov-85  1519	ullman@su-aimvax.arpa 	tomorrow's meeting    
C02401 00462	∂12-Nov-85  1522	MODICA@SU-SCORE.ARPA 	info for TV students   
C02403 00463	∂12-Nov-85  1522	MODICA@SU-SCORE.ARPA 	info for TV students   
C02405 00464	∂12-Nov-85  1519	ullman@su-aimvax.arpa 	another paper    
C02406 00465	∂12-Nov-85  1747	PAPA@SU-SCORE.ARPA 	Re: another paper   
C02408 00466	∂12-Nov-85  1818	ullman@su-aimvax.arpa 	Re: another paper
C02409 00467	∂13-Nov-85  0612	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
C02413 00468	∂13-Nov-85  1026	CAROL@SU-CSLI.ARPA 	Sketch demo    
C02416 00469	∂13-Nov-85  1040	CAROL@SU-CSLI.ARPA 	Sketch demo, correction  
C02419 00470	∂13-Nov-85  1141	NILSSON@SU-SCORE.ARPA 	Herb Grosch 
C02420 00471	∂13-Nov-85  1147	DALRYMPLE@SU-CSLI.ARPA 	party 
C02421 00472	∂13-Nov-85  1353	ACUFF@SUMEX-AIM.ARPA 	Bug reports for Explorers   
C02423 00473	∂13-Nov-85  1403	REGES@SU-SCORE.ARPA 	TAs 
C02425 00474	∂13-Nov-85  1606	NILSSON@SU-SCORE.ARPA 	Fac. Lunch  
C02427 00475	∂13-Nov-85  1704	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	BATS program
C02430 00476	∂13-Nov-85  1749	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #31  
C02437 00477	∂13-Nov-85  1758	EMMA@SU-CSLI.ARPA 	Newsletter November 14, No. 2  
C02456 00478	∂13-Nov-85  1809	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	Driving to Bats  
C02460 00479	∂13-Nov-85  1829	@SU-SCORE.ARPA:coraki!pratt@su-navajo.arpa 	Herb Grosch
C02462 00480	∂13-Nov-85  2319	@SU-SCORE.ARPA:coraki!pratt@su-navajo.arpa 	Fac. Lunch 
C02464 00481	∂14-Nov-85  0002	Operator@SU-SUSHI.ARPA 	bats at dec, 11/22   
C02465 00482	∂14-Nov-85  0830	EMMA@SU-CSLI.ARPA 	Newsletter addition  
C02467 00483	∂14-Nov-85  1109	@SU-SCORE.ARPA:guibas@decwrl.DEC.COM 	Re: info for TV students   
C02469 00484	∂14-Nov-85  1109	@SU-SCORE.ARPA:guibas@decwrl.DEC.COM 	Re: info for TV students   
C02471 00485	∂14-Nov-85  1343	CAROL@SU-CSLI.ARPA 	Meet the Author
C02473 00486	∂14-Nov-85  1405	BSCOTT@SU-SCORE.ARPA 	Proposals    
C02476 00487	∂14-Nov-85  1434	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH plus other business...    
C02482 00488	∂14-Nov-85  1459	PHILOSOPHY@SU-CSLI.ARPA 	Colloquium
C02483 00489	∂14-Nov-85  1515	LANSKY@SRI-AI.ARPA 	oops!
C02484 00490	∂14-Nov-85  2006	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 19 (R. Alterman, UCB) 
C02490 00491	∂14-Nov-85  2017	SF@SU-CSLI.ARPA 	Talk by Rachel Feferman
C02492 00492	∂14-Nov-85  2023	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 19 (R. Alterman, UCB)  
C02498 00493	∂14-Nov-85  2045	@SU-SCORE.ARPA:cbosgd!packard!ihesa!olaf@seismo.CSS.GOV
C02501 00494	∂14-Nov-85  2145	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.berkeley.edu  
C02503 00495	∂15-Nov-85  0734	JF@SU-SUSHI.ARPA 	bats abstracts   
C02504 00496	∂15-Nov-85  0733	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	Abstracts for BATS talks at DEC 
C02511 00497	∂15-Nov-85  0848	ACUFF@SUMEX-AIM.ARPA 	Newest batch of Explorers   
C02513 00498	∂15-Nov-85  0848	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #32  
C02522 00499	∂15-Nov-85  0918	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C02523 00500	∂15-Nov-85  0921	JJW  	MJH-LispM news
C02526 00501	∂16-Nov-85  0750	ullman@su-aimvax.arpa 	Re:  another paper    
C02527 00502	∂16-Nov-85  0751	ark@SALLY.UTEXAS.EDU 	Re:  another paper
C02531 00503	∂16-Nov-85  0752	maier%oregon-grad.csnet@CSNET-RELAY.ARPA 	Re:  another paper
C02533 00504	∂16-Nov-85  0757	ACUFF@SUMEX-AIM.ARPA 	Re: MJH-LispM news
C02534 00505	∂16-Nov-85  0759	NEALE@SU-CSLI.ARPA  
C02536 00506	∂16-Nov-85  0800	ACUFF@SUMEX-AIM.ARPA 	KSL-3640-7 down   
C02537 00507	∂16-Nov-85  0759	RICHARDSON@SU-SCORE.ARPA 	General Faculty Meeting 
C02538 00508	∂16-Nov-85  0801	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #33  
C02547 00509	∂16-Nov-85  0805	PATASHNIK@SU-SUSHI.ARPA 	Upcoming AFLB canceled   
C02548 00510	∂16-Nov-85  0808	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	CS UG program update  
C02552 00511	∂16-Nov-85  1209	BSCOTT@SU-SCORE.ARPA 	Locks   
C02553 00512	∂17-Nov-85  0623	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #34  
C02568 00513	∂17-Nov-85  1850	ullman@su-aimvax.arpa 	Re:  another paper    
C02570 00514	∂17-Nov-85  1851	LANSKY@SRI-AI.ARPA 	PLANLUNCH reminder -- Jeff Finger, 11am 
C02575 00515	∂17-Nov-85  1858	BRESNAN@SU-CSLI.ARPA 	Morphology, Syntax, Discourse Interactions 
C02579 00516	∂17-Nov-85  1859	ACUFF@SUMEX-AIM.ARPA 	Bug Reporting from Explorers
C02581 00517	∂18-Nov-85  1012	WECHSLER@SU-CSLI.ARPA 	XenoneX
C02583 00518	∂18-Nov-85  1018	INGRID@SU-CSLI.ARPA 	Symposium on AI    
C02587 00519	∂19-Nov-85  0827	@SU-SUSHI.ARPA:RUTENBURG@SU-SCORE.ARPA 	ICALP
C02588 00520	∂19-Nov-85  0843	@SU-CSLI.ARPA:BrianSmith.pa@Xerox.ARPA 	A day's mail lost   
C02590 00521	∂19-Nov-85  0845	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #35  
C02613 00522	∂19-Nov-85  0852	cheriton@su-pescadero.ARPA 	Re:  another paper    
C02615 00523	∂19-Nov-85  0856	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	PALMS     
C02616 00524	∂19-Nov-85  0925	RICHARDSON@SU-SCORE.ARPA 	Tuesday CSD Lunch  
C02617 00525	∂19-Nov-85  0938	ullman@su-aimvax.arpa 	meeting
C02618 00526	∂19-Nov-85  0939	LIBRARY@SU-SCORE.ARPA 	Socrates:  Additional Terminal In Math/CS Library   
C02620 00527	∂19-Nov-85  0944	NILSSON@SU-SCORE.ARPA 	Visiting Professorship
C02623 00528	∂19-Nov-85  0945	NEUMANN@SRI-CSL.ARPA 	RISKS-1.23   
C02650 00529	∂19-Nov-85  1014	DAVIES@SUMEX-AIM.ARPA 	No meeting tomorrow, November 20
C02651 00530	∂19-Nov-85  1014	CAROL@SU-CSLI.ARPA 	Sketch demo    
C02653 00531	∂19-Nov-85  1013	LIBRARY@SU-SCORE.ARPA 	Math/CS Library:  New Books--CS 
C02655 00532	∂19-Nov-85  1140	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar in Logic and Foundations   
C02657 00533	∂19-Nov-85  1145	JF@SU-SUSHI.ARPA 	p.s. on BATS 11/22/85 
C02658 00534	∂19-Nov-85  1146	JF@SU-SUSHI.ARPA 	bats headcount   
C02659 00535	∂19-Nov-85  1243	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	course description  
C02664 00536	∂19-Nov-85  1315	EMMA@SU-CSLI.ARPA 	Newsletter 
C02665 00537	∂19-Nov-85  1337	ACUFF@SUMEX-AIM.ARPA 	Explorer Hardware Watch
C02667 00538	∂19-Nov-85  1618	RICHARDSON@SU-SCORE.ARPA 	Faculty Accomplishments 
C02673 00539	∂19-Nov-85  2039	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #36  
C02698 00540	∂20-Nov-85  1153	RICE@SUMEX-AIM.ARPA 	36xx -> Explorer Status.
C02704 00541	∂20-Nov-85  1158	PATASHNIK@SU-SUSHI.ARPA 	Next AFLB 
C02706 00542	∂20-Nov-85  1205	NEUMANN@SRI-CSL.ARPA 	RISKS-1.24   
C02735 00543	∂20-Nov-85  1205	LIBRARY@SU-SCORE.ARPA 	Socrates:  Update On Searching Math/CS Technical Reports By Call Numbers (Accession Numbers)
C02738 00544	∂20-Nov-85  1238	YAMARONE@SU-CSLI.ARPA 	HUNGRY? X-TRA SANDWICHES AVAILABLE   
C02739 00545	∂20-Nov-85  1407	@SU-SUSHI.ARPA:PAPA@SU-SCORE.ARPA 	Talk by Victor Klee 
C02741 00546	∂20-Nov-85  1502	@SU-SUSHI.ARPA:PAPA@SU-SCORE.ARPA 	V. Klee's talk 
C02742 00547	∂20-Nov-85  1506	YAMARONE@SU-CSLI.ARPA 	HUNGRY?? TURKEY & AVO AVAILABLE 
C02744 00548	∂20-Nov-85  1604	@SU-SUSHI.ARPA:SILVERBERG@SU-SCORE.ARPA 	Talk by Victor Klee
C02745 00549	∂20-Nov-85  1814	EMMA@SU-CSLI.ARPA 	Newsletter November 21, No. 3  
C02768 00550	∂21-Nov-85  0933	EMMA@SU-CSLI.ARPA 	Newsletter addition  
C02769 00551	∂21-Nov-85  1100	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	International CS Institute 
C02773 00552	∂21-Nov-85  1134	RICE@SUMEX-AIM.ARPA 	3600 -> Explorer Update.
C02774 00553	∂21-Nov-85  1258	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #37  
C02798 00554	∂21-Nov-85  1525	@MIT-MC.ARPA:JCMA@MIT-MC.ARPA 	Classic Quotes (Positivism Assesses Progress in Meaning)   
C02800 00555	∂21-Nov-85  1541	@MIT-MC.ARPA:JCMA@MIT-MC.ARPA 	Classic Quotes (Positivism Assesses Progress in Meaning)   
C02802 00556	∂21-Nov-85  1639	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH -- Haim Gaifman 
C02809 00557	∂21-Nov-85  1704	ullman@su-aimvax.arpa 	paper recieved   
C02810 00558	∂21-Nov-85  1721	ullman@su-aimvax.arpa 	vowel fans take note  
C02811 00559	∂21-Nov-85  1754	WINOGRAD@SU-CSLI.ARPA 	No ENVIRONMENTS meeting until Dec 9  
C02813 00560	∂21-Nov-85  2104	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	ADVANCED COURSE ON PETRI NETS  
C02819 00561	∂22-Nov-85  1248	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #38  
C02851 00562	∂22-Nov-85  1455	@MIT-MC.ARPA:MERRY.pa@XEROX.ARPA 	Re: Classic Quotes (Positivism Assesses Progress in Meaning) 
C02853 00563	∂22-Nov-85  1902	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Call for Algorithms Papers
C02856 00564	∂22-Nov-85  1941	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
C02859 00565	∂23-Nov-85  0351	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 26 (E. Clark, Stanford)    
C02864 00566	∂23-Nov-85  0352	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 26 (E. Clark, Stanford)
C02869 00567	∂23-Nov-85  1159	ACUFF@SUMEX-AIM.ARPA 	3674 and 3647 status   
C02871 00568	∂23-Nov-85  1233	ACUFF@SUMEX-AIM.ARPA 	Air Conditioning in the Welch Road Machine Room 
C02874 00569	∂23-Nov-85  1255	JOHNSON@SU-CSLI.ARPA 	Intro to Computational Linguistics    
C02876 00570	∂24-Nov-85  1238	LANSKY@SRI-AI.ARPA 	tomorrow's planlunch -- 11AM  
C02878 00571	∂25-Nov-85  0924	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C02879 00572	∂25-Nov-85  1405	BERG@SU-SCORE.ARPA 	evaluation forms    
C02881 00573	∂25-Nov-85  1600	EMMA@SU-CSLI.ARPA 	The Next TINLunch    
C02884 00574	∂25-Nov-85  1920	ZEC@SU-CSLI.ARPA 	linguistcs department colloquium
C02886 00575	∂26-Nov-85  0745	@SU-SUSHI.ARPA:karp@ernie.berkeley.edu  
C02888 00576	∂26-Nov-85  0953	YAMARONE@SU-CSLI.ARPA 	NO LUNCH WEDNESDAY    
C02889 00577	∂26-Nov-85  1012	RICHARDSON@SU-SCORE.ARPA 	CSD Lunch
C02890 00578	∂26-Nov-85  1050	ullman@su-aimvax.arpa 	meeting tomorrow 
C02891 00579	∂26-Nov-85  1435	MODICA@SU-SCORE.ARPA 	Grades  
C02894 00580	∂26-Nov-85  2228	BRESNAN@SU-CSLI.ARPA 	interactions of morphology, syntax, and discourse    
C02899 00581	∂27-Nov-85  0015	EMMA@SU-CSLI.ARPA 	Ventura Security
C02900 00582	∂27-Nov-85  0842	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #39  
C02911 00583	∂27-Nov-85  1056	ACUFF@SUMEX-AIM.ARPA 	Patches for Some Explorer Bugs   
C02913 00584	∂27-Nov-85  1411	ullman@su-aimvax.arpa 	algebra of derivations
C02915 00585	∂27-Nov-85  1446	SCHOLZ@SU-SUSHI.ARPA 	meeting of the minds   
C02917 00586	∂27-Nov-85  1648	LANSKY@SRI-AI.ARPA 	future plans for PLANLUNCH    
C02919 00587	∂27-Nov-85  1708	BSCOTT@SU-SCORE.ARPA 	Tuesday Faculty Lunch  
C02921 00588	∂28-Nov-85  1917	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #40  
C02952 00589	∂29-Nov-85  0848	NILSSON@SU-SCORE.ARPA 	Computing Needs  
C02955 00590	∂29-Nov-85  1028	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books--CS   
C02957 00591	∂29-Nov-85  1527	NILSSON@SU-SCORE.ARPA 	Patent agreement 
C02958 00592	∂29-Nov-85  1537	LIBRARY@SU-SCORE.ARPA 	Paper Index To Technical Reports No Longer Available
C02960 00593	∂01-Dec-85  1235	JOHNSON@SU-CSLI.ARPA 	My books..   
C02961 00594	∂01-Dec-85  1639	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #41  
C02988 00595	∂01-Dec-85  2336	ACUFF@SUMEX-AIM.ARPA 	New Explorer Software  
C02991 00596	∂01-Dec-85  2347	ACUFF@SUMEX-AIM.ARPA 	Compatibility between Symbolics 36xx and TI Explorer 
C02993 00597	∂02-Dec-85  0859	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C02994 00598	∂02-Dec-85  0947	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Dec. 3, 1985 
C03001 00599	∂02-Dec-85  0947	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Dec. 3, 1985
C03008 00600	∂02-Dec-85  1002	BSCOTT@SU-SCORE.ARPA 	Tuesday Lunch
C03009 00601	∂02-Dec-85  1015	ZEC@SU-CSLI.ARPA 	interactions of morphology, syntax, and discurse    
C03010 00602	∂02-Dec-85  1041	BSCOTT@SU-SCORE.ARPA 	RFP from IBM 
C03013 00603	∂02-Dec-85  1106	RICHARDSON@SU-SCORE.ARPA 	Faculty Accomplishments 
C03014 00604	∂02-Dec-85  1121	EMMA@SU-CSLI.ARPA 	Bibtex emacs library 
C03016 00605	∂02-Dec-85  1121	REGES@SU-SCORE.ARPA 	anyone free on 11/5?    
C03018 00606	∂02-Dec-85  1123	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Announcing a Logic Meeting at UCLA 
C03023 00607	∂02-Dec-85  1205	EMMA@SU-CSLI.ARPA 	bibtex emacs library 
C03024 00608	∂02-Dec-85  1620	RICHARDSON@SU-SCORE.ARPA 	Near West Campus Invitation  
C03026 00609	∂02-Dec-85  1716	DAVIES@SUMEX-AIM.ARPA 	No meeting this Wednesday  
C03027 00610	∂02-Dec-85  1727	MODICA@SU-SCORE.ARPA 	survey forms 
C03028 00611	∂02-Dec-85  1732	MODICA@SU-SCORE.ARPA 	grades  
C03029 00612	∂02-Dec-85  1802	TAJNAI@SU-SCORE.ARPA 	Bell Fellowship   
C03031 00613	∂03-Dec-85  0923	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar in Logic and Foundations   
C03033 00614	∂03-Dec-85  0923	ullman@su-aimvax.arpa 	CS300 talk  
C03034 00615	∂03-Dec-85  1101	TAJNAI@SU-SCORE.ARPA 	Bell Fellowship   
C03035 00616	∂03-Dec-85  1148	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH  
C03039 00617	∂03-Dec-85  1203	NILSSON@SU-SCORE.ARPA 	Faculty Meeting  
C03043 00618	∂03-Dec-85  1315	MODICA@SU-SCORE.ARPA 	End Quarter Reports    
C03044 00619	∂03-Dec-85  1915	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #42  
C03072 00620	∂03-Dec-85  1934	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #42  
C03100 00621	∂03-Dec-85  2038	BRESNAN@SU-CSLI.ARPA 	interactions of morphology, syntax, and discourse    
C03105 00622	∂04-Dec-85  0735	PATASHNIK@SU-SUSHI.ARPA 	next AFLBs
C03110 00623	∂04-Dec-85  0919	ullman@su-aimvax.arpa 	meeting
C03111 00624	∂04-Dec-85  0949	@SU-SCORE.ARPA:MCCALL@SU-SUSHI.ARPA 	Announcing: CS523 - Survey of Artificial Intelligence
C03117 00625	∂04-Dec-85  1037	DALRYMPLE@SU-CSLI.ARPA 	fun on Friday   
C03118 00626	∂04-Dec-85  1320	MODICA@SU-SCORE.ARPA 	Grade Sheets 
C03120 00627	∂04-Dec-85  1341	CAROL@SU-CSLI.ARPA 	Parking man    
C03121 00628	∂04-Dec-85  1702	EMMA@SU-CSLI.ARPA 	Newsletter December 5, No. 4   
C03137 00629	∂04-Dec-85  2141	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Call for papers - Logic and Applications 
C03142 00630	∂05-Dec-85  0733	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #43  
C03146 00631	∂05-Dec-85  0958	REULING@SU-SCORE.ARPA 	CSD Mailing Lists
C03148 00632	∂05-Dec-85  1223	EMMA@SU-CSLI.ARPA 	Newsletter correction
C03155 00633	∂05-Dec-85  1339	@SU-SCORE.ARPA:LES@SU-AI.ARPA 	Facilities action items 
C03171 00634	∂05-Dec-85  1402	TAJNAI@SU-SCORE.ARPA 	nominations for IBM fellowship   
C03173 00635	∂05-Dec-85  1514	NILSSON@SU-SCORE.ARPA 	Awards 
C03176 00636	∂05-Dec-85  1718	ACUFF@SUMEX-AIM.ARPA 	I'll be gone for 6 days
C03178 00637	∂05-Dec-85  2102	BRESNAN@SU-CSLI.ARPA 	morphology/syntax/discourse interactions   
C03181 00638	∂05-Dec-85  2158	JF@SU-SUSHI.ARPA 	help   
C03182 00639	∂05-Dec-85  2247	CAROL@SU-CSLI.ARPA 	New office
C03184 00640	∂06-Dec-85  1003	CONTRERAS@SU-SCORE.ARPA 	Package   
C03185 00641	∂06-Dec-85  1300	YAMARONE@SU-CSLI.ARPA 	HOLIDAY PARTY NEXT FRIDAY(THE 13TH!) 
C03192 00642	∂06-Dec-85  1309	YAMARONE@SU-CSLI.ARPA 	WHAT TIME'S THE PARTY BEGIN?    
C03194 00643	∂06-Dec-85  1408	RICE@SUMEX-AIM.ARPA 	Explorers have new MIBs 
C03195 00644	∂06-Dec-85  1423	@SU-CSLI.ARPA:PCOHEN@SRI-AI.ARPA 	Seminar by N. S. Sridharan, BBN Labs
C03199 00645	∂06-Dec-85  1629	@SU-CSLI.ARPA:PCOHEN@SRI-AI.ARPA 	time for seminar by N. S. Sridharan 
C03200 00646	∂06-Dec-85  1634	PHILOSOPHY@SU-CSLI.ARPA 	colloquium
C03202 00647	∂06-Dec-85  2050	@SU-SUSHI.ARPA:karp@ernie.berkeley.edu  
C03204 00648	∂07-Dec-85  0846	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #44  
C03215 00649	∂08-Dec-85  2355	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #45  
C03219 00650	∂09-Dec-85  1032	RICHARDSON@SU-SCORE.ARPA 	General Faculty Meeting 
C03222 00651	∂09-Dec-85  1045	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C03223 00652	∂09-Dec-85  1333	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books--CS   
C03225 00653	∂09-Dec-85  1720	@SU-CSLI.ARPA:WALDINGER@SRI-AI.ARPA 	seminar on program transformation wednes, 3:45  
C03227 00654	∂09-Dec-85  2250	DAVIES@SUMEX-AIM.ARPA 	AAP meeting Wednesday 
C03228 00655	∂10-Dec-85  1029	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
C03229 00656	∂10-Dec-85  1214	RICHARDSON@SU-SCORE.ARPA 	CSD 500 Colloquium Series    
C03232 00657	∂10-Dec-85  1235	DAVIES@SUMEX-AIM.ARPA 	Wednesday AAP meeting 9:30 
C03233 00658	∂10-Dec-85  1237	BROWN@SU-CSLI.ARPA 	Message for Sandwich
C03234 00659	∂10-Dec-85  1319	LIBRARY@SU-SCORE.ARPA 	The Connection Machine by Hillis
C03236 00660	∂10-Dec-85  1834	ullman@su-aimvax.arpa 	tomorrow's meeting    
C03238 00661	∂10-Dec-85  2226	vardi@su-aimvax.arpa 	Database Seminar  
C03241 00662	∂11-Dec-85  0617	PATASHNIK@SU-SUSHI.ARPA 	Last AFLB of the quarter 
C03246 00663	∂11-Dec-85  1424	RICE@SUMEX-AIM.ARPA 	Today's talk. 
C03247 00664	∂11-Dec-85  1436	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Technical Debate at Stanford on SDI, December 19   
C03251 00665	∂11-Dec-85  1752	EMMA@SU-CSLI.ARPA 	Newsletter December 12, No. 5  
C03268 00666	∂12-Dec-85  0902	LIBRARY@SU-SCORE.ARPA 	USENIX and UNIFORUM Conference Proceedings
C03270 00667	∂12-Dec-85  0943	YAMARONE@SU-CSLI.ARPA 	HOLIDAY CELEBRATION REMINDER    
C03277 00668	∂12-Dec-85  1116	GOTELLI@SU-SCORE.ARPA 	Correction  
C03278 00669	∂12-Dec-85  1136	GOTELLI@SU-SCORE.ARPA 	New Staff Member 
C03280 00670	∂12-Dec-85  1238	ullman@su-aimvax.arpa 	Recent developments   
C03290 00671	∂12-Dec-85  1537	ullman@su-aimvax.arpa 	oops   
C03294 00672	∂13-Dec-85  0843	TAJNAI@SU-SCORE.ARPA 	Contacts at Apple 
C03295 00673	∂13-Dec-85  0936	MODICA@SU-SCORE.ARPA 	End Quarter Reports    
C03297 00674	∂13-Dec-85  1019	MODICA@SU-SCORE.ARPA 	important correctio    
C03298 00675	∂13-Dec-85  1141	ullman@su-aimvax.arpa 	oops again  
C03300 00676	∂13-Dec-85  1422	ullman@su-aimvax.arpa 	papers received  
C03301 00677	∂13-Dec-85  1441	LANSKY@SRI-AI.ARPA 	revised PLANLUNCH schedule    
C03303 00678	∂13-Dec-85  1609	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.berkeley.edu  
C03305 00679	∂13-Dec-85  1804	@MIT-MC.ARPA:JF@SU-SUSHI.ARPA 	SDI Debate at Stanford, 12/19/85  
C03309 00680	∂14-Dec-85 JF@SU-SUSHI.ARPA	14-Dec-85 JMC	Sleator talk at SRC, Monday, 12/16/85, 2 p.m. 
C03312 ENDMK
C⊗;
∂06-Aug-85  0931	NISSENBAUM@SU-CSLI.ARPA 	CSLI Talk 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  09:31:27 PDT
Date: Tue 6 Aug 85 09:29:27-PDT
From: Helen Nissenbaum <NISSENBAUM@SU-CSLI.ARPA>
Subject: CSLI Talk
To: folks@SU-CSLI.ARPA


Reminder about today's CSLI talk:

time:  4 p.m.
place:  Ventura Seminar Room


			Aaron Sloman
       "The Processing of Motives in Artificial Systems"

-------

∂06-Aug-85  1056	TRUDY@SU-CSLI.ARPA 	Julia Robinson 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  10:56:06 PDT
Date: Tue 6 Aug 85 10:57:35-PDT
From: Trudy Vizmanos <TRUDY@SU-CSLI.ARPA>
Subject: Julia Robinson
To: Researchers@SU-CSLI.ARPA
TEL:  497-1839

     Julia Robinson died this morning, at the age of 66, after a protracted and courageous struggle with leukemia.  On Friday she went into a coma, from which she never recovered.
  
     Professor Robinson is widely known for her outstanding work i
mathematical logic, especially the theory of recursive functions.  She
contributed in essential ways to the attack on Hilbert's 10th problem
regarding the question of effective decidability of Diophantine equations;
this led eventually to the negative solution by Yuri Mathijasevitch.

     Professor of Mathematics at the University of California, Berkeley,
Julia Robinson was also President of the American Mathematical Society
for 1983 and 1984, a member of the National Academy of Sciences since
1976, and a MacArthur Foundation Fellow since 1983.  Through her professional
positions and by her personal example and encouragement, she fostered work
of the highest quality in mathematics.  Liked and admired by all who knew
her, she will be greatly missed.  She is survived by her husband, 
Raphael M. Robinson, and her sister, Constance Reid.

     I am informed that Julia Robinson asked that there be no funeral
services and that those wishing to do something in her memory, should
send a contribution to the Alfred Tarski Fund at the Department of
Mathematics, U.C. Berkeley.  Professor Robinson was instrumental
in helping set us this fund, which will be used to sponsor visitors and
lectures on logic and the foundations of mathematics in Berkeley.  May
I suggest that her name be mentioned when making a contribution.


					Solomon Feferman

-------

∂06-Aug-85  1102	TRUDY@SU-CSLI.ARPA 	Julia Robinson 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  11:01:46 PDT
Date: Tue 6 Aug 85 11:03:21-PDT
From: Trudy Vizmanos <TRUDY@SU-CSLI.ARPA>
Subject: Julia Robinson
To: Researchers@SU-CSLI.ARPA
TEL:  497-1839


Sorry, the actual death of Julia Robinson was July 30, 1985.

*trudy
-------

∂06-Aug-85  1316	EMMA@SU-CSLI.ARPA 	Newsletter 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  13:16:02 PDT
Date: Tue 6 Aug 85 13:14:29-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter
To: folks@SU-CSLI.ARPA
Tel:  497-3479

  I am likely to be quite busy tomorrow, so please mail all items
for the newsletter as early as possible.

  Remember that people draw quite extensively from the newsletter when
writing the year end report.

Emma

Ex nihilo nihil fit

-------

∂06-Aug-85  1702	SBARNES@SUMEX-AIM.ARPA 	SIGLUNCH  Friday  August 9, 1985    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  17:02:46 PDT
Date: Tue 6 Aug 85 16:07:30-PDT
From: Carol Wright/Susie Barnes <SBARNES@SUMEX-AIM.ARPA>
Subject: SIGLUNCH  Friday  August 9, 1985
To: siglunch@SUMEX-AIM.ARPA
Message-ID: <12133063338.65.SBARNES@SUMEX-AIM.ARPA>


                            SIGLUNCH                         



DATE:              Friday,  August 9, 1985

LOCATION:          Chemistry Gazebo,  betweeen Physical and
                   Organic Chemistry

TIME:              12:05

SPEAKER:           Peter Jackson,
                   Department of Artificial Intelligence,
                   University of Edinburgh

TITLE:             A Flexible Toolkit for Building Expert Systems.



This presentation describes the progress to date of a three-year
Alvey-funded project to study, design and implement tools for building
knowledge-based systems.  The parties involved are Edinburgh University's
Department of Artificial Intelligence and GEC's Artificial Intelligence
Group at Great Baddow.  The aim of the seminar is not to present finished
work (the project is only six months old!), but rather to air our ideas and
prejudices in the hope of attracting criticism and other kinds of feedback
from the expert systems community.

Our survey of current expert systems technology has led us to believe that
neither shells such as EMYCIN (and derivatives) nor high-level programming
languages (such as LOOPS) represent the last word in expert system building
tools.  The former are generally restrictive with respect to both the
representation of knowledge and the specification of control, while the
latter present the average programmer with a bewildering array of
possibilities with little indication of how one combines different
programming styles in the construction of an expert systems architecture.
Thus, although there are groups of users for whom shells and AI programming
languages are well-suited, we feel that there is a substantial gap in the
market between the relative beginner or non-programmer, for whom the
majority of commercial shells are intended, and the veteran hacker, for and
by whom systems like LOOPS were developed.

The alternative approach that we are currently exploring can be summed up by
a number of slogans:

(1) The process of choosing a logically adequate representation language and
a heuristically adequate control regime and embedding these into a suitable
architecture should be guided by some analysis of the task one wants one's
system to perform.

(2) It is worth attempting to establish a correlation between a taxonomy of
expert systems tasks and representational schemes based on logical
languages, with respect to both the expressiveness required by the task
(e.g. modal and temporal notions, fuzziness, etc) and the control of
inference (e.g. different problem solving strategies).

(3) It is worth attempting to establish a similar correlation between tasks
and problem solving paradigms (such as ends-means analysis, hypothesize and
test, etc), with a view to helping the user decide on an architecture within
which he can embed the interpretation of this chosen logical language.

The problems we are currently considering include the following:

(1) Is it possible to provide, along with the toolkit, an abstract
architecture that can be instantiated in different ways to implement
different problem solving paradigms?

(2) Could one then embed different interpretations of different logics in
this architecture as part of the instantiation process?

(3) Could one get the behaviour associated with different problem solving
paradigms from this instantiated architecture by running the logical
language under different meta-level regimes?

(4) Will knowledge bases created for use with one instantiation of the
architecture have to be 'recompiled' for use with another instantiation?

(5) How can we help a user to make the 'right' design decisions (assuming we
know what these are)?

We feel that this research program raises a number of very difficult issues,
many of which will not be solved within the scope of the present project.
Nevertheless, we also feel that practical advances in expert systems
development ultimately depend upon theoretical issues of this kind being
addressed, however inadequately.  We still have open minds with regard to
the kinds of utility that a toolkit should provide, and are always ready to
talk to both the builders and users of tools in order to try and gain new
insights into the problem.




-------

∂06-Aug-85  1831	DAVIES@SUMEX-AIM.ARPA 	No meeting this week  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  18:31:07 PDT
Date: Tue, 6 Aug 1985  16:44 PDT
Message-ID: <DAVIES.12133070105.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No meeting this week
cc:   Davies@SUMEX-AIM.ARPA

There will be no Advanced Architectures Project meeting this week.

	-- Byron

∂06-Aug-85  1900	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #9   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 6 Aug 85  18:59:52 PDT
Date:  6 Aug 85 1912-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #9
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Tuesday, 6 Aug 1985        Volume 1 : Issue 9

Today's Topics:

                    Symbolic Computing = Computing
                         First PARSYM Survey

----------------------------------------------------------------------

Date: Fri, 2 Aug 85 22:18:42 pdt
From: coraki!pratt@Navajo (Vaughan Pratt)
Subject: Re: PARSYM Digest   V1 #8

	From: Guy Steele <gls@THINK-AQUINAS.ARPA>
	Subject: Symbolic and numerical computing

	ALL computing applications are symbolic.

My position exactly.  I particularly appreciated GLS's algebraic
examples, which I thought were right on.  Subject closed (as far
as I'm concerned).
-v

------------------------------

Date: Tuesday, 6 August 1985, 5:15 pm PDT
From: Davies@SUMEX
Subject: Symbolic Computing

The cat is out of the bag.  Symbolic computing is indeed all
computing.  PARSYM was designated the netwide mailing list for
*symbolic* computing in order to broaden the domain of discourse
rather than to restrict it to any particular branch of computing.  I
hope it won't be long before someone will ask the analogous question
about parallel computing -- and get the *same* answer.

				-- Your moderator

------------------------------

Date: Tuesday, 6 August 1985, 5:30 pm PDT
From: Davies@SUMEX
Subject: The First PARSYM Survey

As a field of study, parallel computing is nearly as old as computing
itself.  Nonetheless, it is clear that a major shift has recently
occurred in both the academic and industrial communities: parallel
computing is now hot stuff.

In this, the first PARSYM survey, I would like to uncover some of the
motivations for this surge towards parallelism, by asking the
readership the following few questions.  I'll compile the answers and
report back to the mailing list.

1.  What, if any, is your target application for parallel computing?
    (Theoreticians aren't required to reply to this question.)

2.  Is performance improvement your main goal in exploring parallel
    computing? 

3.  (a) If so, what degree of performance improvement does your target
       	application require (in, say, orders of magnitude)?  What
        degree of performance improvement does your current approach
        promise?
    (b) If not, what else do you expect to achieve through the use of
        parallelism? 

To respond to the survey, fill in the following form and return it to:

                     PARSYM-Survey@SUMEX-AIM.ARPA

(not to PARSYM or PARSYM-Request, please).


                      -- First PARSYM Survey --

1. Target application:

2. Is performance your main goal?  (Yes/No)

3. (a)  If so, (i)  required performance improvement:
               (ii) expected performance improvement:

   (b)  Other goals for parallelism:



End of PARSYM Digest
********************

∂07-Aug-85  1133	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	NEXT MONDAY'S PLANLUNCH   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 7 Aug 85  11:33:05 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Wed 7 Aug 85 11:30:25-PDT
Date: Wed 7 Aug 85 11:26:38-PDT
From: LANSKY@SRI-AI.ARPA
Subject: NEXT MONDAY'S PLANLUNCH
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanLehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA

		   Experiments with Belief Resolution

                  Kurt Konolige and Christophe Geissler
                              SRI AI Center

		        11:00 AM, Monday, August 12
	         SRI International, Building E, Room EJ232


In recent work, Konolige developed a resolution rule for a quantified
modal logic of belief.  However, the rule is difficult to apply in
practice, because it takes an arbitrary number of input clauses, and
some instances of the rule may subsume others.  In this talk we
describe a solution to this problem based on a generalization of
semantic attachment, that controls the growth of the search space.  We
have implemented the resulting version of belief resolution with
Stickel's first-order connection-graph theorem prover.  We present
several examples of automatic reasoning about belief using this
system, including a solution to the wise man problem.  




-------

∂07-Aug-85  1532	ullman@diablo 	paper received 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 7 Aug 85  15:32:01 PDT
Date: Wed, 7 Aug 85 15:24:00 pdt
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

Would anybody like a draft of "Earley Deduction," by Harry Porter,
a student of Dave Maier?  It talks about the "LR(k)" style of
deduction that Moshe discussed not so long ago (but the idea
apparently dates back to 1983 at least, maybe "earlier").

I wonder whether one can use this algorithm to lower bound the
number of deductions made by any of the other algorithms
we've been talking about.

∂07-Aug-85  1718	EMMA@SU-CSLI.ARPA 	Newsletter August 8, No. 40    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 7 Aug 85  17:18:14 PDT
Date: Wed 7 Aug 85 16:44:32-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter August 8, No. 40
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
August 8, 1985                  Stanford                       Vol. 2, No. 40
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
           CSLI ACTIVITIES FOR *NEXT* THURSDAY, August 15, 1985

   12 noon		TINLunch
     Ventura Hall       No Lunch this week
     Conference Room    
		
   2:15 p.m.		CSLI Talk
     Ventura Hall	``Relevant Arithmetic and Automatic Theorem Proving''
     Conference Room	Bob Meyer, Australian National University
			(Abstract next week)

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←
                              ANNOUNCEMENT

                     No activities take place today.
                              ←←←←←←←←←←←←
                                CSLI TALK
                         On Situation Semantics
                  Robin Cooper, University of Wisconsin
               Ventura Conference Room, August 14, 2 p.m.
                              ←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
              Summary of the meeting on Thursday, August 1

      Sells, Zaenen, and Zec continued their presentation on a typology
   of reflexive constructions, focussing on the interpretation assigned
   to structures containing reflexive pronouns.  In particular, they
   argued that the traditional conception of the ``bound variable'' vs.
   ``coreferential'' distinction is not fine-grained enough and
   introduced a third interpretation, called ``discourse binding''; this
   builds on the account of anaphora developed in Sells' paper
   ``Restrictive and Non-Restrictive Modification'' (CSLI Report No. 28).
   With regard to the interpretation of reflexive constructions,
   languages fall into two classes:

   --Those that only allow the bound variable interpretation.

   --Those that allow either the bound variable interpretation or the
   discourse binding interpretation.

      A notational system was presented for representing these
   interpretations, using the basics of Kamp's Discourse Representation
   Theory, and rules for constructing such representations from syntactic
   structures were discussed.
      Finally, some speculations were advanced as to the nature of the
   interpretations of other constructions involving reflexives, in
   particular the mediopassive and inchoative use of the `-yi-' reflexive
   in Kungparlang.					--Peter Sells
!
Page 2                     CSLI Newsletter                     August 8, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                     PIXELS AND PREDICATES: AREA P1
                        ``Diagram Understanding:
       The Intersection of Computer Graphics and Computer Vision''
       Fanya S. Montalvo, MIT, Artificial Intelligence Laboratory
                   Summary of the meeting on August 7

      A problem common to computer vision and computer graphics was
   identified.  It deals with the representation, acquisition, and
   validation of symbolic descriptions for visual properties.  The
   utility of treating this area as one was explained in terms of
   providing the facility for diagrammatic conversations with systems.  I
   call this area ``diagram understanding'', which is analogous to
   natural language understanding.  The recognition and generation of
   visual objects are two sides of the same symbolic coin.  A paradigm
   for the discovery of higher-level visual properties was introduced,
   and its application to computer vision and computer graphics
   described.  The notion of denotation was introduced in this context.
   It is the map between linguistic symbols and visual properties.  A
   method was outlined for associating symbolic descriptions with visual
   properties in such a way that human subjects can be brought into the
   loop in order to validate (or specify) the denotation map.  Secondly,
   a way of discovering a natural set of visual primitives was
   introduced.					--Fanya Montalvo
                              ←←←←←←←←←←←←
                           SUMMARY OF NL2 TALK
           ``Some Null Subject Constructions in Modern Irish''
                Jim McCloskey, University College Dublin
                         Tuesday, July 23, 1:00

      The paper discusses a group of related constructions in Modern
   Irish which have two characteristics in common.  They have subjects
   which are phonologically null and which are pleonastic.  The paper is
   particularly concerned with the interaction between unaccusatives and
   a passive construction.
                              ←←←←←←←←←←←←
                             NEW CSLI REPORT

      Report No. CSLI-85-28, ``Restrictive and Non-restrictive
   Modification'' by Peter Sells, has just been published.  This report
   may be obtained by writing to David Brown, CSLI, Ventura Hall,
   Stanford, CA 94305 or Brown@su-csli.
-------

∂08-Aug-85  1244	YAMARONE@SU-CSLI.ARPA 	SANDWICH ORDERS <9:30 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 8 Aug 85  12:44:19 PDT
Date: Thu 8 Aug 85 12:43:20-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: SANDWICH ORDERS <9:30
To: folks@SU-CSLI.ARPA



THE VENTURA SANDWICH CORP. IS GOING TO ATTEMPT A PSYCHOLOGICAL 
MANEUVER TO AVOID LATE SANDWICH ORDERS.

FROM NOW ON , SANDWICH ORDERS MUST BE IN BY 9:30 AM THE DAY 
OF PURCHASE.. ORDERING THE DAY BEFORE IS ENCOURAGED IF YOU'RE
NOT A MORNING TYPE.. AND REMINDERS WILL BE SENT TO THOSE ORDERING
AFTER 9:30.

SO,             ORDER BEFORE 9:30 AND YOU'LL ALL 

                KNOW VICTUAL SATISFACTION!!!!


YOURS IN LUNCH,


THE VENTURA SANDWICH CORP.



-------

∂08-Aug-85  1320	SELLS@SU-CSLI.ARPA 	using Imprint  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 8 Aug 85  13:20:17 PDT
Date: Thu 8 Aug 85 13:17:43-PDT
From: Peter Sells <Sells@SU-CSLI.ARPA>
Subject: using Imprint
To: folks@SU-CSLI.ARPA
cc: bboard@SU-CSLI.ARPA

Several people seem to be experiencing difficulty with the new Canon
software.  The Imprint command no longer works with .dvi files created by
TeX and LaTeX.  To Imprint these, you must run the .dvi files through
Makimp, i.e., you do:

@makimp foo.dvi

This gives you an .imp file which you can Imprint.  However, the pages will
reverse; to turn the reversal off you do:

@imprint foo.imp/REVERSE:NO

Maybe this will alleviate some of the problems.

	Peter
-------

∂08-Aug-85  1527	PIERRE@SU-CSLI.ARPA 	Printing DVI files 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 8 Aug 85  15:26:55 PDT
Date: Thu 8 Aug 85 15:25:17-PDT
From: Peter King <PIERRE@SU-CSLI.ARPA>
Subject: Printing DVI files
To: folks@SU-CSLI.ARPA
cc: bboard@SU-CSLI.ARPA, summer@SU-CSLI.ARPA


We hope to reinstate automagic printing of DVI files soon.  As you
know, we have switched over to a new brand of Canon software and
things are a little rough around the edges.

Peter
-------

∂09-Aug-85  0944	BETSY@SU-CSLI.ARPA 	SDF Visit 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Aug 85  09:44:28 PDT
Date: Fri 9 Aug 85 09:40:46-PDT
From: Betsy Macken <BETSY@SU-CSLI.ARPA>
Subject: SDF Visit
To: folks@SU-CSLI.ARPA



The Board of Trustees of the System Development Foundation is making a
site visit on Monday morning (August 12).  They will be hearing
presentations, and would like a brief tour of the facilities during
their break (probably around 10:15 or 10:30).  They have expressed an
interest in visiting informally with any of you who are around during
the tour.

The members of the Board who will be here are:

		Arnold Beckman, Chairman
		Chairman, Beckman Instruments, Inc.
	
		Ralph Tyler, President
		Director Emeritus, Center for Advanced Study in the
		Behavioral Sciences

		Edwin Huddleson, Assistant Secretary and Financial Officer
	        Partner, Cooley, Godward, Castro, Huddlesson & Tatum
	
		Lloyd Morrisett, Chairman, Investment Committee
	        President, The John and Mary R. Markle Foundation

The following members of the SDF staff will also be here:

		Carl York

		Roberta Ishihara

		John Kerns


-------

∂09-Aug-85  1340	DAVIES@SUMEX-AIM.ARPA 	Advanced Architecture Meeting Wednesday August 14!! 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Aug 85  13:40:33 PDT
Date: Fri, 9 Aug 1985  13:39 PDT
Message-ID: <DAVIES.12133822837.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA, BHayes-Roth@SUMEX-AIM.ARPA
Subject: Advanced Architecture Meeting Wednesday August 14!!
cc:   Davies@SUMEX-AIM.ARPA

Harold will present some initial thoughts on a parallel blackboard
system.  He hopes to get plenty of feedback.

Meeting time is 9:30 to 11 am.

	-- Byron

∂11-Aug-85  1837	ULLMAN@SU-SCORE.ARPA 	["Prof. David Maier" <maier%oregon-grad.csnet@csnet-relay.arpa>: Complexity of LP evaluation]
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 11 Aug 85  18:37:23 PDT
Received: from SU-SCORE.ARPA by diablo with TCP; Sun, 11 Aug 85 18:31:15 pdt
Date: Sun 11 Aug 85 18:29:11-PDT
From: Jeffrey D. Ullman <ULLMAN@SU-SCORE.ARPA>
Subject: ["Prof. David Maier" <maier%oregon-grad.csnet@csnet-relay.arpa>: Complexity of LP evaluation]
To: nail@SU-AIMVAX.ARPA
Message-Id: <12134399851.18.ULLMAN@SU-SCORE.ARPA>

I'm forwarding this from Dave Maier.  Interesting...
                ---------------

Return-Path: <maier%oregon-grad.csnet@csnet-relay.arpa>
Received: from csnet-relay by SU-SCORE.ARPA with TCP; Sun 11 Aug 85 07:03:19-PDT
Received: from oregon-grad by csnet-relay.csnet id aa08598; 11 Aug 85 10:02 EDT
Received: by ogcvax.ogc.uucp (4.12/OGC←1.1)
	id AA02062; Sat, 10 Aug 85 10:01:09 pdt
Date: Sat, 10 Aug 85 10:01:09 pdt
From: "Prof. David Maier" <maier%oregon-grad.csnet@csnet-relay.arpa>
Message-Id: <8508101701.AA02062@ogcvax.ogc.uucp>
To: ullman@su-score.ARPA
Subject: Complexity of LP evaluation


While at MCC, Bancilhon filled me in about his method of measuring complexity
of rule evaluators with `firings'.  I assume that you understand his
notions of failed, duplicate, redundant and useless firings.  He made
an interesting point, that failed and useless firings are necessary to
demonstrate that the answer set is complete, that you haven't missed any
answers.

We kicked around the following idea: For a given rule set and goal,
consider all proof trees for that goal under all possible databases.
For a single database, you must invalidate all those proof trees that
don't work for that database.  You invalidate a tree by finding a useless
or failed firing that covers (subsumes) one of the nodes in the tree.  Thus,
the notion of optimal should be extended so that the set of firings
invalidates all illegal proof trees, but no smaller set invalidates all
trees (and finds all answers).

Francois speculated that all the methods he knew used the same firings
to invalidate the same trees, but I was able to find a counterexample.

Dave

PS. Harry's paper assumes familiarity with the Pereira and Warren paper
he references.
-------

∂11-Aug-85  2059	DAVIES@SUMEX-AIM.ARPA 	New mailing list: CARE-Users    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 11 Aug 85  20:58:51 PDT
Date: Sun, 11 Aug 1985  20:58 PDT
Message-ID: <DAVIES.12134427011.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: New mailing list: CARE-Users
cc:   Davies@SUMEX-AIM.ARPA

I've created a new mailing list for people who use CARE (the
architecture simulator): CARE-Users@SUMEX.

The mailing list is in <architects>care-users.dis
The archive is in <architects>care-users.mail

Please send notifications of enhancements, desired capabilities, bugs,
and fixes to the list.

	-- Byron

∂11-Aug-85  2138	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #10  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 11 Aug 85  21:38:19 PDT
Date: 11 Aug 85 2124-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #10
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 12 Aug 1985       Volume 1 : Issue 10

Today's Topics:

                 First PARSYM Survey: More Responses?
            Flat Concurrent Prolog: Beta-Test Announcment
                            Bench Marking
                       HIGH TECHNOLOGY Article

----------------------------------------------------------------------

Date: Fri, 9 Aug 85 15:16:33 pdt
From: Byron Davies <PARSYM-Request@SUMEX-AIM.ARPA>
Subject: First PARSYM Survey

You still have a chance to respond to the First PARSYM Survey (to
PARSYM-Survey@SUMEX-AIM.ARPA).  See V1 #9 for details.  I will be
putting together a special digest of responses: if it's OK for me to
forward your response to PARSYM, let me know.

	-- Byron

------------------------------

Date: Tue 6 Aug 85 23:26:40-MDT
From: Uday Reddy <U-REDDY@UTAH-20.ARPA>

Date: Fri, 26 Jul 85 15:25:56 -0200
From: Ehud Shapiro  <Udi%wisdom.bitnet@WISCVM>
Subject: Beta test-sites for our Flat Concurrent Prolog

	[Forwarded from Prolog Digest]

We will be ready in a month or so to release our Flat Concurrent
Prolog system to beta test sites (FCP is the And-parallel subset of
Concurrent Prolog).  The system consists of an emulator for a
Warren-style FCP abstract machine, written in C, which also includes
kernels and a garbage collector; an FCP compiler, written entirely in
FCP (including the tokenizer, parser, precompiler, encoder, and
assembler); and a basic interactive programming environmnt that
includes a shell, I/O routines, and a source-level debugger, which are
also written in FCP.

The system has an interesting module system, which implements remote
procedure calls via message-passing between module manager processes.
It supports separate compilation and runtime linking.

The system runs on VAX and on the Sun under Berkeley Unix 4.2.  It
runs at about the third of the speed of Quintus Prolog on the Vax for
similar programs, on a wide range of examples.

Some statistics: compiling the main module of the compiler, the
encoder, which is 8937 bytes and 418 lines of code long, takes 258 CPU
seconds on the VAX/750.  The resulting binary file is 17560 bytes of
code long (at present we use word-encoding, rather then byte encoding
for the abstract machine instructions).  The compilation consumes
about 1.5 Mbytes of heap memory (without garbage collection).

During this computation, 29000 processes are generated (yes, twenty
nine thousand), and altogether they perform 92000 reductions.  During
the computation 13000 process suspensions occur.  This gives an
effective rate of 350 process reductions per second, and an avarage of
3 reductions per process.

The system's C code is currently 5270 lines of code long, 2942 for the
emulator, 1624 for kernels, and 704 for the garbage collector.  Its
FCP code is currently 4529 lines of code long, 2060 for the compiler,
324 for the debugger, 722 for I/O, and 1423 for the rest of the
system.  Of course the interactive shell and the compiler share the
tokenizer and parser.

The main designers and implementors of the system are Avshalom Houri
and myself.  Other contributors include Bill Silverman, Michael
Hirsch, Jim Crammond, Colin Mierowski, Steve Taylor, Muli Safra, Nir
Friedman, and Shimon Cohen.  The development of the system was
supported by IBM Poughkeepsie, Data Systems Division.

The system is still under development. The major avenues of
improvement being investigated at present are extending the module
system to be hierarchical, and to integrate better the debugging and
module concepts; integrating the system with a partial evaluator;
improving the performance by optimizing the emulator and improving the
instruction set (we do not plan at present rewriting the emulator in
assembler or in micro-code); adding a window system; an independent
file-system; and other gadgets.

Longer term research includes full compilation and implementation
on a multiprocessor.

We plan to distribute the system following standard university based
software distribution practices.  Before releasing the system for the
general public, we would like to obtain some feedback, and improve the
system some more.  We would like to deliver it to some selected groups
with strong logic programming or concurrent programming background,
who are interested in one or more of the following:

1. Improve and extend the system in various ways.
2. Use it as a research tool for developing pilot FCP
   applications.
3. Use it to teach a course in concurrent logic programming.
   (we believe it is reliable enough even at present for this.
   It has just begun to be used for doing course projects in
   my Concurrent Prolog Programming course at the Hebrew
   University and at the Weizmann Institute, so we will know
   better in a few weeks).

Interested parties should contact me, with relevant information,
at:
        udi%wisdom.bitnet@WISCVM.arpa
     or
        ...!decvax!humus!wisdom!udi

We hope to sort out the details of the distribution mechanism
by mid-August.

-- Ehud Shapiro

------------------------------

Date: Fri, 9 Aug 85 15:16:33 pdt
From: Russell Brand <brand@lll-crg>
Subject: Bench Marking


What type of problems should be given to bench mark a massively parrallel
Symbolic Computing System?  Clearly something from pattern matching,
tree manipulation, sort, searching, game trees...
But what would be good bench mark programs?
Suggestions for inclusion in such a suite are requested by
Russell Brand [brand@lll-crg.ARPA]

------------------------------

Date: Sat 10 Aug 85 20:10:49-EDT
From: Randy Haskins <rh%MIT-EECS@MIT-MC.ARPA>
Subject: Article about our favorite topic

The July issue of HIGH TECHNOLOGY has a pretty reasonable article on the
current progress of parallel processing technology, including an overview
on a few systems working now and some that will be up soon.

Random

------------------------------

End of PARSYM Digest
********************

∂12-Aug-85  1244	YAMARONE@SU-CSLI.ARPA 	Extra Turkey Sandwich 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Aug 85  12:44:07 PDT
Date: Mon 12 Aug 85 12:41:39-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: Extra Turkey Sandwich
To: folks@SU-CSLI.ARPA


There is a turkey sandwich available.

First come, First full.

2.50 .

The Ventura Sandwich Corp.

-------

∂12-Aug-85  1303	@SU-CSLI.ARPA:GEORGEFF@SRI-AI.ARPA 	Monday Planlunch   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Aug 85  13:03:41 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 12 Aug 85 12:58:44-PDT
Date: Sun 11 Aug 85 17:01:38-PDT
From: Michael Georgeff <georgeff@SRI-AI.ARPA>
Subject: Monday Planlunch
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanLehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA

		   Experiments with Belief Resolution

                  Kurt Konolige and Christophe Geissler
                              SRI AI Center

		        11:00 AM, Monday, August 12
	         SRI International, Building E, Room EJ232


In recent work, Konolige developed a resolution rule for a quantified
modal logic of belief.  However, the rule is difficult to apply in
practice, because it takes an arbitrary number of input clauses, and
some instances of the rule may subsume others.  In this talk we
describe a solution to this problem based on a generalization of
semantic attachment, that controls the growth of the search space.  We
have implemented the resulting version of belief resolution with
Stickel's first-order connection-graph theorem prover.  We present
several examples of automatic reasoning about belief using this
system, including a solution to the wise man problem.  

-------

∂13-Aug-85  1027	ullman@diablo 	tomorrow's meeting  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  10:27:11 PDT
Date: Tue, 13 Aug 85 10:16:31 pdt
From: Jeff Ullman <ullman@diablo>
Subject: tomorrow's meeting
To: nail@diablo

Remember that we have a speaker-- abstract attached.
The meeting is at 1PM in MJH 352.
*****************************************

           AND/OR PARALLELISM  IN LOGIC PROGRAMS
 
                              by
 
                         Simon Kasif
                Department of Computer Science
             University of Maryland, College Park
 
 
 
                         ABSTRACT
 
 
      The separation of logic and control in  logic  programs
 has  been  shown  to  allow the programmer to write declara-
 tively lucid programs whose execution is determined  by  the
 interpreter. This appealing characteristic of logic program-
 ming spurred  research  directed  towards  diversifying  the
 means  for  controlling  the execution of logic programs. In
 particular, parallelism in logic programs may  be  exploited
 even  when it is impossible to state a priori that two goals
 may be executed concurrently, but such an opportunity may be
 detected during the course of the execution.
 
      This talk will address the problem of AND/OR  parallel-
 ism in logic programming. We  describe a computational model
 for AND/OR parallel execution of logic programs.  The  model
 provides  the  primitives  to  describe and analyze parallel
 interpreters, emphasizing the  data-flow  among  conjunctive
 goals. The effectiveness of our computational model is esta-
 blished through its ability to support both old and new com-
 munication  paradigms  for  the  parallel  interpretation of
 logic programs.
 
      Several methods  to  implement  AND/OR  parallelism  in
 logic  programs are investigated based on notation developed
 in the model. The methods are shown to define a spectrum  of
 communication  schemes,  ranging  from  the set intersection
 method  where   communication  is   eliminated   altogether,
 through methods based on producers-consumers, where communi-
 cation is uni-directional and finally ending at  a  flexible
 bi-directional  scheme  introduced  in the paper, called the
 Intelligent Channel.
 
      The primitives that comprise the model are used to syn-
 thesize  two  new  parallel interpreters: Disjunctive System
 (DS) interpreters and Intelligent Channel Interpreters.  The
 Intelligent  Channel is a scheme we propose to constrain the
 combinatorial explosion of active processes,  and  to  elim-
 inate  the  need  to maintain a separate binding environment
 for every active OR-branch.

∂13-Aug-85  1040	ullman@diablo 	another talk   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  10:40:51 PDT
Date: Tue, 13 Aug 85 10:33:21 pdt
From: Jeff Ullman <ullman@diablo>
Subject: another talk
To: nail@diablo

This takes place 9AM on Aug. 26 (a Monday, yet).
Nail folks may be interested.
***********************
         The Magic of Partial Evaluation,
                      or
          Meta Interpreters for Real


                Ehud Shapiro

       The Weizmann Institute of Science



Abstract

Enhanced meta-interpreters can implement sophisticated functions within
complex software system.  Examples are explanation facilities in expert
systems, algorithmic debuggers in programming environments, and layers of
protection and control in operating systems.
However, the execution overhead of the added layer of interpretation
is unacceptable in many applications.

Partial evaluation can eliminate the overhead of meta-interpreters.
A partial-evaluator can specialize an enhanced meta-interpreter
with respect to a given program,
generating a variant of this program which inherits the enhanced
functionality of the meta-interpreter, but not its overhead.

An application of a Concurrent Prolog partial evaluator
to operating system development will be shown.

∂13-Aug-85  1225	YAMARONE@SU-CSLI.ARPA 	Two Turkey Sands.Available 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  12:19:30 PDT
Date: Tue 13 Aug 85 12:16:21-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: Two Turkey Sands.Available
To: folks@SU-CSLI.ARPA


There are Two Turkey sandwiches available .

2.50/each

First come , First served.


The Ventura Sandwich Corp.




-------

∂13-Aug-85  1548	ullman@diablo 	Porter paper   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  15:48:04 PDT
Date: Tue, 13 Aug 85 15:38:40 pdt
From: Jeff Ullman <ullman@diablo>
Subject: Porter paper
To: nail@diablo

Anybody who picked up a copy of Porter's paper on
Earley deduction and would also like the *even* numbered
pages, stop by my office.

∂13-Aug-85  1611	vardi@diablo 	Re:  Porter paper    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  16:11:39 PDT
Date: Tue, 13 Aug 85 15:54:59 pdt
From: Moshe Vardi <vardi@diablo>
Subject: Re:  Porter paper
To: nail@diablo, ullman@diablo

The amazing thing is that I have read the paper and it was completely coherent.

Moshe

∂13-Aug-85  1937	YM@SU-AI.ARPA 	Seminar on Control in Prolog  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  19:36:55 PDT
Received: from SU-AI.ARPA by diablo with TCP; Tue, 13 Aug 85 19:30:39 pdt
Date: 13 Aug 85  1929 PDT
From: Yoni Malachi <YM@SU-AI.ARPA>
Subject: Seminar on Control in Prolog 
To: nail@SU-AIMVAX.ARPA

 ∂13-Aug-85  1546	GEORGEFF@SRI-AI.ARPA 	Seminar on Control in Prolog
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  15:43:45 PDT
Date: Tue 13 Aug 85 15:40:21-PDT
From: Michael Georgeff <georgeff@SRI-AI.ARPA>
Subject: Seminar on Control in Prolog
To: aic-associates@SRI-AI.ARPA


Lee Naish, from The University of Melbourne, will be giving a short
talk (which will also be presented at IJCAI) in EK242, immediately
after Chris Stuart's IJCAI talk at 11am, Wednesday 14th.  Here is the
title and abstract:

			PROLOG CONTROL RULES

We present a broad overview of the many proposals for control in
PROLOG.  Two features of computation rules are used to evaluate them.
The first is the well known idea of detecting failure as soon as
possible (where is is unavoidable).  The second idea is to avoid
failures, where possible.  The extend to which various language
constructs and heuristics embody these two ideas is discussed.  By
examining current systems in this light, several conclusions are
reached.  We propose an idealized system, with a hierarchy of goals
and a fair computation rule.

-------

∂13-Aug-85  2205	@MIT-MC.ARPA:HEWITT@MIT-MC.ARPA 	Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  22:05:14 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 AUG 85  01:04:41 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 14 Aug 85 01:02-EDT
Date: Wed, 14 Aug 85 01:03:48 EDT
From: Carl E. Hewitt <HEWITT@MIT-MC.ARPA>
Subject:  Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language
To: phil-sci@MIT-OZ
Message-ID: <[MIT-MC.ARPA].611660.850814.HEWITT>

Folks,

This list has been rather dormant.  To liven things up, I would like to throw
out the following little ticker:

Prolog (like APL before it) will fail as the foundation for Artificial
Intelligence because of competition with Lisp.  There are commercially
viable Prolog implementations written in Lisp but not conversely.

LOGIC as a PROGRAMMING Language will fail as the foundation for AI because:

  1.  Logical inference cannot be used to infer the decisions that need to be
      taken in open systems because the decisions are not determined by
      system inputs.

  2.  Logic does not cope well with the contradictory knowledge bases inherent
      in open systems.  It leaves out counterarguments and debate.

  3.  Taking action does not fit within the logic paradigm.


∂13-Aug-85  2332	DAVIES@SUMEX-AIM.ARPA 	QLISP on CARE    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 13 Aug 85  23:32:36 PDT
Date: Tue, 13 Aug 1985  23:32 PDT
Message-ID: <DAVIES.12134979312.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: QLISP on CARE
cc:   Davies@SUMEX-AIM.ARPA

During the last two weeks, I brought up a QLISP on CARE.  It's a very
basic Lisp -- the only "primitive" functions supported so far are CAR,
CDR, CONS, some arithmetic functions, a form of DEFUN, and LOAD, to
get function definitions from files.  There are no flavors, no
defstructs, no packages, no keyword arguments, no trigonometric
routines, but there is a Lisp on CARE, and it supports QLET and will
soon support the QLAMBDA form as well.

QLISP on CARE runs with all the CARE instrumentation so you can watch
QLISP processes being spawned as the QLISP computation distributes
itself through the CARE network.  The QLISP-on-CARE top level is
implemented as separate servers at separate sites: there is a READ
server which takes input from the terminal; an EVAL server -- at a
separate CARE site -- that accepts expressions from the READ server
and evaluates them using QLISP; and a PRINT server which accepts
output expressions from the EVAL server and prints them to the screen.

In the first sentence, I said I implemented "a" QLISP because the
QLISP I implemented is actually a dialect of Scheme, hence might be
called QScheme.  For the general advantages of Scheme, see the article
by Hal Abelson and Kent Dybvig appended below.  There are a lot of
excellent technical reasons for using Scheme: it is the most
principled Lisp around and contains some interesting features that
make it better than Zetalisp or Common Lisp, as a language.  But for
me, the main reason for implementing Scheme was its simplicity and
clarity.  A meta-circular interpreter for Scheme (a Scheme interpreter
written in Scheme itself) is only a few pages of code.  My
semicircular interpreter, written in Zetalisp, is also just a few
pages of code.

The implementation of QLET for Scheme-on-CARE took about 40 lines of
code, plus a few changes to CARE itself.  The resulting code spawns
CARE processes to evaluate the QLET subexpressions and gets the
results back to the spawning process by means of CARE streams.

Because of the distributed nature of the CARE architecture, it was
easiest initially to implement a distributed version of QScheme.
There is no central shared memory: the initial top-level process
spawns computations at neighboring sites, transmitting the necessary
data and lexical environment along with the function invocation.

When Penny remarked this morning that I had diverged from my initial
plan to implement a shared-memory QLISP on CARE, I decided to see what
it would take to do that.  It turns out that the elegance and
simplicity of Scheme make it very easy to see the leverage points for
converting QScheme to simulate different architectures within the CARE
scheme.  The only change required to make QScheme-on-CARE a shared
memory Lisp was to have the single function APPLY-PRIMITIVE-PROCEDURE
do all memory allocation at one selected CARE node rather than
distributed through the network.  Similar small changes can transform
QScheme to use a larger subset of CARE nodes as memory servers.

So much for the advantages of Scheme from the viewpoint of
implementation.  Will Scheme be an adequate language for implementing
our application?  This I can't answer by myself.  If QLISP needs to
include Common Lisp as a subset, then we may have to wait for Lucid to
do it: Common Lisp is just too complicated for one person -- or at
least this person -- to implement.  I tend to think that Scheme has a
very good set of abstraction mechanisms to support arbitrarily
complex applications, but there is no denying that it will require
some relearning and some reimplementation.

While I'm glad to have a multiprocessing Lisp running on CARE, I'd be
much happier if I could be sure that it would be useful to the
implementors of applications.  I would appreciate any comments on this
issue.

	-- Byron

                       What's Good About Scheme


The following description  of Scheme  is taken from  the "Chez  Scheme
Manual" by Kent Dybvig, with some editing by Hal Abelson:

Scheme is an applicative-order, lexically-scoped, and block-structured
dialect of  Lisp.   Like  almost  all  Lisp  dialects,  Scheme  is  an
interactive  language  with  automatic  storage  management  for  data
objects, including  lists,  strings, various  numeric  datatypes,  and
symbols.  Because of this,  Scheme is ideally  suited to symbolic  and
dynamic computation.  Because  Scheme is  interactive, it  is easy  to
learn Scheme by experimenting with it.

Scheme also employs lexical scoping and block structure.  In this way,
it  is  similar  to  Algol  60  and  Pascal  and  unlike   traditional
(dynamically-scoped) Lisp  dialects.  But  Scheme goes  beyond  either
traditional Lisp or  Algol-like languages by  providing procedures  as
"first-class" data objects.  A  Scheme procedure may  be passed as  an
argument, returned  as  a  value,  made  part  of  a  compostive  data
structure,  and  stored   indefinitely  while   still  retaining   the
environment of its definition.

Like Scheme, Common Lisp also provides lexical scoping and first-class
procedures.  However, Common Lisp does  not treat procedures and  data
in a  uniform manner:  Procedure identifiers  are separate  from  data
identifiers, evaluation  of the  operator  position of  a  combination
follows different rules than evaluation of the operand positions,  and
procedures must be "quoted" in a special way if they are to be treated
as data.   Scheme really  does treat  procedures and  data  uniformly,
greatly simplifying evaluation rules and cutting down namespaces,  and
gaining expressive power with no loss of efficiency.

Another difference between Scheme  and Common Lisp  is that Scheme  is
specified to be  tail-recursive --  procedure calls  can be  evaluated
without building  up  space for  "control  stack."  This  permits  the
definition of a wide variety of "imperative" constructs (such as loops
and  other  iterators)  purely  in  terms  of  procedure  application.
Additionally,  Scheme  does  not  prescribe  the  order  of   argument
evaluation (as does  Common Lisp)  so there  is more  freedom for  the
Scheme implementor  to  rearrange  programs for  optimization  or  for
parallel evaluation.

As a  consequence  of  lexical  scoping  and  tail  recursion,  Scheme
encourages   functional   (side-effect    free)   programming.     The
higher-order procedures that a Scheme programmer can easily  construct
and use provide alternative and more  elegant ways to perform most  of
the computations that are usually accomplished with side-effects.

Another feature of Scheme  places it beyond  other Lisp dialects,  and
most other programming languages as  well.  This is the provision  for
continuations, a  general control  facility  based on  solid  semantic
principles.   Continuations  allow   the  implementation  of   control
structures such as  coroutines, non-blind  backtracking, and  multiple
tasks.

These semantic  attributes of  Scheme help  the programmer  to  create
clear,  concise,  and  maintainable  programs  and  program   systems.
However, the most important attribute  of Scheme supporting this  goal
is its simplicity.   The Scheme community  has staunchly resisted  the
addition of new language features  that have not proven themselves  to
be general enough to warrant making the language larger.  As a result,
Scheme is  a fairly  small language  that  relies on  a small  set  of
underlying concepts, which once mastered, provide more power than  the
less general mechanisms found in other programming languages.

A complete description of Scheme can be found in the "Revised  Revised
Report on Scheme,"  edited by Will  Clinger.  This is  published as  a
joint technical  report of  the  Indiana University  Computer  Science
Department and the MIT Artificial Intelligence Laboratory (June 1985).

∂14-Aug-85  0115	@MIT-MC.ARPA:Wayne@MIT-OZ 	Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language 
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  01:15:47 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 AUG 85  04:15:31 EDT
Date: 14 Aug 1985  04:12 EDT (Wed)
Message-ID: <MDCG.WAYNE.12134997608.BABYL@MIT-OZ>
Sender: MDCG.WAYNE%MIT-OZ@MIT-MC.ARPA
From: Wayne McGuire <Wayne%MIT-OZ@MIT-MC.ARPA>
To:   "Carl E. Hewitt" <HEWITT@MIT-MC.ARPA>
Cc:   phil-sci%MIT-OZ@MIT-MC.ARPA, wayne%MIT-OZ@MIT-MC.ARPA
Subject: Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language
In-reply-to: Msg of 14 Aug 1985  01:03-EDT from Carl E. Hewitt <HEWITT at MIT-MC.ARPA>

The pro-Prolog/logic programming people have been out in force lately.
Following are a few comments that have struck my eye:

From an interview with Donald Michie, Research Director of the Turing
Institute, in /Expert Systems/, 2(1), Jan. '85, pp. 20-23:

\begin

ES: Are the ideas of induction and logic programming being taken up in
the United States?

DM: There are three points on which the conventional position in
America is tremendously vulnerable, to such a degree that they are
rapidly becoming a backward area in the core topics of this field.  In
the surface topics to do with how to dress the screen and have icons
and so on they are so good and so professional that for a long time
they may hold their own.  But in certain core themes they are missing
out.  First, they haven't fully taken Prolog-style logic programming
on board.  Secondly, they certainly haven't taken inductive
programming on board.  Thirdly they have no conception of the critical
importance of the design of the rule language in which their skill is
expressed.

Each one of these is going to prove to be a fundamental pillar on
which the science of knowledge engineering will be built and it will
be obvious to every schoolboy in the year 2000.

\end

From ''Prolog Goes to Work,'' by Clara Y. Cuadrado and John L.
Cuadrado, /Byte/, August '85, pp. 151-158:

\begin

The verbosity of an imperative language like FORTRAN (the more modern
ones like Pascal and Ada tend to be even worse) comes as a result of
the need to specify each control path through the program in explicit
detail.  Then, after large programs have been written, it is very
difficult to figure out what they do, partly because the
problem-solving logic cannot be made transparent in this manner.

Studies have shown that the amount of code a programmer produces tends
to be essentially constant regardless of the specific programming
language used.  It is obvious that the fewer lines of clearer code we
use to accomplish a task, the better.  And a logic-programming
language seems to be the perfect candidate for getting more work out
of fewer lines of code....

[The Cuadrados then describe, from personal experience, the vast
superiority of Prolog over Ada in writing a simulator of a data-flow
model for a high-performance signal-processing system.]

One problem with Prolog has been the fact that, when developing an
application, we are forced to write the whole thing as one giant, flat
program.  Some of the new implementations are now providing a module
facility that greatly enhances the software-engineering appeal of the
language.

From the point of view of many applications programmers, what is
really needed is a facility like the ''demo'' predicate proposed by
Bowen and Kowalski [Bowen, Ken, and Robert Kowalski.  Amalgamating
language and metalanguage in logic programming.  Logic Programming,
K.L. Clark and S.A. Tarnlund, eds.  London: Academic Press, 1982,
pages 153-172].  Their basic idea is to be able to have any number of
''theories'' active at any one time.  When attempting to prove a goal,
we specify not only the goal to be proved but also the theory under
which the goal is to be proved.  This facility allows us to maintain
incomplete and even inconsistent theories about an object of interest.
It provides us with a very attractive mechanism for the implementation
of nonmonotonic features.

\end

From ''Logic Programming,'' by Robert Kowalski, /Byte/, August '85,
pp. 161-177:

\begin

Many psychologists believe that human beings are not logical.  Certain
schools of artificial intelligence also argue that knowledge is
inadequate for representing knowledge and belief.  They argue that
logic is rigid and inflexible and that it requires human beings to be
consistent and their knowledge to be complete.

In my opinion, the critics are mistaken.  The practical application of
logic requires a framework within which new knowledge can be
assimilated and beliefs can change.  Such a framework for /knowledge
assimilation/ is completely compatible with logic and has many
potential applications both inside and outside computing.

[Kowalski then goes on to describe just how Prolog can handle
contradictory information, knowledge assimilation and belief
revision.]

\end

It seems possible that Prolog, and its extended logic programming
descendants, might well come into vogue and replace Lisp as the AI
language of choice.  The Japanese and Europeans may be more prescient
in this case than we like to admit.  Although Lisp is theoretically
more flexible and powerful than Prolog (Lisp is often described as the
''assembly language'' of AI), if, practically speaking, most of the
sorts of things we want to do in AI (natural language understanding,
expert systems, etc.) can be written more quickly and compactly, and
be debugged, maintained, and revised more easily in Prolog, then
Prolog will triumph.  Perhaps Symbolics should begin to think about
marketing Prolog Machines.

∂14-Aug-85  0243	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #11  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  02:43:40 PDT
Date: 14 Aug 85 0002-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #11
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 14 Aug 1985     Volume 1 : Issue 11

Today's Topic:

   Results of the First PARSYM Survey: Goals for Parallel Computing


----------------------------------------------------------------------
Date: 14 August 1984, 1:37 am
From: Byron Davies <PARSYM-Request@SUMEX-AIM.ARPA>

                    Results of the First PARSYM Survey

First, I thank the respondents to the First PARSYM survey.  At the
very least, they have provided insight into the variety of goals held
by people working on parallel computing.  There are no conclusive
results: one can't say that 63% of researchers in parallel computing
view performance as their main goal or that the average desired
performance is 1.56 orders of magnitude.  What we can learn from the
survey is that some researchers are interested in specific
applications (rather fewer than might have been expected since my
original question was highly biased towards applications) while others
are interested in understanding parallelism as a physical or
metaphysical phenomenon.  Some are interested in developing tools for
parallel programming; others are interested in incorporating
parallelism into existing problem-solving tools.

Here are the original questions:

  1.  What, if any, is your target application for parallel computing?
      (Theoreticians aren't required to reply to this question.)

  2.  Is performance improvement your main goal in exploring parallel
      computing? 

  3.  (a) If so, what degree of performance improvement does your target
          application require (in, say, orders of magnitude)?  What
          degree of performance improvement does your current approach
          promise?
      (b) If not, what else do you expect to achieve through the use of
          parallelism? 

Immediately below are my capsule summaries of the 15 projects
described by the 14 respondents.  Each project is reviewed on two or
three lines.  The first line describes the target application or
motivation for studying parallelism.  The second line describes the
performance requirements/desires of the project.  The third line, if
present, describes other goals and motivations for studying
parallelism.

Following the capsule summaries are the actual responses submitted.
When I initiated the survey I planned to distribute only a summary
rather than the actual responses.  Since the volume of response was
more manageable than I might have hoped, and since all the respondents
gave permission for their responses to be distributed to PARSYM, I
have included the text below.  Additional survey responses are
welcomed: please send them to PARSYM-Survey@SUMEX-AIM.ARPA.  If you
have any comments on the conduct or the results of the survey, send
them on in.

	-- Byron

<Application or motivation>
  <Performance a goal? How much?>
    <Other goals>

Object-oriented simulation
  Performance not main goal
     Selecting parallel architectures for specific simulation problems

AI reasoning
  Performance main goal: desire 10 orders of magnitude (!)
    Understanding parallelism in natural intelligence

Image understanding
  Performance main goal: need 1-3 orders of magnitude
    Implementing mammalian-type low-level vision systems

Mechanisms for describing parallel systems
  Performance not a goal
    Mechanisms for describing systems in general

Diagnostic expert systems
  Performance main goal: expect 5 or more

Modeling spreading activation in associative memory
  Performance not main goal

Ease of parallel programming
  Performance main goal: desired speedup unspecified

Parallel logic programming
  Performance one goal: 1-2 orders of magnitude hoped for
    New models of computation

Multiprocessing Lisp
  Performance main goal: desire improvements in mean AND variance,
                         wish for O(n), hope for O(n↑.667)
    Understand new programming paradigms

Tools for parallel programming
  Performance not main goal
    Parallel programming language design; description of parallel systems

Blackboard-based problem solving
  Performance main goal: 100x needed
    Better handling of asynchronous events

Production systems in neural networks
  Performance not main goal
    Representation of symbolic information in the brain

General-purpose symbolic computing
  Performance main goal: 10x for small applications, ??? for large

Use of large databases
  Performance main goal: O(n) desired; > O(log n) hoped for
    Interaction of asynchronous actors

General purpose AI applications
  Performance main goal: "orders of magnitude" desired

------------------------------


Date: 07 Aug 85 07:50:48 EDT (Wed)
From: sokol@mitre.ARPA

We have one project which is looking at Object-Oriented programs relative to
parallel architectures.  Specifically we have a large military simulation whic
which contain military objects which have real behaviours and procedures and
which communicate by passing messages.  This forthcoming year we plan to
place our simulation on JPL's Hypercube, and possibly on several other
parallel architectures.

No, performance is not our main goal.

Our primary interest in parallelism is that it can speed up large slow AI
simulations; however our research questions are more specifically how best
to place this type of simulation on a parallel architecture, and which types
of parallel architectures are suitable for which type of simulations.

Lisa Sokol (Sokol@MITRE)

------------------------------

Date: Wed, 7 Aug 1985  09:44 PDT
From: ZVONA@SRI-AI

I'm one of the MIT connection machine designers.

    1. Target application:

AI reasoning.  Most immediately, semantic net access, truth
maintenance, congruence closure, semantic modulation (a
theorem-proving technique), production rule interpretation, graph
isomorphism, pattern recognition, statistical learning, ...

    2. Is performance your main goal?  (Yes/No)

Yes.  I don't see how it could be otherwise, given that you can always
simulate parallel with serial.

    3. (a)  If so, (i)  required performance improvement:

Hard to say.  Ten orders of magnitude?

                   (ii) expected performance improvement:

Depends on for which algorithm.  Also depends on how fast the TMC
connection machine runs, which I haven't heard yet.  I suspect that it
won't do much better than break-even (with say a 3600) for most of
these things.  The next generation of machine (eg the MIT/GE
connection machine) will be faster; how much so, I don't know.

       (b)  Other goals for parallelism:

I believe that natural intelligence has the same computational
constraints artificial intelligence does, so understanding how to use
parallelism for AI may lead to useful psychological theories.

------------------------------

Date: Wed, 7 Aug 85 10:24:25 pdt
From: glick@aids-unix (Jay Glicksman)

	
1. Target application:

	Image understanding.  Both low level and high level
(model-based) ends of the spectrum.  Immediate applications are for IU
workstations and on-board real-time systems for an autonomous land
vehicle.
	
2. Is performance your main goal?  (Yes/No)
	
	I think this is a moot point since anything that can be done on
a parallel machine can be done on a serial one (via simulation in the
worst case).  In other words I think everyone should answer yes (even in
the cases where performance improvements will make certain techniques
feasible for the first time).

3. (a)  If so, (i)  required performance improvement:

	For the workstation application about 1 (order of magnitude)
	For the ALV about 2-3 (at a minimum)

               (ii) expected performance improvement:

	As in (i) for the immediate future.  As the machines (and
algorithms) improve add 1 order of magnitude to each.
	
   (b)  Other goals for parallelism:
	
	Implementation of mammalian-type low-level vision algorithms.
Also (although not my particular area) mammalian-type memory and
learning structures.  This will require smaller-grain processor units
(i.e. more like the Connection Machine than a Butterfly).

	Jay Glicksman (glick@aids-unix)

------------------------------

Date: Wed, 7 Aug 85 10:37:53 pdt
From: coraki!pratt@Navajo (Vaughan Pratt)

1. Target application:

	To develop concepts and languages for describing systems, and
	to develop methods for manipulating those descriptions.
	"System" is very broadly construed to cover Fortran programs,
	silicon, PC boards, buses, keyboards, video displays,
	information networks, telephone exchanges, logical inference
	systems, systems of pulleys, molecular systems, planetary
	systems, ecosystems, etc.  Relevant buzzwords are "broad
	applications spectrum," "concurrency," "mixed analog and
	digital," and "real-time."

2. Is performance your main goal?  (Yes/No)

	No.  It is not even a secondary goal.

3. (a)  If so, (i)  required performance improvement:
               (ii) expected performance improvement:

	Not applicable.

   (b)  Other goals for parallelism:

	I will be satisfied if I can give good descriptions of existing
	systems.  Practically the only serial systems in existence are
	those implied by conventional programming languages.  To
	describe almost any other system one must be able to cope with
	parallelism, along with a number of other concepts not normally
	encountered in programming languages.  Even that first cousin
	of the programming language, the logical inference system, is
	not naturally serial - the natural structure of a proof is a
	dag, not a line.  My goal is to refine existing concepts and
	develop new concepts that deal with variety of behavior,
	temporal progression of events, concurrency of events, mixed
	digital and analog levels in both data (binary data vs. analog
	voltages, positions, velocities etc.) and scheduling (the
	distinction between temporal precedence and real time), etc.

	Some idea of the direction of the present efforts of myself and
	my students can be gleaned from

	Pratt, V.R., The Pomset Model of Parallel Processes:  Unifying
	the Temporal and the Spatial, in Proc. CMU/SERC Workshop on
	Analysis of Concurrency, Springer-Verlag Lecture Notes in
	Computer Science, ed. S. Brookes and A. Roscoe, Pittsburgh,
	1984.

	Pratt, V.R., Some Constructions for Order-Theoretic Models of
	Concurrency, in Logics of Programs, Lecture Notes in Computer
	Science LNCS 193, ed. R. Parikh, Springer-Verlag, Brooklyn,
	June 1985.

	Gischer, J., Partial Orders and the Axiomatic Theory of Shuffle,
	Ph.D. thesis, STAN-CS-84-1033, Dept. of CS, Stanford University,
	Dec. 1984.

	Time permitting, I will try to convey some of the ideas from
	these efforts to this list, to the extent that they may be
	of interest to the list.  It seems pretty reasonable to assume
	of interest the question of what it means to be "parallel."

	I am pursuing this work in parallel with research into the
	graphical aspects of font design for document preparation
	systems, in the course of which some unexpected and striking
	parallels between algebras for process modelling and algebras
	for calligraphics have emerged.

	I am just now putting together a proposal for a grant to carry
	out such work at Stanford and would appreciate leads on funding
	agencies who might have a particular interest in this topic.
	For the past decade I have relied mainly on the NSF, but
	academic year support from this source has over the years been
	steadily decreasing, and has now dwindled to the point where I
	must begin looking elsewhere.  In the interests of minimizing
	bureaucracy and maximizing time available for research the work
	would ideally be covered by a single grant.  Under these
	conditions the ideal agency would therefore presumably be one
	with a particular interest in this subject, or at least some
	faith in its importance.

	I am sure there must be others on this mailing list who are in the
	same boat with respect to both topic and funding who would also be
	well served by such leads.

	--Vaughan Pratt

------------------------------

Date: Wed, 7 Aug 85 14:46:52 edt
From: James A. Reggia <reggia@maryland>

Here are responses to the PARSYM survey on two projects at the
University of Maryland.  For further information, contact Jim
Reggia at       reggia@umcp-cs
via arpanet.

----------------------------------------------------------
FIRST PROJECT


1. Target application:
Parallel processing algorithms for diagnostic expert systems.

2. Is performance your main goal?  (Yes/No)
Yes.

3. (a)  If so, (i)  required performance improvement:
                        Whatever one can obtain.
               (ii) expected performance improvement:
                          Speedup of 5 or more.

   (b)  Other goals for parallelism:
Note: This project is just starting this Fall.

---------------------------------------------------------
SECOND PROJECT

1. Target application:
Spreading activation model of associative memory.

2. Is performance your main goal?  (Yes/No)
No.

3. (a)  If so, (i)  required performance improvement:
               (ii) expected performance improvement:

   (b)  Other goals for parallelism:
    The purpose of this project is to study a new method for controlling
spreading activation in parallel activation networks. This is a project
in cognitive science.  (See forthcoming paper in IJCAI Proceedings
late August).

------------------------------

Date: Wednesday, 7 August 1985  12:03-PDT
From: Tony Li <Tli at Usc-Eclb>

    1.  What, if any, is your target application for parallel computing?
        (Theoreticians aren't required to reply to this question.)

Hmmm...  I'm not a theoretician, but I don't think I have a simple
answer.  Basically, I'm interested in developing programming
languages that make it *trivial* to write parallel programs.

    2.  Is performance improvement your main goal in exploring parallel
        computing?

Yes...

    3.  (a) If so, what degree of performance improvement does your target
           	application require (in, say, orders of magnitude)?  What
            degree of performance improvement does your current approach
            promise?

Any significant amount of speedup would be welcome.


;-)

------------------------------

From: "David M. Meyer"
From: <meyer%tekchips%tektronix.csnet@csnet-relay.arpa>
Date: Wednesday, 7 Aug 85 08:03:27 PDT


1. Target application:	

	I am interested in other (parallel) models of computation,
	specifically, parallel interpretation of "logic" programs.
	However, I am in general interested in novel models of computation.

2. Is performance your main goal?
	
	Yes/..and No -- We need to show that we can do things
	(for instance, execute benchmarks) faster, but there
	are other reasons for exploring parallelism
	(see 3 (b) below). Sorry.

3. (a)  If so, (i)  required performance improvement:

	Probably at least an order of magnitude (or 2)...

        (ii) expected performance improvement:

	????	

   (b)  Other goals for parallelism:

	We/I would like to break out of our "von Neumann" mold,
	and develop some truely "non-Von" models of computation.
	This, of course, requires some novel approaches to an entire
	programming system: the language used, and systems architecture.
	Perhaps here a more top-down, or "language first" approach will
	be useful.

	Dave ---

	David Meyer	

	Tektronix, Inc.
	P.O. Box 500
	MS 50-662
	Beaverton, OR	97077
	
	usenet:		{decvax, allegra}!tektronix!tekchips!meyer	
	csnet:		meyer%tekchips@tektronix.csnet
	arpanet:	tekchips!meyer%tektronix.csnet@csnet-relay.arpa

------------------------------

Date: Thu, 8 Aug 85 12:12 EDT
From: Seth Steinberg <sas@BBN-VAX.ARPA>


1. Target application:

A multiprocessor LISP system.  We hope to support up to 256 68000
processors on a Scheme/Common LISP hybrid.

2. Is performance your main goal?  (Yes/No)

Yes, in two senses.  Overall time improvement.  Statistical stability
in solution time.  (Make the solution time less dependent on the
particular answer and more dependent on the question).


3. (a)  If so, (i)  required performance improvement:
			O(number of processors)
               (ii) expected performance improvement:
			O(number of processors ↑ 2/3)
		[How's that for speculation.]

   (b)  Other goals for parallelism:

   Understanding new programming paradigms.
   [Good and vague.]

					Seth Steinberg

(These opinions are my own and while they are consonant with those of
my employer they are not necessarily identical).

------------------------------

Date: Thu,  8 Aug 85 18:31:27 EDT
From: Alan Bawden <ALAN@MIT-MC.ARPA>

    1. Target application:

I'm not sure how to answer this.  I'm on this mailing list because I'm
interested in developing tools for use on parallel architectures.

    2. Is performance your main goal?  (Yes/No)

Not -my- goal.  It is certainly important to the people who might use the
tools I'm interested in designing.

    3. (a)  If so, (i)  required performance improvement:
                   (ii) expected performance improvement:

       (b)  Other goals for parallelism:

I'm interested in parallel programming language design.  I'm interested in
how a programing language can help people explain how to do a zillion
things at once without becoming confused.

------------------------------

Date: Thu 8 Aug 85 19:32:27-PDT
From: Mike Hewett <HEWETT@SUMEX-AIM.ARPA>


1. Target application:           BB1 (blackboard) Problem-Solving environment

2. Is performance your main goal?  (Yes/No)    Yes.  Both Speed and Versatility

3. (a)  If so, (i)  required performance improvement:   100x
               (ii) expected performance improvement:  unknown

   (b)  Other goals for parallelism:

            Provide versatility for handling blackboard events which
            are caused by "outside" sources.

------------------------------

Date: 10 Aug 85 11:49 EDT
From: Dave.Touretzky@CMU-CS-A.ARPA

1.  My target application is to implement a production system interpreter
in a neural network architecture.  This is work I am doing with Geoffrey
Hinton, also at CMU, as part of the Boltzmann Machine research effort.

2.  Performance improvement is not our main goal.

3.  Our objective is to explore ways in which symbolic information can
be represented and used within the brain.  We are pursuing this question
by designing neural network models of increasing sophistication, in order
to elicit techniques for organizing massively parallel symbolic reasoning
that might, eventually, be shown to have some analog in actual living
systems.  We consider this goal to be a long way off, but we have had some
short term successes, most notably in successfully implementing a
production system interpreter using 7-8000 simulated neurons on a Symbolics
3600.  This work is described in our forthcoming IJCAI paper.

-- Dave Touretzky

------------------------------

Date: Mon, 12 Aug 85 11:00:34 edt
From: Bert Halstead <rhh%mit-vax@mit-mc>

For the Multilisp project at MIT:

1. Target application:  general-purpose symbolic computing (this is kind
of a cop-out, but the intent is to provide tools to facilitate high
performance symbolic computing, rather than to implement any one
application).

2. Is performance your main goal?  Yes

3. (a) (i) required performance improvement:  the more the better
       (ii) expected performance improvement:  highly application
		dependent, but 10x seems easy even for many rather
		small applications.  An interesting question is how
		much we can achieve for larger applications.

------------------------------

From: Gio Wiederhold <WIEDERHOLD@SUMEX-AIM.ARPA>
Date: Wed 7 Aug 85 00:20:58-PDT


1. Target application:  Knowledge-base technology for large databases.

2. Is performance your main goal?  (Yes)

3. (a)  If so, (i)  required performance improvement: O(n), i.e., modest  --
                        (other techniques will be explored prior)
               (ii) expected performance improvement: > O(log n)

   (b)  Other goals for parallelism:  Interaction of asynchronous actors

Gio Wiederhold  KBMS - KSYS project

------------------------------

From: linus!brando@Mitre-Bedford
Date: Tue, 13 Aug 85 11:15:09 edt


1. Target application:  General-purpose AI applications

2. Is performance your main goal?  Yes

3.  (i) Our stated goal is "orders of magnitude improvement in processing
           speed for all classes of AI computations"
   (ii) We cannot predict at this time the actual performance improvement of
	   our current approach

-------------------------------------------------------------------------------

Thom Brando                          linus!brando@MITRE-BEDFORD.ARPA
{allegra,decvax,ihnp4,philabs,utzoo,uw-beaver,...}!linus!brando.uucp

------------------------------

End of PARSYM Digest
********************

∂14-Aug-85  0922	@MIT-MC.ARPA:DAM@MIT-OZ 	Symbolics Prolog    
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  09:22:45 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 AUG 85  12:15:07 EDT
Date: Wed, 14 Aug 1985  11:09 EDT
Message-ID: <DAM.12135073491.BABYL@MIT-OZ>
From: DAM%MIT-OZ@MIT-MC.ARPA
To:   Wayne%MIT-OZ@MIT-MC.ARPA
cc:   phil-sci%MIT-OZ@MIT-MC.ARPA
Subject: Symbolics Prolog


   Date: Wednesday, 14 August 1985  04:12-EDT
   From: Wayne McGuire <Wayne>

   Perhaps Symbolics should begin to think about marketing Prolog
   Machines.

Symbolics recently anounced a PROLOG system.  I think it
has been billed as "the fastest PROLOG in the world".  This
software product has proved to be extremely popular;
Symbolics is selling PROLOG systems like hotcakes.

∂14-Aug-85  0936	@SU-CSLI.ARPA:barwise.pa@Xerox.ARPA 	Robin Cooper's Title   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  09:34:16 PDT
Received: from Xerox.ARPA by SU-CSLI.ARPA with TCP; Wed 14 Aug 85 09:31:47-PDT
Received: from Semillon.ms by ArpaGateway.ms ; 14 AUG 85 09:31:04 PDT
Date: 14 Aug 85 09:30 PDT
From: barwise.pa@Xerox.ARPA
Subject: Robin Cooper's Title
To: Researchers@su-csli.ARPA
cc: barwise.pa@Xerox.ARPA
Message-ID: <850814-093104-5099@Xerox>

Robin Cooper is talking today at 2 PM. His title is "Unification in
Situation Semantics: Linguistic Arguments?"  The talk will  be in the
Ventura Conference Room.

∂14-Aug-85  0958	@MIT-MC.ARPA:Vijay.Saraswat@CMU-CS-C.ARPA 	On logic programming as a foundation for AI.   
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  09:57:45 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 AUG 85  12:48:07 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 14 Aug 85 12:19-EDT
Received: from CMU-CS-C.ARPA by MIT-MC.ARPA 14 Aug 85 12:07:55 EDT
Received: ID <SARASWAT@CMU-CS-C.ARPA>; Wed 14 Aug 85 12:07:08-EDT
Date: Wed 14 Aug 85 12:07:05-EDT
From: vijay <Vijay.Saraswat@CMU-CS-C.ARPA>
Subject: On logic programming as a foundation for AI.
To: phil-sci%MIT-OZ@MIT-MC.ARPA
cc: saraswat@CMU-CS-C.ARPA

This is a response to some of the issues recently raised by Hewitt in
a post to the PHIL-SCI digest. (Phil SCI Digest, Aug 14).

For the sake of argument,  Prolog is probably a `more convenient' language for
writing compilers  than Lisp; on the other hand, the current state of
sophistication of Lisp implementations makes them extremely attractive for
developing rapid prototype implementations.   For the near future I see
the ideal programming environment as being a Lisp Machine with a very fast
Prolog on it (If Symbolics Prolog lives up to rumours of 100K Lips that will
do, thank you.)  Why would anyone EVER want to write a CommonLisp interpreter
in  Prolog?  There are much better things to do ... particularly when you
already have CommonLisp and Prolog!

What does this have to do with "being a foundation for AI"? If he means that
"X is a foundation for AI" if X is used by most people for writing their AI
programs, then that is irrelevant.  I would say that "X is a foundation for AI"
if most work being carried out in AI can fit in the theoretical framework of
X. With such a working definition it is rather doubtful if any one paradigm can
be a foundation for AI.  But I would submit that the concept of logic
programming is a step towards reaching such a foundational understanding. 

"Logic as a programming language will fail as the foundations for AI
because..." 

Let's be careful how we bandy terms around.  Logic is not a programming
language. Logic is logic, a formal system. The first axiom of logic programming
(the LOGIC axiom) is that the definite clause subset of logic has an appealing
interpretation as a programming language, if the process of SLD-refutation
(which is complete for this subset) is taken as the inference mechansim.  In
this programming language, the user has DON'T KNOW CHOICE, i.e. the power of
specifying existential searches.  This power is rather unnerving.  The second
axiom of logic programming (the CONTROL axiom) is that it is possible to
provide general control mechanisms which can be exploited by a programmer for
controlling the search. This led to the rise of logic programming languages,
which allow, in general, the programming of incomplete searches, (hence the
handling of non-montonic inference, defaults, inheritance ...). 

The most ancient of these is PROLOG, which essentially provided the control
structure of SEQUENTIALITY. The language PROLOG has something to do with logic
of course, but, because of its incomplete proof procedure, its semantics is
best given via denotational means, rather than logical means.  This is the
first corollary of logic programming: a formal understanding of logic
programming languages has to appeal to traditional computer science techniques
for giving semantics to programming languages.  The surprising discovery is
that because of the simplicity of the underlying execution mechanism such
semantics are surprisingly simple: far simpler than semantics for ALGOL,
LISP (has anyone ever attempted a semantics for something like COMMONLISP?),
ACTORS...  This has very powerful connotations with respect to semantically
based programming environments, program transformations, and meta-programming. 

There have been more recent languages which have provided other control
features: PARLOG, GHC, Concurrent Prolog and CP[!,|,&], to name just a few in
the main stream of Horn logic programming, have attempted to also provide don't
care choice and parallelism.  To frame it in the current parlance, these
languages essentially provide support for object-oriented programming. 
While a formal denotational semantics for the last is still being developed
(the others do not yet have formal semantics), it has already been demonstrated
that it has a surprisingly simple partial correctness semantics. 

More important, these languages demonstrate that while keeping the LOGIC
component more or less identical, it is possible to achieve a great variety of
operational behaviours by changing the CONTROL component.  Hence logic
programming is not a LANGUAGE, it is a PROGRAMMING PARADIGM encompassing all
these operational behaviours. 

To sum up, logic programming langauges are not just any programming languages,
at some level, most programs written in such languages have a declarative
interpretation compatible with a logic system, typically universally quantified
definite clause logic. Logic programming languages are not just logic systems
because their control component plays an important part.  The art of designing
logic programming languages is concerned with maintaining a delicate balance
between these two divergent themes. Even with their control components, logic
programming languages can generally be given simple semantics, which reflects
their underlying conceptual simplicity. 

As far as supporting most of the current AI paradigms is concerned, it should
be clear that Definite clause logic supports naturally the notions of 
goal-driven backward chaining. 

It is my contention (as yet unsupported) that the basic style of computation
provided by Concurrent Prolog and CP[!,|,&] naturally supports data-driven,
forward-chaining inference and also knowledge-structuring a la semantic
network based languages, again via the primary mechanism of message-passing. 
("reserach in progress", though there is an article by Shapiro andd Takeuchi
on object oriented programming in Concurrent Prolog). 

Hence I would conclude that, to the extent that logic programming languages
support such  programming paradigms, and to the extent that they themselves
have secure theoretical foundations, one can at least claim that logic
programming offers a step towards a unified understanding of the foundations of
(programming paradigms used in) AI.

Note that I am not saying anything at all about the psychological plausibility
of controlled inference as the essential "problem-solving capability" in 
intelligent agents. 


Let me now discuss three specific poinst raised by Hewitt:

  "1.  Logical inference cannot be used to infer the decisions that need to be
       taken in open systems because the decisions are not determined by system
       inputs.

   2. Logic does not cope well withthe contradictory knowledge bases inherent
      in open systems.  It leaves out counterarguments and debate.

   3.  Taking action does not fit within the logic paradigm."

Some of these contentions have been made by Hewitt in other contexts and I
still remain as mystified by them as I was then. 

Let us keep in mind that we are talking of programming langauges, albeit
peculiar programming languages. How does a LISP function 'take action'?
Presumably by doing a (setq foo 'take-action).  How does an OPS-5 program take
action? Presumably by executing a (make task-312 ↑task take-action) on the
right hand side of a production. So also  logic programming languages
take action by instantiating a variable to some value. If that variable was
actually implemented as a particular memory cell which is being monitored by a
hard-wired coke-dispenser, then it would 'take action': deliver a coke bottle. 

Presumably what is confusing Hewitt is that in Prolog bindings can be done on
backtracking.  But all that is in the programmer's control! If the programmer
intended that the action to be taken is irrevocable, he would write his axioms
in such a fashion that the binding would not be backtracked over. That is
a control problem!

About contradictory data bases.  First let me make the obvious statement that a
certain level of understanding, every set of definite clause has a model,
indeed an infinity of models. Hence there is no scope for representing
contradictory knowledge directly via axioms in a Horn logic program. 
So how do logic programming systems do it?  Well you have to PROGRAM IN your
handling of such data.  You have to write a program, just as you would write a
program in LISP or ATOLIA which does to this data what you want to do with it.
The axioms you write (the program you write) will be executed in the fashion
determined by the logic programming language you choose and by the control you
provide: usually this approximates some kind of inferencing. BUt the action
that 
gets taken when you execute your program depends upon your program.  If you had
written your program such that when you encounter an inconsistency it would
print out "The moon is made of green cheese" and stop, then, if an inconsitency
is encountered, believe it or not, it will print out "The moon is made of green
cheese" and stop. If you wanted to program in counterargumentsd and debate then
go ahead and do that.  Logic programming does not provide "counterarguments and
debate" as a primtive concept.   The advantage that you get by writing your
program in a logic programming language is that you can reason about it, you
can (hopefully) prove that when the program encounters an inconsistency it will
print out "The moon is made of greencheese" and terminate.  And because of the
simplicity of the programming paradigm that you are using, your proofs would be
relatively simple.  If you were careful in writing your program and in choosing
your logic programming langauge, the answer that you would get would be a
logical consequence of YOUR AXIOMS (program) which of course COULD HAVE DONE
whatever it wanted to to the data. 

Hence, while on the face of it Hewitt's statement (2) is true, it is totally
irrelevant.  

This also takes care of the argument that "logical inference cannot be used to 
infer decisions that need to be taken in open systems ..." because logical
inference is being used to execute YOUR PROGRAM, which determines how to make
those decisions.  If your program is actually an OPS-5 interpreter which acts
on the data that it gets by using a knowledge base of productions, then, by
heavens, YOUR PROGRAM is NOT doing logical inference.  Hence even if Hewitt's
contention is correct, it is totally irrelevant. 


Vijay A. Saraswat
CMU-CSD


(References for some of the work cited herein are available on request.  Or if
there is sufficient interest, I can send them to the Phil-sci digest.
-------

∂14-Aug-85  1107	@MIT-MC.ARPA:Laws@SRI-AI.ARPA 	Challenge Problem  
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  11:07:17 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 AUG 85  14:07:15 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 14 Aug 85 13:53-EDT
Received: from SRI-AI.ARPA by MIT-MC.ARPA 14 Aug 85 13:51:34 EDT
Date: Wed 14 Aug 85 10:49:57-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Challenge Problem
To: Phil-Sci%MIT-OZ@MIT-MC.ARPA

I have long believed that any algorithm developed in a procedural
language could be reformulated more elegantly in a logic language.
This may be true, but I have come to doubt its utility.  Restating
an algorithm in nonprocedural form can be a useful exercise, but does
not improve on the procedural nature of the algorithm.  Advocates
of logic programming would have a weak case if their languages were
just notations for telling the computer what to do, step by step.

In order to sharpen my understanding of logic programming, I would be
interested in hearing about logic-based solutions to a problem that I
consider particularly "algorithmic": extraction of connected
components in images.  I would like to learn of any logic-based
approach to the problem that would entail neither massive search nor
such extensive procedural embedding as to make the "logic" portion
trivial.


The problem:  Assume that you have an array of integers, and are to
identify all maximal connected components -- i.e., all clusters of
identical integers.  The integers range from 0 to some known maximum,
with 0 being a special code that connects with the space "outside" the
array.  You are free to use any bounded intermediate data structures.
The output is to consist of an array in which elements of each connected
component are marked with the same integer code, and elements of different
components are marked with different codes.  Other outputs (boundary
traces, lists of region adjacencies and inclusion relationships, number
of holes per region, etc.) are desirable but of secondary importance.


I know of several good algorithms for solving this problem.  Most involve
an intermediate map or equivalent data structure that records nonmaximal
connected components found during a row-by-row scan, together with a
list of discovered equivalences that can be used to link these into
maximal components.  Such algorithms are quite efficient (aside from
having to scan the array completly, in the worse case, before starting
to build the output map).  Others involve visiting (and marking) every
point in the array, tracing region boundaries until they close and then
raster scanning to find the next unvisited point.  (This is a reasonable
procedure if the boundary traces are of primary interest.)

Is there any hope that a logic-based language could compile a
[nonalgorithmic] declarative statement of this problem into similar
code, or would the compiler have to be guided by domain knowledge that
essentially supplies it with the right answer?  If the latter is true,
in what sense is logic programming the answer to our current needs?

					-- Ken Laws
-------

∂14-Aug-85  1258	NET-ORIGIN@MIT-MC.ARPA 	Re:  Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language    
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  12:58:29 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 AUG 85  15:57:15 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 14 Aug 85 15:52-EDT
Received: from Uiuc.ARPA by MIT-MC.ARPA 14 Aug 85 11:23:51 EDT
Received: from uiucdcsp.Uiuc.ARPA (uiucdcsp.ARPA) by Uiuc.ARPA (4.12/4.7)
	id AA19858; Wed, 14 Aug 85 10:24:03 cdt
Received: by uiucdcsp.Uiuc.ARPA (4.12/4.7)
	id AA07991; Wed, 14 Aug 85 10:23:10 cdt
Date: Wed, 14 Aug 85 10:23:10 cdt
From: forbus%uiucdcsp@Uiuc.ARPA (Kenneth Forbus)
Message-Id: <8508141523.AA07991@uiucdcsp.Uiuc.ARPA>
To: HEWITT@MIT-MC.ARPA, Wayne%MIT-OZ@MIT-MC.ARPA
Subject: Re:  Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language
Cc: phil-sci%MIT-OZ@MIT-MC.ARPA, wayne%MIT-OZ@MIT-MC.ARPA

(Carl, I wish you had found a fresher topic to re-open the list with.
And doing it right before conference season...)

To equate "prolog" with "logic programing, as Carl & Wayne's messages comes
very close to doing, is somewhat less than accurate.  True, prolog is
currently the logic programming language of choice.  True, prolog provides a
handy notion of variable that takes painful effort to build in Lisp.  But
prolog really is "Micro-Planner done right", as its advocates claim, and
that is enough to allow most AI people to tune out prolog because we found
out a long time ago that Micro-Planner isn't what we want.

Prolog suffers from being both too old and too new.  It is too old, in the
sense that it has chronological backtracking built into its roots and the
standard versions don't take into account new ideas (such as TMS').
Yes, of course one can encode these things in logic and "run them" in
Prolog (If you really believe prolog is "programming in logic" I have
a bridge to sell you).  However, you may not get an answer in the lifetime
of the universe, and doing three or four experiments will have you sitting
fruitlessly at the console long enough to write a fast lisp program to do
the same thing!  But prolog is too new, in that it has also not kept pace
with progress in programming environments and language design.  Talking
to prolog programmers is a bit like talking to lisp programmers 10
years ago -- there are many dialects, often changing rapidly, wild
disagreements on syntax, etc.  The days of CommonProlog are likely to be
pretty far off.

Judging the idea of logic programming by prolog is a bit like judging lisp
by Lisp 1.5.  I believe that only a tiny fraction of the interesting ideas
which should be encapsulated in a real logic programming language have been
developed.  If some venture capitalist were to ask me into whose pocket
should he put $1M to get a signficant advance in the state of logic
programming, I wouldn't suggest spending it on the prolog community.  I'd
probably suggest splitting it between McAllester & de Kleer, whose ideas on
TMS', while rather different, are both far more interesting and potentially
productive ideas than, say, improving the speed of unification or studying
and-parallelism.  We simply don't have enough experience writing and using
reasoning languages yet to either cast our ideas into stone (i.e., adopt
prolog and spend our time optimizing it) OR pass final judgement on the
merits of logic programming.

On logic and action:  Why shouldn't Shakey be considered a counter-example
to Carl's claim about the impossibility of action in a system based on
logic?  Shakey used resolution theorem proving to decide what to do and
other kinds of routines (such as A* search) to flesh out the details.  If
the existence of these other routines is considered sufficient to discount
Shakey as a counterexample then the claim seems rather uninterestingly
obvious.


∂14-Aug-85  1614	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH break -- for next three weeks  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  16:14:21 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Wed 14 Aug 85 16:08:11-PDT
Date: Wed 14 Aug 85 16:05:38-PDT
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH break -- for next three weeks
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanLehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA

Due to IJCAI, summer vacations, and Labor Day, we will be having NO planlunches
for the next three weeks.  The series will resume on September 9, so keep
posted.  Also, if anyone out there would like to give a planlunch talk, please
contact either myself (LANSKY@SRI-AI) or Mike Georgeff (GEORGEFF@SRI-AI).

-Amy Lansky
-------

∂14-Aug-85  1743	EMMA@SU-CSLI.ARPA 	Newsletter August 15, No. 41   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Aug 85  17:43:24 PDT
Date: Wed 14 Aug 85 16:54:40-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter August 15, No. 41
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
August 15, 1985                 Stanford                       Vol. 2, No. 41
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
           CSLI ACTIVITIES FOR *THIS* THURSDAY, August 15, 1985

   12 noon		TINLunch
     Ventura Hall       No Lunch this week
     Conference Room    
		
   2:15 p.m.		CSLI Talk
     Ventura Hall	``Relevant Arithmetic and Automatic Theorem Proving''
     Conference Room	Bob Meyer, Australian National University
			(Abstract on page 1)

   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←
                              ANNOUNCEMENT

   No activities have been scheduled for next Thursday, August 22.
                              ←←←←←←←←←←←←
                    ABSTRACT FOR THIS WEEK'S TALK
          ``Relevant Arithmetic and Automatic Theorem Proving''
                      Thursday, August 15, 2:15 pm

      Relevant logics were first developed in the 1950's, as systems
   satisfying improved versions of the deduction theorem, designed better
   to capture the relation between premises and conclusion in an ordinary
   valid argument.  The time has come to implement the design philosophy
   by exhibiting some valid arguments.  Those familiar with relevant
   deductive technique will appreciate the point that one would not wish
   to do this without a computer.  Relevant technique, whose major root
   has always been deduction-theoretic, does lend itself quite nicely to
   mechanization.  The program KRIPKE developed at LaTrobe, Melbourne,
   and Australian National universities with (and by, mainly)
   Thistlewaite and McRobbie realizes this technique for the system LR,
   applying a sophisticated decision procedure due to Kripke.  KRIPKE is
   Gentzen-based, but invokes semantic constraints to keep proof searches
   within reasonable bounds.  The pay-off is the obvious one; by limiting
   the supply of premises from which a conclusion can reasonably come
   (which was the idea behind relevant logics all along), the path to a
   proof is fast and efficient.  Our aim now is to adapt these methods to
   concrete theories, as part of a 5-year project beginning in early
   1986.  We have chosen arithmetic as our first area of concentration,
   since it has a smooth first-order relevant formalization (in the
   system R#), with many distinctive features.		--Bob Meyer

!
Page 2                     CSLI Newsletter                     August 15, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                CSLI TALK
    ``On the Complementarity of Subject and Subject-verb Agreement''
                            Edit Doron, CSLI
           Wednesday, August 21, 1:30 pm, Ventura Seminar Room

      ``Pro-drop'' languages allow a null subject in conjunction with
   rich inflectional morphology on the verb.  This paper is concerned
   with the other side of the ``pro-drop'' coin: a null subject is
   sometimes REQUIRED under those conditions.  The Celtic languages
   typically impose such complementarity, and Hebrew does so to some
   extent.  I will point out some problems with McCloskey and Hale's
   ``agreement'' analysis for the data, and will propose a variant of the
   ``incorporation'' analysis.
                              ←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
              Summary of the meeting on Thursday, August 8

      A class of five morphemes in Finnish, traditionally called
   possessive suffixes (henceforeward Px), raises interesting questions
   about the relationship of morphological structure to syntactic
   functions.  Px's appear to be pronominal (or anaphoric) elements
   attached to nominal words (nouns and some adjectives, including
   nominalized verbs and participals) following number and case suffixes.
   Recent analyses have treated Px's as clitics, that is, parts of
   phonological words that are not placed within words exclusively by the
   morphology.  I argue in two parts, however, that Finnish possessive
   suffixes are best analyzed as true suffixes.
      The talk, comprising part one of my argument, dealt with
   phonological, morphological, and semantic evidence for the suffixal
   (or morphological word-internal) status of the Px's.  I argued that
   any allomorphy or morphophonological alternation in Finnish that is
   sensitive to word boundaries treats the undisputed suffixes and Px's
   alike as being inside the word and treats a class of clitics as being
   outside the word.  Furthermore, a variety of semantically
   idiosyncratic lexical items containing Px's provide further support
   for a suffixal analysis of Px's, insofar as suffixes are more
   susceptible to idiosyncratic lexicalization than clitics.  I then
   argued against the possibility that Px's are lexical level clitics
   (i.e., clitics that attach to words at the morphological level) by
   showing that it is quite costly to the theory of lexical phonology to
   have a lexical level in Finnish that contains all of the undisputed
   suffixes yet excludes the Px's; hence Px's must occupy the same
   lexical level as other suffixes.  Considering, then, all of the
   evidence favoring a suffixal analysis for the Px's, it is extremely
   weak to set Px's apart from the other suffixes solely on the basis of
   morpheme order.  If the syntactic facts involving Px's can be analyzed
   competently from an entirely lexical basis, then a clitic analysis is
   unmotivated and a suffix analysis is correct.  Part two of my
   argument, to be presented in a later talk, involves such a syntactic
   analysis of the Px's.			--Jonni Kanerva
!
Page 3                     CSLI Newsletter                     August 15, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                             SDF BOARD VISIT

      The Board of Trustees of the System Development Foundation visited
   Ventura Hall last Monday morning (August 12). 
      The members of the Board who were here are:

	Arnold Beckman, Chairman
   	Chairman, Beckman Instruments, Inc.

   	Ralph Tyler, President
   	Director Emeritus, Center for Advanced Study in the Behavioral
          Sciences

   	Edwin Huddleson, Assistant Secretary and Financial Officer
   	Partner, Cooley, Godward, Castro, Huddlesson & Tatum

   	Lloyd Morrisett, Chairman, Investment Committee
   	President, The John and Mary R. Markle Foundation

   Carl York and Roberta Ishihara, two members of the SDF staff, were
   also here.
-------

∂15-Aug-85  0819	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	THE SEVENTH THEORY DAY at Columbia University
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 15 Aug 85  08:18:53 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 15 Aug 85 08:15:20-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Thu 15 Aug 85 08:15:16-PDT
Received: by wisc-rsch.arpa; Thu, 15 Aug 85 09:57:50 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Thu, 15 Aug 85 08:31:54 cdt
Message-Id: <8508151331.AA28042@wisc-crys.arpa>
Received: from COLUMBIA-20.ARPA by wisc-crys.arpa; Thu, 15 Aug 85 08:31:47 cdt
Date: Thu 15 Aug 85 09:30:36-EDT
From: Susan A Maser <MASER@COLUMBIA-20.ARPA>
Subject: THE SEVENTH THEORY DAY at Columbia University
To: theory@WISC-CRYS.ARPA
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

THE
SEVENTH THEORY DAY
at Columbia University

Sponsored by the Department of Computer Science

Friday, September 27, 1985


10:00  PROFESSOR JOHN E. HOPCROFT
       Cornell University
 
       "MATHEMATICAL PROBLEMS IN OBJECT
        REPRESENTATION SYSTEMS"


11:00  PROFESSOR LASZLO LOVASZ
       Eotvos Lorand University, Budapest, Hungary
 
       "MATCHINGS, POLYHEDRA,
        LATTICES, AND ALGORITHMS"


 2:00  DR. FAN K.R. CHUNG
       Bell Communications Research

       "DYNAMIC SEARCH ON GRAPHS"


 3:00  PROFESSOR TOM LEIGHTON
       Massachusetts Institute of Technology

       "WAFER-SCALE INTEGRATION AND
        THE GRID RECONSTRUCTION PROBLEM"


Coffee will be available at 9:30 a.m.
All lectures will be in the Kellogg Conference Center on the
fifteenth floor of the International Affairs Building, 118th
Street and Amsterdam Avenue.
The lectures are free and open to the public.
Call (212) 280-2736 for more information.
-------

∂15-Aug-85  0831	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	my new address 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 15 Aug 85  08:31:34 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 15 Aug 85 08:15:28-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Thu 15 Aug 85 08:15:26-PDT
Received: by wisc-rsch.arpa; Thu, 15 Aug 85 09:55:47 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Wed, 14 Aug 85 22:30:16 cdt
Message-Id: <8508150330.AA25588@wisc-crys.arpa>
Received: from csnet-relay.arpa by wisc-crys.arpa; Wed, 14 Aug 85 22:30:11 cdt
Received: from ibm-sj by csnet-relay.csnet id ct02026; 14 Aug 85 23:13 EDT
Date: Tue, 13 Aug 85 15:25:42 PDT
From: Avi Wigderson <avi%ibm-sj.csnet@csnet-relay.arpa>
To: theory@wisc-crys.ARPA
Subject: my new address
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

Hi
 
To those of you who may send me (regular) mail or call me,
starting Monday, August 19 I will be at Berkeley:
 
Mathematical Sciences Research Institute
1000 Centenial Dr.
Berkeley, CA 94720
Tel: 415-643-6056
 
Avi Wigderson.


∂15-Aug-85  1425	ullman@diablo 	papers received
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 15 Aug 85  14:25:01 PDT
Date: Thu, 15 Aug 85 14:15:26 pdt
From: Jeff Ullman <ullman@diablo>
Subject: papers received
To: nail@diablo

Thanks to our flurry of visitors yesterday, I have received the
following papers:
Lloyd and TOpor: "A basis for deductive database systems II"
	This is related to Lloyd's recent talk--shame on database
	people for not proving "correctness" of relational algebra/calculus
	in a model-theoretic sense (I guess).

Kasif: "Control ad Data Driven Execution of Logic Programs"
	"The Intelligent Channel"
	"Are the logic and control in logic programs always disjoint?"

Naish:	A complete copy of his thesis.
	"All-solutions predicates in Prolog"

I also have many copies of the even numbered pages of Porter's paper.
Drop by if you want to look at any of this.
Perhaps someone might even (mirable dictu) want to present a seminar
on one.

∂16-Aug-85  0125	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #12  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Aug 85  01:22:27 PDT
Date: 16 Aug 85 0111-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #12
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Friday, 16 Aug 1985       Volume 1 : Issue 12

Today's Topics:
                  First PARSYM Survey: Another entry
               Seminar -- Parallelism in Logic Programs
         Parallel processing and economics & theorem proving
               Parallel algorithms for linear programming


----------------------------------------------------------------------

Date: 15 August 85 10:09-EDT
From: BYKAT%UTCVM.BITNET@Berkeley
Subject: First PARSYM Survey

I am a bit late, but hope not too late! Here is one of my projects:

1. Target application:
     Parallelity in parsing NL utterances ( + generation of
     representation of their meaning which supports parallel
     retrieval of information from a knowledge base).

2. Performance a goal?
     Yes, performance is an important goal.

3. Required performance?
     Expect, and hope for, an improvement of at least one order
     of magnitude


Alex Bykat
BitNet address:  BYKAT@UTCVM


------------------------------

Date: Tue, 13 Aug 85 01:11:28 EDT
From: Steven A. Swernofsky <SASW@MIT-MC.ARPA>
Subject: Seminar -- Parallelism in Logic Programs

         [Excerpted from the IBM-SJ Calendar by Laws@SRI-AI.]

                    IBM San Jose Research Lab
                         5600 Cottle Road
                        San Jose, CA 95193

 Fri., Aug. 16 Computer Science Seminar
 2:00 P.M.     PARALLELISM IN LOGIC PROGRAMS
 Aud. A        The separation of logic and control in logic
               programs has been shown to allow the programmer
               to write declaratively lucid programs whose
               execution is determined by the interpreter.
               This appealing characteristic of logic
               programming spurred research directed towards
               diversifying the means for controlling the
               execution of logic programs.  In particular,
               parallelism in logic programs may be exploited
               even when it is impossible to state a priori
               that two goals may be executed concurrently, but
               such an opportunity may be detected during the
               course of the execution.  This talk will address
               the problem of and/or parallelism in logic
               programming.  We describe a computational model
               for and/or parallel execution of logic programs.
               The model provides the primitives to describe
               and analyze parallel interpreters, emphasizing
               the data-flow among conjunctive goals.  The
               effectiveness of our computational model is
               established through its ability to support both
               old and new communication paradigms for the
               parallel interpretation of logic programs.

               Prof. S. Kasif, Department of Computer Science,
                 University of Maryland, College Park
               Host:  P. Lucas

------------------------------

Date: 02 Aug 85 15:31:01 EDT (Fri)
From: Emil J. Volcheck <volcheck at UDel-Dewey.ARPA>
To:   AIList
Re:   Parallel Processing and Economics & Theorem Proving

[Forwarded from AIList by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

        Economists use mathematical models to analyze the behavior of
large systems, in fact some of these models are basically simulations
of an open system in which many individuals are making economic decisions.
If someone can prove theorems pertaining to parallel processing on
distributed systems, these results may carry over to economics, and
vice-versa.  Does anyone know of research pertaining to this idea, or
does anyone have thoughts about this?

        Also -- Does anyone know of a place in the literature about
theorem-proving where proving a theorem is defined as "The unification of
necessary (given) with sufficient (goal) conditions" ?  I'd appreciate
your thoughts on this too.

                                                        --Emil Volcheck

------------------------------

Date: 13 Aug 1985 1710-CDT
From: SATISH <THATTE%CSL60%ti-csl.csnet@csnet-relay.arpa>
Subject: Parallel algorithms for linear programming

Do you know anything about solving linear programming problems (for
example, using the Simplex method) using a parallel algorithm ?

Any pointers to literature, feedback or your opinions will be greatly
appreciated.

-- Regards, Satish Thatte

Texas Instruments, Dallas

------------------------------

End of PARSYM Digest
********************

∂16-Aug-85  0925	ullman@diablo 	a random thought before I forget it
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 16 Aug 85  09:25:25 PDT
Date: Fri, 16 Aug 85 09:18:59 pdt
From: Jeff Ullman <ullman@diablo>
Subject: a random thought before I forget it
To: nail@diablo

Has anyone thought about transformations that convert nonlinear
rules into linear ones, thereby showing that they are in NC?
The canonical example is the transitive closure:
	T -> E.   T -> TT
which can be rewritten T -> E. T -> ET.
There was a paper in the '82 TOulouse conference that talked
about "Greibach normal form" using this as the example,
but it didn't deal with when the result becomes linear---it was
viewed as a way to eliminate left recursion.

∂16-Aug-85  1010	PAPA@SU-SCORE.ARPA 	Re: a random thought before I forget it 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 16 Aug 85  10:10:47 PDT
Received: from SU-SCORE.ARPA by diablo with TCP; Fri, 16 Aug 85 10:04:30 pdt
Date: Fri 16 Aug 85 10:00:32-PDT
From: C. Papadimitriou <PAPA@SU-SCORE.ARPA>
Subject: Re: a random thought before I forget it
To: ullman@SU-AIMVAX.ARPA
Cc: nail@SU-AIMVAX.ARPA
In-Reply-To: Message from "Jeff Ullman <ullman@diablo>" of Fri 16 Aug 85 09:18:59-PDT
Message-Id: <12135617974.33.PAPA@SU-SCORE.ARPA>

Athena Roussou at NTU Athens has thought a little about transforming rules.
She convinced me at some point that, if I remember correctly, the
transitive closure example is essentially the only pattern without
use of auxiliary relations and new rules.

Christos.
-------

∂17-Aug-85  1306	BRAD@SU-CSLI.ARPA 	System clock    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Aug 85  13:06:24 PDT
Date: Sat 17 Aug 85 13:03:19-PDT
From: Brad Horak <BRAD@SU-CSLI.ARPA>
Subject: System clock
To: folks@SU-CSLI.ARPA

Turing's clock was set to July 17th sometime after 8am Saturday, August 17.
The clock was reset at 12:40 Saturday afternoon.  Mail sent during this time
may need to be resent, as the receiving mailer may mark the mail as old.

--Brad
-------

∂18-Aug-85  1216	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	birth announcement  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 18 Aug 85  12:15:54 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Sun 18 Aug 85 12:12:29-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Sun 18 Aug 85 12:12:34-PDT
Received: by wisc-rsch.arpa; Sun, 18 Aug 85 14:00:58 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Sun, 18 Aug 85 10:21:04 cdt
Received: from mitre.ARPA by wisc-crys.arpa; Sun, 18 Aug 85 10:20:58 cdt
Received: by mitre.ARPA (4.12/4.7)
	id AA16009; Sun, 18 Aug 85 11:22:20 edt
Message-Id: <8508181522.AA16009@mitre.ARPA>
To: theory@wisc-crys.ARPA
Subject: birth announcement
Date: 18 Aug 85 11:22:05 EDT (Sun)
From: laskowsk@mitre.ARPA
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;


We would like to announce the birth of our daughter:

        Nicole Marie Jaja

         July 28, 1985

         8 lbs. 7 oz.


     --the proud parents    Joseph Jaja (eneevax!joseph @ maryland)
                            Sharon Laskowski (laskowski @ mitre)


∂18-Aug-85  1505	@SU-SUSHI.ARPA:ANDERSON@SU-SCORE.ARPA 	Thesis Orals, August 22   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 18 Aug 85  15:05:33 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Sun 18 Aug 85 15:01:46-PDT
Date: Sun 18 Aug 85 15:01:52-PDT
From: Richard Anderson <ANDERSON@SU-SCORE.ARPA>
Subject: Thesis Orals, August 22
To: aflb.all@SU-SCORE.ARPA
Message-ID: <12136197117.16.ANDERSON@SU-SCORE.ARPA>

My thesis orals are next Thursday, August 22, 2:15,  
Building 160 (pol sci), room 161K

The title of the talk is:
	The Complexity of Parallel Algorithms

-- Richard Anderson
-------

∂18-Aug-85  1535	ACUFF@SUMEX-AIM.ARPA 	Local BUG-LISPM   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 18 Aug 85  15:35:32 PDT
Date: Sun 18 Aug 85 15:35:11-PDT
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Local BUG-LISPM
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12136203185.27.ACUFF@SUMEX-AIM.ARPA>

   When there are problems using 3600's, please feel free to send reports
or fixes to BUG-LISPM@SUMEX, and read them from <BBOARD>KSL-BUG-LISPM on
SUMEX.  This means you can use the 3600 software to submit bug reports,
as well as conventional mail software.  I will read msgs sent to this
mailbox, and respond as soon as possible, though anyone interested is
encouraged to contribute knowledge.

	-- Rich
-------

∂19-Aug-85  0948	@MIT-MC.ARPA:Hewitt@MIT-MC.ARPA 	Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 19 Aug 85  09:48:01 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 19 AUG 85  12:48:00 EDT
Received: from MIT-XX.ARPA by MIT-OZ via Chaosnet; 19 Aug 85 12:46-EDT
Date: Mon, 19 Aug 1985  12:47 EDT
Message-ID: <HEWITT.12136401984.BABYL@MIT-XX>
Subject:  Prolog will fail as the foundation for AI; so will LOGIC as a PROGRAMMING Language
To:   phil-sci@MIT-OZ
Sender: HEWITT@MIT-XX
From: Hewitt@MIT-MC.ARPA
CC:   Hewitt@MIT-XX, PEREIRA@SRI-AI.ARPA

    Date: Saturday, 17 August 1985  13:36-EDT
    From: PEREIRA at SRI-AI.ARPA
    To:   HEWITT at MIT-MC.ARPA
    Re:   Prolog will fail as the foundation for AI; so will LOGIC as

    Carl Hewitt's message is based on several misconceptions:

    1. (the least interesting one) All the so-called commercially viable
    Prolog systems in Lisp are not really Prolog systems written IN Lisp,
    but rather Prolog systems written FOR Lisp machines. Or is it that a
    microcode sublanguage or Lisp machine pointer-smashing operations are
    part of List as we know it? 

Yes.  They are DEFINITELY part of Common Lisp as we know it being
implementations of reading and writing operations on record
structures.  Such implementation methods are NOT part of Logic as a
Programming language.

    Without those machine-level operations,
    those Prolog systems would run too slow and use too much memory to be
    useful for serious Prolog programming. From the Prolog implementation
    point of view, what is important about the Lisp machines is not that
    they run Lisp, but that they can be microcoded and have good support
    for tagged data types and stack operations.

It is important to many users that they can make use of ALL the software
available to the community and not just be limited to the tiny amount
in Prolog.  Furthermore in the future good software will be ported
from stand alone Prolog systems to Prolog implemented on Lisp.  However
to good Lisp software will not be able to be ported to the stand
alone Prolog systems.

    2. If the decisions (actions) of a system are not determined by its
    inputs, the system is nondeterministic. Nondeterminism in a system can
    be either an artifact of our incomplete knowledge (or lack of
    interest) of the detailed operation of the system; or it can be
    ``real physical'' nondeterminism. It would take us to far to discuss
    whether the second kind of nondeterminism is ``real'' or also an
    artifact. In any case, most uses of nondeterminism, say in models of
    concurrency, are of the first kind, and can be expressed appropriately
    in various temporal/dynamic logics. Admittedly, these are not Prolog,
    but then Common Lisp is not Lisp 1.5! (Prolog is 13 years old, Lisp
    25).

Yes indeed there is a large problem here that poses fundamental problems
for using Logic as a Programming language to make decisions in Open
Systems.

    3. The first logic course dictum ``from a contradiction one can
    conclude anything'' is getting in the way. Notice that the dictum says
    ``can'', not ``must''. There is an enormous difference between things
    that are in principle true and things that an agent knows to be true
    in a way that can affect its action. An agent might know ``p'' and
    ``not p'', but it might well never come to infer the dreaded totally
    unrelated ``q'' which IN PRINCIPLE follows from the contradiction.
    This inference might not happen either because of inference control
    mechanisms of the agent (eg. it uses the set-of-support strategy) or
    because the agent's logic is just TOO WEAK to conclude anything from a
    contradiction (vide Hector Levesque's work in the proceedings of the
    last AAAI). In any case, Horn clauses (the basis of Prolog) are too
    weak to represent contradictions... :-)

I claim that in practice the contradictions lie close to the surface and
occur in any nontrivial application.  Thus the contradictions
pose fundamental problems for using Logic as a Programming Language.

    4. The question of whether ``taking action'' fits in a logic paradigm
    tends to be answered negatively after an hour's worth of
    consideration.  If you persist for several years, though, this
    question becomes a source of insight on the relations between
    knowledge, state and action that is not available to those who started
    by dismissing the question after that initial hour. There is just too
    much work on logics of knowledge and action in AI and computer science
    for me to try to discuss it here. Some of this work has been applied
    to logic programming, either in the form of new logic programming
    languages based on temporal or dynamic logics or in representations of
    temporal reasoning and decision in, say, Prolog. 

I have been thinking about the problem for many years having designed
Micro-Planner, the first "procedural embedding of logic" programming
language in 1967.  I claim that the problem of taking action poses
fundamental problems for using Logic as a Programming language.

    5. It is curious to see someone by implication defend Lisp as a
    language for expressing the taking of action!

I claim that current Lisp systems are better than current Prolog
systems for taking action because the only ways to take action in
current Prolog systems are kludges.


    We know by now that the
    most difficult issue in ``reactive systems'' is not EXPRESSING action,
    but rather keeping it under control to prevent unwanted interactions.
    In this area, less is REALLY more, and highly complex languages like
    Lisp are just not suitable for the formal reasoning about programs
    that is needed to help us believe our programs do what we intend. To
    those who say ``...but we can keep to a carefully constrained subset
    of Lisp, use only message passing for interactions,...'' I will answer
    that the history of all large Lisp programs I know (and I think that
    is a representative sample) is littered with the brutalized corpses of
    constrained programming styles. Anyone who has looked at the current
    flavor mechanism in Zetalisp and its use in the window system will
    know what I mean...

    5. Underlying Carl Hewitt's misconceptions is the old chestnut ``logic
    is static, systems are dynamic''.

Note that the above quotation is NOT anything that I said.

    Any language, be it first-order
    logic or Lisp, is static; it is its USE which is dynamic (changes the
    state of communicating agents). A good analogy here is the use of
    differential equations to model dynamic systems in classical
    mechanics. The differential equations themselves do not say directly
    what happens when (they are ``static'' in Hewitt's jargon).

I do not deny that dynamic systems can be DESCRIBED using logic only
that they can be CONTROLLED.

    It is
    their SOLUTIONS that tell us the sequence of events. Even the
    solutions are given as static objects (functions from an open interval
    of the reals to some space). Does anyone worry that such equations do
    not ``really'' describe the dynamic behavior of a system? Is it not
    possible to combine such ``static'' entities with an incremental
    solution procedure to build systems that actually control their
    (classical mechanical) environment?

I do not believe that the control system can be implemented using Logic
as a Programming language

∂19-Aug-85  1005	ullman@diablo 	events this week    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 19 Aug 85  10:04:58 PDT
Date: Mon, 19 Aug 85 09:59:12 pdt
From: Jeff Ullman <ullman@diablo>
Subject: events this week
To: nail@diablo

We'll meet 10AM in 301MJH as usual for our weekly gabfest.
Kathy Morris is going to lead a discussion on the structures
that she has set up to represent the rules in a NAIL! source
program.

We also have a special seminar on Thursday, 8/22 11AM in Rm. 352MJH.
COme find out about the mysterious "APEX" method of processing
logic.  A title and abstract are attached.
*************************

Inference by Generating and Structuring Deductive Databases
		E. L. Lozinskii
		Hebrew Univ.

A strong advantage of bottom-up generating techniques is that they
guarantee finiteness of all inferences in a deductive database system
involving recursive axioms.  But a "brute force" generation of answers
to a query would be very inefficient, producing many facts useless
for the query evaluation.  An economical generating method is presented,
based on discovering explicit facts relevant to the query and applying
preselected axioms as generating rules.  The method is complete,
in the sense that it generates all the existing answers to the query
in a finite time.
	An important problem of designing deductive
databases and their structuring is the question of what knowledge
should be represented by explicit facts and what by means of axioms.
A performance-oriented dynamic structuring of databases is proposed,
influenced by the query evaluation method described.

∂19-Aug-85  1516	avg@diablo 	linear approximations  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 19 Aug 85  15:16:43 PDT
Date: Mon, 19 Aug 85 15:06:21 pdt
From: Allen VanGelder <avg@diablo>
Subject: linear approximations
To: nail@diablo

Consider the NC program P1 in which q and e are EDB relations:
	p(X,Y) :- p(X,Z), q(Z,V), p(V,Y).
	p(X,Y) :- e(X,Y).
Apparently the following linear program P2 has the same minimum models:
	p(X,Y) :- e(X,Z), q(Z,V), p(V,Y).
	p(X,Y) :- e(X,Y).

Proof: Let < denote subset and let P1 stand for "min model of P1", etc.
Clearly P2 < P1.  For the other direction, the inductive hypothesis is
that all atoms derivable from P1 in k steps are also derivable from P2.
It is true for k=1.  For k > 1, the last step must use rule 1.  Say the
specific application is
	p(a,b) :- p(a,c), q(c,d), p(d,b).
Then P2 derives p(a,c) and p(d,b) because P1 derives them in fewer than k
steps.  It is easy to show that p(a,c) is derivable by P2 if and only if
there is a chain of EDB atoms:
	e(a,c0), q(c0,a1), e(a1,c1), q(c1,a2), ... , e(an, c)
(Just use induction on the number of steps P2 requires.)
Therefore P2 derives p(an,b) by means of
	p(an,b) :- e(an,c), q(c,d), p(d,b).
Working back down the chain, P2 derives p(a←{n-1}, b), ..., p(a1,b), p(a,b).
QED.

The proof is trivial, but I put it down so we could see what problems
arise in trying to generalize.  We can confine ourselves to one recursive
predicate p and one base rule for it, p :- e.  I believe these are not
essential restrictions.  Now we can consider 3 avenues of generalization:
	(1) Multiple recursive rules
	(2) Different arrangements of the variables among the subgoals,
	    leading to different types of joins.
	(3) More than 2 recursive subgoals per rule.

Here is P3 to test our mettle:
	p(X,Y) :- p(X,Z1), q1(Z1,V1), p(V1, Y).
	p(X,Y) :- p(X,Z2), q2(Z2,V2), p(Y, V2).
	p(X,Y) :- e(X,Y).

One possibly useful way to rewrite this is to use r(X,Y) = p(Y,X) and try
to make all derivations left-to-right joins.  To further this cause we use
q3(X,Y)= q1(Y,X)  and  q4(X,Y) = q2(Y,X)  and  f(X,Y) = e(Y,X).  We get:

	p(X,Y) :- p(X,Z1), q1(Z1,V1), p(V1, Y).
	p(X,Y) :- p(X,Z2), q2(Z2,V2), r(V2, Y).
	p(X,Y) :- e(X,Y).
	r(X,Y) :- r(X,V1), q3(V1,Z1), r(Z1, Y).
	r(X,Y) :- p(X,V2), q4(V2,Z2), r(Z2, Y).
	r(X,Y) :- f(X,Y).

Because of the regular structure we can omit the arguments and write these
rules as a CFG:
	P -> e | P q1 P | P q2 R
	R -> f | R q3 R | P q4 R
Beginning a Greibach normal form translation to remove RIGHT recursion
(the opposite HU), we remove R as rightmost symbol.
	R -> f | S f
	S -> R q3 | P q4 | S R q3 | S P q4
	P -> e | P q1 P | P q2 f | P q2 S f
But now we can get rid of R altogether:
	P -> e | P q1 P | P q2 f | P q2 S f
	S -> f q3 | S f q3 | P q4 | S f q3 | S S f q3 | S P q4
Continue by removing rightmost P:
	T -> P q1 | T P q1
	P -> e | P q2 f | P q2 S f | T e | T P q2 f | T P q2 S f
	S -> f q3 | S f q3 | P q4 | S f q3 | S S f q3 | S P q4

Now it is time to apply the paradigm
	"Greibach normal form + mistakes = linear recursion"
I'll start by forgetting S -> S S f q3.
	T -> P q1 | T P q1
	P -> e | T e | P q2 U f | T P q2 U f
	U -> [] | S	where [] = "empty"
	S -> V | S V
	V -> f q3 | P q4
Now with liberal abuse of notation, I'll represent U by the regular
expression (f q3 + P q4)*.
	T -> W P q1
	P -> W e | W P q2 (f q3 + P q4)* f
	W -> [] | T
It looks like W is (P q1)*, and (a+b)* = (a*b*)*, giving:
	P -> (P q1)* e  |  (P q1)* P q2 ((f q3)* (P q4)*)* f

I think this is correct, but can anybody do anything with it?

∂19-Aug-85  1620	@MIT-MC.ARPA:august@JPL-VLSI.ARPA 	Got to keep trying! Sorry about all the copies.   
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 19 Aug 85  16:19:51 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 19 AUG 85  19:20:32 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 19 Aug 85 19:18-EDT
Received: from JPL-VLSI.ARPA by MIT-MC.ARPA 19 Aug 85 19:19:14 EDT
Received: from vlsidc by VLSIDC with VMS ; Mon, 19 Aug 85 16:19:24 PDT
Date:    Mon, 19 Aug 85 16:19:24 PDT
From:     august@JPL-VLSI.ARPA
Subject: Got to keep trying! Sorry about all the copies.
To:       phil-sci%mit-oz%mit-mc@JPL-VLSI.ARPA, august@JPL-VLSI.ARPA

From:	VLSIDC::ST%"" 19-AUG-1985 16:16
To:	AUGUST
Subj:	Msg of Monday, 19 August 1985 19:15-EDT

Received: from MIT-MC.ARPA by JPL-VLSI.ARPA with INTERNET ; Mon, 19 Aug 85 16:16:24 PDT
Date: Mon, 19 Aug 85 19:15:25 EDT
From: Communications Satellite <COMSAT@MIT-MC.ARPA>
Subject: Msg of Monday, 19 August 1985 19:15-EDT
To: "august@JPL-VLSI.ARPA"@JPL-VLSI.ARPA
Message-ID: <[MIT-MC.ARPA].618302.850819>
 
============ A copy of your message is being returned, because: ============
"PHIL-SCI-REQUEMIT-OZ" at MIT-MC.ARPA is an unknown recipient.
============ Failed message follows: ============
Received: from JPL-VLSI.ARPA by MIT-MC.ARPA 19 Aug 85 19:14:26 EDT
Received: from vlsidc by VLSIDC with VMS ; Mon, 19 Aug 85 16:14:57 PDT
Date:    Mon, 19 Aug 85 16:14:57 PDT
From:     august@JPL-VLSI.ARPA
Subject: try again
To:       phil-sci-request%mit-oz%mit-mc@JPL-VLSI.ARPA, august@JPL-VLSI.ARPA
 
From:	VLSIDC::AUGUST       "Richard B. August" 19-AUG-1985 16:11
To:	ST%"request-phil-sci%mit-oz@mit-mc",AUGUST      
Subj:	Addition to mailing list
 
Please add me to the Phil-Sci list.
 
Regards,
Richard
 

∂19-Aug-85  2011	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #36
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Aug 85  19:27:36 PDT
Date: Monday, August 19, 1985 9:33AM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #36
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Tuesday, 20 Aug 1985      Volume 3 : Issue 36

Today's Topics:
               Query - LP Based Specification Language,
                   LP Philosophy - Hewett Challenge
----------------------------------------------------------------------

Date: Sat, 10-Aug-85 14:36:30 PDT
From: P. Allen Jensen gatech!gitpyr!allen@UCB-Vax
Subject: Prolog based Software Specification Language

I am trying to find information on Software Specification
Languages written in Prolog.  I have looked at two non
procedural description techniques, one based on regular
expressions, the other utilizing a data base approach.
Bibliographic information would be of help.  It seems
to me that Prolog would be a good language to use for
developing a software specification language.

-- P. Allen Jensen

------------------------------

Date: Sun, 11-Aug-85 10:34:32 PDT
From: P. Allen Jensen  gatech!gitpyr!allen@UCB-Vax
Subject: Program Specification Languages

I am doing some research on Program Specification Languages.
I am aware of only two system currently available:

o - Program Statement Language/Program Statement Analyzer
    (PSL/PSA) University of Michigan

o - Software Development System (SDS, SREM)
    Ballistic Missile Defense and TRW, Inc.

PSL uses Objects  and Relationships to  describe a system.
The language allows 22 possible objects and 36 relationships.
These descriptions are then analyzed by  PSA for  redundancies
or  logical  inconsistencies.  PSA,  however, is  not
rigorous  and therefore cannot provide a mathematically correct
verification of the logical consistency of the specifications.

I am  not  familiar with  SDS,  but understand  that it is more
extensive than PSL/PSA.

Any  further information  on currently  available products  or
research in this area would be  appreciated.  I am considering
developing a  prototype language for specifications in Prolog.
All comments and suggestions are welcome.

-- P. Allen Jensen

------------------------------

Date: Wed, 14 Aug 85 01:03:48 EDT
From: Carl E. Hewitt <HEWITT@MIT-MC.ARPA>
Subject: Prolog will fail as the foundation for AI; so will

           [ The following four messages are reprinted from
                       the Phil-Sci list. -ed ]

Folks,

This list has been rather dormant.  To liven things up, I
would like to throw out the following little ticker:

Prolog (like APL before it) will fail as the foundation for
Artificial Intelligence because of competition with Lisp.
There are commercially viable Prolog implementations written
in Lisp but not conversely.

LOGIC as a PROGRAMMING Language will fail as the foundation
for AI because:

1.  Logical inference cannot be used to infer the decisions
    that need to be taken in open systems because the decisions
    are not determined by system inputs.

2.  Logic does not cope well with the contradictory knowledge
    bases inherent in open systems.  It leaves out
    counterarguments and debate.

3.  Taking action does not fit within the logic paradigm.

------------------------------

Date: Wed 14 Aug 85 10:49:57-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Challenge Problem

I have long believed that any algorithm developed in a procedural
language could be reformulated more elegantly in a logic language.
This may be true, but I have come to doubt its utility.  Restating
an algorithm in nonprocedural form can be a useful exercise, but
does not improve on the procedural nature of the algorithm.
Advocates of logic programming would have a weak case if their
languages were just notations for telling the computer what to do,
step by step.

In order to sharpen my understanding of logic programming, I would
be interested in hearing about logic-based solutions to a problem
that I consider particularly "algorithmic": extraction of connected
components in images.  I would like to learn of any logic-based
approach to the problem that would entail neither massive search nor
such extensive procedural embedding as to make the "logic" portion
trivial.

The problem:  Assume that you have an array of integers, and are
to identify all maximal connected components -- i.e., all clusters
of identical integers.  The integers range from 0 to some known
maximum, with 0 being a special code that connects with the space
"outside" the array.  You are free to use any bounded intermediate
data structures.  The output is to consist of an array in which
elements of each connected component are marked with the same
integer code, and elements of different components are marked with
different codes.  Other outputs (boundary traces, lists of region
adjacencies and inclusion relationships, number of holes per
region, etc.) are desirable but of secondary importance.

I know of several good algorithms for solving this problem.  Most
involve an intermediate map or equivalent data structure that records
nonmaximal connected components found during a row-by-row scan,
together with a list of discovered equivalences that can be used to
link these into maximal components.  Such algorithms are quite
efficient (aside from having to scan the array completly, in the
worse case, before starting to build the output map).  Others involve
visiting (and marking) every point in the array, tracing region
boundaries until they close and then raster scanning to find the
next unvisited point.  (This is a reasonable procedure if the
boundary traces are of primary interest.)

Is there any hope that a logic-based language could compile a
[nonalgorithmic] declarative statement of this problem into similar
code, or would the compiler have to be guided by domain knowledge that
essentially supplies it with the right answer?  If the latter
is true, in what sense is logic programming the answer to our
current needs?

-- Ken Laws

------------------------------

Date: Wed, 14 Aug 85 10:23:10 cdt
From: Kenneth Forbus Forbus%uiucdcsp@Uiuc
Subject: Prolog will fail as the foundation for AI

(Carl, I wish you had found a fresher topic to re-open the
list with. And doing it right before conference season...)

To equate "prolog" with "logic programing, as Carl & Wayne's
messages comes very close to doing, is somewhat less than
accurate.  True, prolog is currently the logic programming
language of choice.  True, prolog provides a handy notion
of variable that takes painful effort to build in Lisp.
But prolog really is "Micro-Planner done right", as its
advocates claim, and that is enough to allow most AI people
to tune out prolog because we found out a long time ago that
Micro-Planner isn't what we want.

Prolog suffers from being both too old and too new.  It is
too old, in the sense that it has chronological backtracking
built into its roots and the standard versions don't take
into account new ideas (such as TMS').  Yes, of course one
can encode these things in logic and "run them" in Prolog
(If you really believe prolog is "programming in logic" I
have a bridge to sell you).  However, you may not get an
answer in the lifetime of the universe, and doing three or
four experiments will have you sitting fruitlessly at the
console long enough to write a fast lisp program to do the
same thing!  But prolog is too new, in that it has also not
kept pace with progress in programming environments and
language design.  Talking to prolog programmers is a bit
like talking to lisp programmers 10 years ago -- there are
many dialects, often changing rapidly, wild disagreements
on syntax, etc.  The days of CommonProlog are likely to be
pretty far off.

Judging the idea of logic programming by prolog is a bit like
judging lisp by Lisp 1.5.  I believe that only a tiny fraction
of the interesting ideas which should be encapsulated in a real
logic programming language have been developed.  If some venture
capitalist were to ask me into whose pocket should he put $1M to
get a signficant advance in the state of logic programming, I
wouldn't suggest spending it on the prolog community.  I'd
probably suggest splitting it between McAllester & de Kleer,
whose ideas on TMS', while rather different, are both far more
interesting and potentially productive ideas than, say, improving
the speed of unification or studying and-parallelism.  We simply
don't have enough experience writing and using reasoning languages
yet to either cast our ideas into stone (i.e., adopt prolog and
spend our time optimizing it) OR pass final judgement on the
merits of logic programming.

On logic and action:  Why shouldn't Shakey be considered a counter
example to Carl's claim about the impossibility of action in a
system based on logic?  Shakey used resolution theorem proving to
decide what to do and other kinds of routines (such as A* search)
to flesh out the details.  If the existence of these other routines
is considered sufficient to discount Shakey as a counterexample
then the claim seems rather uninterestingly obvious.

------------------------------

Date: Wed 14 Aug 85 12:07:05-EDT
From: Vijay <Vijay.Saraswat@CMU-CS-C.ARPA>
Subject: On logic programming as a foundation for AI.

This is a response to some of the issues recently raised
by Hewitt in a post to the PHIL-SCI digest. (Phil SCI
Digest, Aug 14).

For the sake of argument,  Prolog is probably a `more convenient'
language for writing compilers  than Lisp; on the other hand,
the current state of sophistication of Lisp implementations makes
them extremely attractive for developing rapid prototype
implementations.   For the near future I see the ideal programming
environment as being a Lisp Machine with a very fast Prolog on it
(If Symbolics Prolog lives up to rumours of 100K Lips that will
do, thank you.)  Why would anyone EVER want to write a CommonLisp
interpreter in  Prolog?  There are much better things to do ...
particularly when you already have CommonLisp and Prolog!

What does this have to do with "being a foundation for AI"?
If he means that "X is a foundation for AI" if X is used by
most people for writing their AI programs, then that is
irrelevant.  I would say that "X is a foundation for AI"
if most work being carried out in AI can fit in the theoretical
framework of X. With such a working definition it is rather
doubtful if any one paradigm can be a foundation for AI.  But I
would submit that the concept of logic programming is a step
towards reaching such a foundational understanding.

"Logic as a programming language will fail as the foundations
for AI because..."

Let's be careful how we bandy terms around.  Logic is not a
programming language. Logic is logic, a formal system. The
first axiom of logic programming (the LOGIC axiom) is that
the definite clause subset of logic has an appealing
interpretation as a programming language, if the process of
SLD-refutation (which is complete for this subset) is taken
as the inference mechansim.  In this programming language,
the user has DON'T KNOW CHOICE, i.e. the power of specifying
existential searches.  This power is rather unnerving.  The
second axiom of logic programming (the CONTROL axiom) is that
it is possible to provide general control mechanisms which can
be exploited by a programmer for controlling the search. This
led to the rise of logic programming languages, which allow,
in general, the programming of incomplete searches, (hence the
handling of non-montonic inference, defaults, inheritance ...).

The most ancient of these is PROLOG, which essentially provided
the control structure of SEQUENTIALITY. The language PROLOG has
something to do with logic of course, but, because of its
incomplete proof procedure, its semantics is best given via
denotational means, rather than logical means.  This is the
first corollary of logic programming: a formal understanding of
logic programming languages has to appeal to traditional computer
science techniques for giving semantics to programming languages.
The surprising discovery is that because of the simplicity of the
underlying execution mechanism such semantics are surprisingly
simple: far simpler than semantics for ALGOL, LISP (has anyone
ever attempted a semantics for something like COMMONLISP?),
ACTORS...  This has very powerful connotations with respect to
semantically based programming environments, program
transformations, and meta-programming.

There have been more recent languages which have provided other
control features: PARLOG, GHC, Concurrent Prolog and CP[!,|,&],
to name just a few in the main stream of Horn logic programming,
have attempted to also provide don't care choice and parallelism.
To frame it in the current parlance, these languages essentially
provide support for object-oriented programming.  While a formal
denotational semantics for the last is still being developed
(the others do not yet have formal semantics), it has already
been demonstrated that it has a surprisingly simple partial
correctness semantics.

More important, these languages demonstrate that while
keeping the LOGIC component more or less identical, it is
possible to achieve a great variety of operational
behaviours by changing the CONTROL component.  Hence logic
programming is not a LANGUAGE, it is a PROGRAMMING PARADIGM
encompassing all these operational behaviours.

To sum up, logic programming langauges are not just any
programming languages, at some level, most programs written
in such languages have a declarative interpretation
compatible with a logic system, typically universally quantified
definite clause logic. Logic programming languages are not just
logic systems because their control component plays an important
part.  The art of designing logic programming languages is
concerned with maintaining a delicate balance between these two
divergent themes. Even with their control components, logic
programming languages can generally be given simple semantics,
which reflects their underlying conceptual simplicity.

As far as supporting most of the current AI paradigms is concerned,
it should be clear that Definite clause logic supports naturally
the notions of  goal-driven backward chaining.

It is my contention (as yet unsupported) that the basic style of
computation provided by Concurrent Prolog and CP[!,|,&] naturally
supports data-driven, forward-chaining inference and also knowledge
structuring a la semantic network based languages, again via the
primary mechanism of message-passing.  ("reserach in progress",
though there is an article by Shapiro andd Takeuchi on object
oriented programming in Concurrent Prolog).

Hence I would conclude that, to the extent that logic programming
languages support such  programming paradigms, and to the extent
that they themselves have secure theoretical foundations, one can
at least claim that logic programming offers a step towards a
unified understanding of the foundations of (programming paradigms
used in) AI.

Note that I am not saying anything at all about the psychological
plausibility of controlled inference as the essential "problem
solving capability" in  intelligent agents.

Let me now discuss three specific poinst raised by Hewitt:

"1.  Logical inference cannot be used to infer the decisions that
need to be taken in open systems because the decisions are not
determined by system inputs.

2.   Logic does not cope well withthe contradictory knowledge
bases inherent in open systems.  It leaves out counterarguments
and debate.

3.   Taking action does not fit within the logic paradigm."

Some of these contentions have been made by Hewitt in other
contexts and I still remain as mystified by them as I was then.

Let us keep in mind that we are talking of programming langauges,
albeit peculiar programming languages. How does a LISP function
'take action'? Presumably by doing a (setq foo 'take-action).
How does an OPS-5 program take action? Presumably by executing a
(make task-312 ↑task take-action) on the right hand side of a
production. So also  logic programming languages take action by
instantiating a variable to some value. If that variable was
actually implemented as a particular memory cell which is being
monitored by ahard-wired coke-dispenser, then it would 'take
action': deliver a coke bottle.

Presumably what is confusing Hewitt is that in Prolog bindings
can be done on backtracking.  But all that is in the programmer's
control! If the programmer intended that the action to be taken
is irrevocable, he would write his axioms in such a fashion that
the binding would not be backtracked over. That is a control
problem!

About contradictory data bases.  First let me make the obvious
statement that a certain level of understanding, every set of
definite clause has a model, indeed an infinity of models. Hence
there is no scope for representing contradictory knowledge
directly via axioms in a Horn logic program.  So how do logic
programming systems do it?  Well you have to PROGRAM IN your
handling of such data.  You have to write a program, just as you
would write a program in LISP or ATOLIA which does to this data
what you want to do with it.  The axioms you write (the program
you write) will be executed in the fashion determined by the logic
programming language you choose and by the control you provide:
usually this approximates some kind of inferencing. BUt the action
that gets taken when you execute your program depends upon your
program.  If you had written your program such that when you
encounter an inconsistency it would print out "The moon is made of
green cheese" and stop, then, if an inconsitency is encountered,
believe it or not, it will print out "The moon is made of green
cheese" and stop. If you wanted to program in counterargumentsd
and debate then go ahead and do that.  Logic programming does not
provide "counterarguments and debate" as a primtive concept.
The advantage that you get by writing your program in a logic
programming language is that you can reason about it, you can
(hopefully) prove that when the program encounters an inconsistency
it will print out "The moon is made of greencheese" and terminate.
And because of the simplicity of the programming paradigm that you
are using, your proofs would be relatively simple.  If you were
careful in writing your program and in choosing your logic
programming langauge, the answer that you would get would be a
logical consequence of YOUR AXIOMS (program) which of course COULD
HAVE DONE whatever it wanted to to the data.

Hence, while on the face of it Hewitt's statement (2) is true,
it is totally irrelevant.

This also takes care of the argument that "logical inference
cannot be used to  infer decisions that need to be taken in open
systems ..." because logical inference is being used to execute
YOUR PROGRAM, which determines how to make those decisions.  If
your program is actually an OPS-5 interpreter which acts on the
data that it gets by using a knowledge base of productions, then,
by heavens, YOUR PROGRAM is NOT doing logical inference.  Hence
even if Hewitt's contention is correct, it is totally irrelevant.

-- Vijay A. Saraswat

------------------------------

End of PROLOG Digest
********************

∂20-Aug-85  1240	@SU-SUSHI.ARPA:ANDERSON@SU-SCORE.ARPA 	Complexity Year info 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Aug 85  12:40:21 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 20 Aug 85 12:38:04-PDT
Date: Tue 20 Aug 85 12:38:00-PDT
From: Richard Anderson <ANDERSON@SU-SCORE.ARPA>
Subject: Complexity Year info
To: aflb.local@SU-SCORE.ARPA
Message-ID: <12136695215.9.ANDERSON@SU-SCORE.ARPA>

This year Complexity Year is being held at the Mathematical Sciences
Research Institute (MSRI) at Berkeley.  A large number of computer
scientists will be at the institute during the year.

There will be talks on Tuesdays at 2 pm and 4 pm, with a break for
afternoon tea.  I will attempt to keep you informed as to what the talks
will be.  I will maintain a mailing list of people wanting to know what the
talks will be, so send me a note if you want to be included in the 
mailing list.

MSRI is located in the Berkeley hills, just above the Lawrence Hall of
Science.  I can provide more complete and more complicated directions on how 
to get there.

Professor C. P. Schnorr will be giving a talk this Friday, August 23 at
2 pm in the MSRI Lecture Hall.  The title of the talk is:
	Polynomial time algorithms for integer relations,
	and lattice basis reduction

-- Richard

-------

∂20-Aug-85  2251	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	"Official" FOCS airline  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Aug 85  22:51:21 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 20 Aug 85 22:49:00-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Tue 20 Aug 85 22:49:17-PDT
Received: by wisc-rsch.arpa; Wed, 21 Aug 85 00:35:03 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Tue, 20 Aug 85 18:41:02 cdt
Message-Id: <8508202340.AA04487@wisc-crys.arpa>
Received: from csnet-relay.arpa by wisc-crys.arpa; Tue, 20 Aug 85 18:40:51 cdt
Received: from uoregon by csnet-relay.csnet id ac09844; 20 Aug 85 19:34 EDT
Received: by uo-vax3.UONET (4.12/1.0.uoregon)
	id AA01054; Tue, 20 Aug 85 11:17:49 pdt
Date: Tue, 20 Aug 85 11:17:49 pdt
From: Fall Conference Mail <focs%uoregon.csnet@csnet-relay.arpa>
Posted-Date: Tue, 20 Aug 85 11:17:49 pdt
To: theory@WISC-CRYS.ARPA
Subject: "Official" FOCS airline
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;



Here are some more details on United Airlines offer to FOCS travelers.

United Airlines has been designated as the "Official Carrier for FOCS '86"
UA will offer special airfares, not available to the general public, when
you attend FOCS and travel between October 18, 1985 and October 26, 1985.

You can obtain a 35% discount from the unrestricted Day Coach fare OR a 
15% discount from the Easy Saver fare (requiring a Saturday night stay).
The discounts are available only on United Airlines flights in the
continental U.S.  To take advantage of these fares:

    1. Call United's Convention Desk toll-free at 800-521-4041, Monday-Friday,
       8:30am-8:00pm Eastern time.  Your travel agent may make the call.

    2. Give the Focs account number 562T.

    3. United specialists will provide information and make reservations for
       all United flights and fares, including the special FOCS fares.

    4. United can arrange to mail tickets to your home or office, or you may
       purchase them from your travel agent.
      
Seats are limited, so call early to ensure availability.  Fares are
guaranteed at time of purchase.


------------------------------------------------------------------------------

Please feel free to direct any questions relative to FOCS local arrangements
to   focs@uoregon





∂20-Aug-85  2302	ullman@diablo  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 20 Aug 85  23:02:25 PDT
Date: Tue, 20 Aug 85 22:58:36 pdt
From: Jeff Ullman <ullman@diablo>
To: Jskud@diablo, avg@diablo, cohn@diablo, freeman@diablo, karlin@diablo,
        kuper@diablo, morris@diablo, nail@diablo, naughton@score, pph@sail,
        roy@diablo, trickey@diablo

Here are some more details on United Airlines offer to FOCS travelers.

United Airlines has been designated as the "Official Carrier for FOCS '86"
UA will offer special airfares, not available to the general public, when
you attend FOCS and travel between October 18, 1985 and October 26, 1985.

You can obtain a 35% discount from the unrestricted Day Coach fare OR a 
15% discount from the Easy Saver fare (requiring a Saturday night stay).
The discounts are available only on United Airlines flights in the
continental U.S.  To take advantage of these fares:

    1. Call United's Convention Desk toll-free at 800-521-4041, Monday-Friday,
       8:30am-8:00pm Eastern time.  Your travel agent may make the call.

    2. Give the Focs account number 562T.

    3. United specialists will provide information and make reservations for
       all United flights and fares, including the special FOCS fares.

    4. United can arrange to mail tickets to your home or office, or you may
       purchase them from your travel agent.
      
Seats are limited, so call early to ensure availability.  Fares are
guaranteed at time of purchase.


------------------------------------------------------------------------------

Please feel free to direct any questions relative to FOCS local arrangements
to   focs@uoregon

∂20-Aug-85  2302	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Roommates for FOCS  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Aug 85  23:02:38 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 20 Aug 85 22:49:47-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Tue 20 Aug 85 22:50:02-PDT
Received: by wisc-rsch.arpa; Wed, 21 Aug 85 00:35:47 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Tue, 20 Aug 85 18:41:48 cdt
Message-Id: <8508202341.AA04494@wisc-crys.arpa>
Received: from csnet-relay.arpa by wisc-crys.arpa; Tue, 20 Aug 85 18:41:39 cdt
Received: from uoregon by csnet-relay.csnet id af09844; 20 Aug 85 19:35 EDT
Received: by uo-vax3.UONET (4.12/1.0.uoregon)
	id AA02160; Tue, 20 Aug 85 13:25:58 pdt
Date: Tue, 20 Aug 85 13:25:58 pdt
From: Fall Conference Mail <focs%uoregon.csnet@csnet-relay.arpa>
Posted-Date: Tue, 20 Aug 85 13:25:58 pdt
To: theory@WISC-CRYS.ARPA
Subject: Roommates for FOCS
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;



                Looking for a roommate for FOCS meeting?

We are maintaining a listing of those looking for person(s) with whom to
share a hotel room at FOCS.  Send following information to focs@uoregon

   Name ..................................................
   Institution ...........................................
   Phone number ..........................................
   Electronic address ....................................
   Student [ ]    Non-student [ ]
   Desired number of room occupants (2,3,4) .......

Note: To obtain special room discount for students, all occupants must be
full-time students.


We shall mail complete listing to all those on list twice a week.  We leave it
to individuals to contact one another.  Please send us a note when you have
found roommates so that we can remove your name from list.





∂21-Aug-85  0912	EMMA@SU-CSLI.ARPA 	Ventura Hall and air conditioning   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 21 Aug 85  09:11:58 PDT
Date: Wed 21 Aug 85 09:04:27-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Ventura Hall and air conditioning
To: bboard@SU-CSLI.ARPA
cc: folks@SU-CSLI.ARPA
Tel:  497-3479


  The air conditioning in Ventura Hall has been turned off TODAY
(August 21).

Please make sure that your dandelions do not suffer from heat stroke;
turn them off when the room gets too warm.

This message does not apply to the trailers.

Any questions should be sent to Jamie@csli.

-------

∂21-Aug-85  0925	ullman@diablo 	don't forget   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 21 Aug 85  09:25:27 PDT
Date: Wed, 21 Aug 85 09:20:20 pdt
From: Jeff Ullman <ullman@diablo>
Subject: don't forget
To: nail@diablo

meeting 10AM today, 301 MJH
Lozinskii talk, 11AM tomorrow (Thursday), 352 MJH
Shapiro talk, 9AM Monday 8/26, 352 MJH

∂21-Aug-85  1158	STUCKY@SU-CSLI.ARPA 	D-lion Users  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 21 Aug 85  11:58:21 PDT
Date: Wed 21 Aug 85 11:50:28-PDT
From: Susan Stucky <Stucky@SU-CSLI.ARPA>
Subject: D-lion Users
To: folks@SU-CSLI.ARPA
cc: consultants@SU-CSLI.ARPA

Dear All, 

Since many of us are at different sites, the usual method for
disseminating information about machines (i.e., walking down the hall
and asking) isn't available.  And then, much of the information I know
of, at least, of is at the level of hearsay.  For instance, rumor has it
that Kris Halvorsen has a program that takes LaTex files and translates
them into Tedit files (or is it the other way around??).  And I notice
other people's screens and wonder where the program is that does THAT.

Anyway, Jon and I thought a D-lion users fest was in order. We suggest
an informal meeting, in the Trailer Dandelion room this next Tuesday,
the 27th at 1:30 p.m., say, for an hour and a half or so.  Though we
hope we can ask of the experts all our naive questions, we thought too
that the experts at various sites might have something to share with
each other.

Bring ideas and questions, in particular, ideas and questions for
across-site problems.  Getting files from one machine to another.  From
one editor to another.  We hope you'll come.

-Susan
-------

∂21-Aug-85  1658	ullman@diablo 	paper received 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 21 Aug 85  16:58:13 PDT
Date: Wed, 21 Aug 85 16:52:22 pdt
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

Anyone like to see: "LOGIN: a logic programming language with
built-in inheritance" by H. Ait-Kaci and R. Nasr?

∂21-Aug-85  1736	EMMA@SU-CSLI.ARPA 	Newsletter August 22, No. 42   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 21 Aug 85  17:36:08 PDT
Date: Wed 21 Aug 85 17:22:48-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter August 22, No. 42
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
August 22, 1985                 Stanford                       Vol. 2, No. 42
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
           CSLI ACTIVITIES FOR *NEXT* THURSDAY, August 29, 1985

   12 noon		TINLunch
     Ventura Hall       ``A Unified Indexical Analysis of ``same'' and
     Conference Room    ``different'': A Response to Stump and Carlson''
			by David Dowty
			Discussion led by Mats Rooth
			(Abstract on page 1)
		
   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week
     Conference Room	
			

   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←
                              ANNOUNCEMENTS

   No activities have been scheduled for this Thursday, August 22.  Next
   week, TINLunch will resume.
                              ←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
      ``A Unified Indexical Analysis of ``same'' and ``different'':
                    A Response to Stump and Carlson''

      Stump's ``A GPSG Fragment for `Dependent Nominals' '' is concerned
   with sentences such as 
         ``Mary saw ``Amadeus,'' but John saw a different movie.''
   and
            ``Every student saw a different movie.'' 
   where the interpretation of ``different movie'' is said to be
   dependent on the interpretation of another NP in the sentence or
   discourse.  Stump's analysis involves quantifier storage; Dowty
   criticizes some of the data which motivated this approach, and
   proposes an indexical or contextual analysis which posits a number of
   free variables in the interpretaton of ``same'' and ``different,'' the
   interpretation of which is determined by context.  In the second
   example above, the interpretation of ``different'' includes a free
   variable which is bound by the quantifier ``every student.''
							--Mats Rooth
!
Page 2                     CSLI Newsletter                    August 22, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                     CSLI WORKSHOP ON GERMAN GRAMMAR

      On Monday, August 26, and Tuesday, August 27, a small and rather
   informal workshop on problems of German syntax and semantics will be
   held at CSLI.
      The event was not planned as a conference-style workshop with
   fixed-length papers, restricted discussion periods and a large
   non-participating audience.  The goal is rather to present affiliates
   and visitors of CSLI who have worked on different theoretically
   interesting aspects of German with an opportunity to learn about each
   other's work and to get feedback on their own results.  We consider
   the mix of syntacticians and semanticists among the participants
   especially fortunate for the success of the workshop.
      Drop-in participants are welcome but are forwarned not to rely too
   much on the announced schedule since talks and discussions may
   overrun.  On both days, participants will be invited to a simple lunch
   on the trailer patio.

   Monday morning 9-12:

   David Perlmutter		On German Causatives
   (UCSD)

   Mark Johnson			Constituent Structure of the
   (Stanford and CSLI)		German VP

   Hans Uszkoreit		Ordering Principles
   (SRI and CSLI)

   Lunch on the patio

   Monday afternoon 1:30:

   John Nerbonne		Tense and Temporal Adverbs
   (HP and CSLI)

      After John's talk, interested participants will be given a brief
   overview over projects on German in the area of computational
   linguistics.  Guenter Goerz (U. Erlangen), Manfred Pinkal (U.
   Duesseldorf), Uwe Reyle (U. Stuttgart), and Hans Uszkoreit (SRI and
   CSLI) will give 10 minute talks on recent and current projects such as
   HAM-ANS, KIT, METAL, PLIDIS, EVAR, SUSY, the Stuttgart LFG
   implementation, GPSG in Berlin.

   Tuesday morning 9-12:

   Godehard Link		Generalized Quantifiers and Plural:
   (U. Muenchen and CSLI) 	The case of German 'je' (each)

   Dietmar Zaefferer		Bare Plurals and Naked Relatives:
   (U. Muenchen and CSLI)	Semantics of German wh-constructions

   Manfred Pinkal		Syntactic and Semantic Gender
   (U. Duesseldorf)

   Lunch on the patio
   Departure of the participants to their offices
!
Page 3                     CSLI Newsletter                     August 22, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
     ``Logophoricity: SELF and SOURCE in Discourse and Morphology''
              Summary of the meeting on Thursday, August 15

      Peter Sells gave a presentation on the notion of ``logophoricity''
   that is grammatically expressed in many languages, and proposed that
   there are more primitive notions of SELF and SOURCE in terms of which
   logophoric domains are created and perpetuated.  He proposed that the
   grammatical conditions on the antecedent of the Japanese reflexive
   `zibun' are that:

      (a) its antecedent is a grammatical subject, or
      (b) its antecedent realises the SELF of the discourse

      The SELF may be realised in two ways, either concomitant with the
   SOURCE of a verb of communication, as with `Max' in ``Mary heard from
   Max that ...''  or ``Max said that ....''  Alternatively, the SOURCE
   may be the external speaker, as with the `psychological' predicates,
   such as ``That so-and-so distressed Max;'' again `Max' realises the
   SELF here.  Intuitively, psychological predicates are those predicates
   with which an external speaker says something about a mental state of
   a sentence-internal protagonist.  A simple notion of logophoricity
   cannot distinguish these two cases.
      A framework for representing these constructs was given, in terms
   of the Discourse Representation Theory developed by Kamp, and various
   differences between the communicational and psychological predicates
   were discussed.
      There was also discussion of the English prefix ``self-,'' as in
   ``self-confidence,'' which also gives evidence in favor of the
   constructs proposed by Sells.  In particular, nouns like
   ``self-deception'' give a clear indication that the speaker is
   classifying the mental state of some other person.	--Peter Sells

-------

∂21-Aug-85  2018	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:Robert←S.←Printis.osbunorth@Xerox.ARPA 	Re: Complexity Year info 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 21 Aug 85  20:17:55 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 21 Aug 85 20:13:04-PDT
Received: from Xerox.ARPA by SU-SCORE.ARPA with TCP; Wed 21 Aug 85 20:13:19-PDT
Received: from Salvador.ms by ArpaGateway.ms ; 21 AUG 85 20:15:11 PDT
Sender: "Robert S. Printis.osbunorth"@Xerox.ARPA
Date: 21 Aug 85 20:14:04 PDT (Wednesday)
Subject: Re: Complexity Year info
From: Printis.osbunorth@Xerox.ARPA
To: ANDERSON@SU-SCORE.Arpa
cc: aflb.local@SU-SCORE.Arpa
In-Reply-to: ANDERSON%SU-SCORE:ARPA:Xerox's message of 20-Aug-85
 22:00:10 
Message-ID: <850821-201511-4020@Xerox>

Please add my name to the distribution list.  Thanks,
 
Bob

∂22-Aug-85  1609	MARJORIE@SU-CSLI.ARPA 	phone lines 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Aug 85  16:08:53 PDT
Date: Thu 22 Aug 85 16:04:36-PDT
From: Marjorie Maxwell <MARJORIE@SU-CSLI.ARPA>
Subject: phone lines
To: folks@SU-CSLI.ARPA

This is to inform everyone of our change of phone lines as there seems
to be some confusion.  Rich Cower's line is now 1909. The programmers/
consulants now share 1932. Brad Horak has 2607, Clay Andres 5565, Joe
Zingheim 2658, and I remain at 9030. 
Marj
-------

∂22-Aug-85  1651	JAMIE@SU-CSLI.ARPA 	security  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Aug 85  16:51:40 PDT
Date: Thu 22 Aug 85 16:45:17-PDT
From: Jamie Marks <JAMIE@SU-CSLI.ARPA>
Subject: security
To: folks@SU-CSLI.ARPA




Two people recently reported lost belongings.  I feel I ought to warn
you to be careful about locking up, and about leaving purses, briefcases,
packs, etc. out in the open when you're not in the office.

Please let me know if other things are missing.


					- Jamie
-------

∂22-Aug-85  2229	JAMIE@SU-CSLI.ARPA 	security  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Aug 85  22:29:00 PDT
Date: Thu 22 Aug 85 17:24:47-PDT
From: Jamie Marks <JAMIE@SU-CSLI.ARPA>
Subject: security
To: folks@SU-CSLI.ARPA




I should have worded my previous message more strongly.  Here's the text of a
message I just received:

	Both thefts were of things that WERE locked up. It's
	misleading, therefore, to warn people to lock up and
	not leave things in the open. They should be warned not
	to leave any valuables in their offices overnight.

Once again, please let me know if anything is missing.

					- Jamie

-------

∂23-Aug-85  0905	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Messages to SIAM    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 23 Aug 85  09:05:05 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 23 Aug 85 08:59:34-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Fri 23 Aug 85 08:53:56-PDT
Received: by wisc-rsch.arpa; Thu, 22 Aug 85 20:52:03 cdt
Received: from SU-SCORE.ARPA by wisc-rsch.arpa; Thu, 22 Aug 85 18:27:33 cdt
Date: Thu 22 Aug 85 16:25:14-PDT
From: Gene Golub <GOLUB@SU-SCORE.ARPA>
Subject: Messages to SIAM
To: udi@WISC-RSCH.ARPA
Message-Id: <12137260870.24.GOLUB@SU-SCORE.ARPA>
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

It is now possible to correspond directly by e-mail to SIAM. Correspondence
concerning manuscripts, meetings, dues, etc can be sent to SIAM@SCORE.

Gene Golub
President,SIAM
-------


∂25-Aug-85  0130	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #13  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 25 Aug 85  01:30:49 PDT
Date: 25 Aug 85 0107-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #13
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Sunday, 25 Aug 1985       Volume 1 : Issue 13

Today's Topics:

                           Query: IBM GF11
              Parallel algorithms for linear programming
                          Symbolic computation


----------------------------------------------------------------------


[Your moderator will be away from his desk for a couple of weeks.
Submissions and requests will be buffered until September 8.  If
anyone has summaries of or comments on recent conferences relevant to
PARSYM (e.g., IJCAI or the Parallel Computing conference), please send
them in.  I am sure that many readers would be interested.  -- BD]


Date: Sat, 17 Aug 1985  07:44 EDT
Sender: GZT.TDF%MIT-OZ@MIT-MC.ARPA
From: "David D. Story" <FTD%MIT-OZ @ MIT-MC.ARPA>
Subject: IBM GF11

Has anyone out there looked at SIGARCH proceeding of last month.  I
was wondering what comments might be found on IBM's GF11.  I heard of
this machine years ago.  I thought at that time it was for a group of
problems General Foods had taken to the Yorktown Group.

I see that they have added some new micro-code features resembling
Hewitt's Actors and Semantic Massage for a society of experts.  Though
I dunno as of yet what is happening with AI at Yorktown it seems
they've tried to get this under the hood.  If the machine was meant
for the task of computing chromodynamics seems to me that the word
size would have been re-engineered since there is no mention of double
or quad wording in the article.  Whatta think ?

				Dave

------------------------------

Date: Friday, 16 August 1985  14:59-PDT
From: Deepak D. Sherlekar <deepak at maryland>
To:   PARSYM-REQUEST at SUMEX-AIM
Re:   Parallel Algorithm for Linear Programming.

Linear Programming is P-complete.  There is strong evidence suggesting
that P-Complete problems are unlikely to have fast parallel solutions.
By 'fast' is meant an algorithm that runs in time O((logn)↑k) for some
fixed k on a shared memory parallel random access machine (PRAM) using
polynomial number of processors.  Some references to literature on
P-Complete problems and the theory of synchronous parallel algorithms
that I could find/remember readily are:

1.	A.K. Chandra, D.C. Kozen and L.J. Stockmeyer, "Alternation", JACM, '81.

2.	S.A. Cook, "Towards a Complexity Theory of Synchronous Parallel
	Computation", (U of Toronto TR??).

3.	D. Dobkin, R. Lipton and S. Reiss, "Linear Programming is P-Complete",
	in (same authors) "Excursions into Geometry", Rept. No. 71, Dept.
	of Computer Sc., Yale U.

4.	M.R. Garey and D.S. Johnson, "Computers and Intractability: A Guide
	to the Theory of NP-Completeness", Chapter 7.6.

5.	L.M. Goldschlager, "Synchronous Parallel Computation", Ph.D. thesis,
	U. of Toronto.  See also Proc. 10th STOC and some '82 JACM.

--  Deepak

------------------------------

Date: Wed, 21 Aug 85 12:41:54 BST
From: Fitch@ucl-cs.arpa
Subject: Symbolic Computation

Guy Steele has not got it right with his example of rings, fields and
groups.  I compute over a field (of rational polynomials), and this is
normally considered symbolic.

John Fitch

------------------------------

End of PARSYM Digest
********************

∂25-Aug-85  1031	DAVIES@SUMEX-AIM.ARPA 	Vacation    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 25 Aug 85  10:31:07 PDT
Date: Sun, 25 Aug 1985  10:29 PDT
Message-ID: <DAVIES.12137982517.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: Vacation
cc:   Davies@SUMEX-AIM.ARPA

I'll be gone on vacation for the next two weeks.  I'm asking Jim Rice
to look after weekly meetings, if there are any.  See you in
September.

	-- Byron

∂26-Aug-85  0059	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #36
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 26 Aug 85  00:57:31 PDT
Date: Monday, August 19, 1985 9:33AM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #36
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Tuesday, 20 Aug 1985      Volume 3 : Issue 36

Today's Topics:
               Query - LP Based Specification Language,
                   LP Philosophy - Hewett Challenge
----------------------------------------------------------------------

Date: Sat, 10-Aug-85 14:36:30 PDT
From: P. Allen Jensen gatech!gitpyr!allen@UCB-Vax
Subject: Prolog based Software Specification Language

I am trying to find information on Software Specification
Languages written in Prolog.  I have looked at two non
procedural description techniques, one based on regular
expressions, the other utilizing a data base approach.
Bibliographic information would be of help.  It seems
to me that Prolog would be a good language to use for
developing a software specification language.

-- P. Allen Jensen

------------------------------

Date: Sun, 11-Aug-85 10:34:32 PDT
From: P. Allen Jensen  gatech!gitpyr!allen@UCB-Vax
Subject: Program Specification Languages

I am doing some research on Program Specification Languages.
I am aware of only two system currently available:

o - Program Statement Language/Program Statement Analyzer
    (PSL/PSA) University of Michigan

o - Software Development System (SDS, SREM)
    Ballistic Missile Defense and TRW, Inc.

PSL uses Objects  and Relationships to  describe a system.
The language allows 22 possible objects and 36 relationships.
These descriptions are then analyzed by  PSA for  redundancies
or  logical  inconsistencies.  PSA,  however, is  not
rigorous  and therefore cannot provide a mathematically correct
verification of the logical consistency of the specifications.

I am  not  familiar with  SDS,  but understand  that it is more
extensive than PSL/PSA.

Any  further information  on currently  available products  or
research in this area would be  appreciated.  I am considering
developing a  prototype language for specifications in Prolog.
All comments and suggestions are welcome.

-- P. Allen Jensen

------------------------------

Date: Wed, 14 Aug 85 01:03:48 EDT
From: Carl E. Hewitt <HEWITT@MIT-MC.ARPA>
Subject: Prolog will fail as the foundation for AI; so will

           [ The following four messages are reprinted from
                       the Phil-Sci list. -ed ]

Folks,

This list has been rather dormant.  To liven things up, I
would like to throw out the following little ticker:

Prolog (like APL before it) will fail as the foundation for
Artificial Intelligence because of competition with Lisp.
There are commercially viable Prolog implementations written
in Lisp but not conversely.

LOGIC as a PROGRAMMING Language will fail as the foundation
for AI because:

1.  Logical inference cannot be used to infer the decisions
    that need to be taken in open systems because the decisions
    are not determined by system inputs.

2.  Logic does not cope well with the contradictory knowledge
    bases inherent in open systems.  It leaves out
    counterarguments and debate.

3.  Taking action does not fit within the logic paradigm.

------------------------------

Date: Wed 14 Aug 85 10:49:57-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Challenge Problem

I have long believed that any algorithm developed in a procedural
language could be reformulated more elegantly in a logic language.
This may be true, but I have come to doubt its utility.  Restating
an algorithm in nonprocedural form can be a useful exercise, but
does not improve on the procedural nature of the algorithm.
Advocates of logic programming would have a weak case if their
languages were just notations for telling the computer what to do,
step by step.

In order to sharpen my understanding of logic programming, I would
be interested in hearing about logic-based solutions to a problem
that I consider particularly "algorithmic": extraction of connected
components in images.  I would like to learn of any logic-based
approach to the problem that would entail neither massive search nor
such extensive procedural embedding as to make the "logic" portion
trivial.

The problem:  Assume that you have an array of integers, and are
to identify all maximal connected components -- i.e., all clusters
of identical integers.  The integers range from 0 to some known
maximum, with 0 being a special code that connects with the space
"outside" the array.  You are free to use any bounded intermediate
data structures.  The output is to consist of an array in which
elements of each connected component are marked with the same
integer code, and elements of different components are marked with
different codes.  Other outputs (boundary traces, lists of region
adjacencies and inclusion relationships, number of holes per
region, etc.) are desirable but of secondary importance.

I know of several good algorithms for solving this problem.  Most
involve an intermediate map or equivalent data structure that records
nonmaximal connected components found during a row-by-row scan,
together with a list of discovered equivalences that can be used to
link these into maximal components.  Such algorithms are quite
efficient (aside from having to scan the array completly, in the
worse case, before starting to build the output map).  Others involve
visiting (and marking) every point in the array, tracing region
boundaries until they close and then raster scanning to find the
next unvisited point.  (This is a reasonable procedure if the
boundary traces are of primary interest.)

Is there any hope that a logic-based language could compile a
[nonalgorithmic] declarative statement of this problem into similar
code, or would the compiler have to be guided by domain knowledge that
essentially supplies it with the right answer?  If the latter
is true, in what sense is logic programming the answer to our
current needs?

-- Ken Laws

------------------------------

Date: Wed, 14 Aug 85 10:23:10 cdt
From: Kenneth Forbus Forbus%uiucdcsp@Uiuc
Subject: Prolog will fail as the foundation for AI

(Carl, I wish you had found a fresher topic to re-open the
list with. And doing it right before conference season...)

To equate "prolog" with "logic programing, as Carl & Wayne's
messages comes very close to doing, is somewhat less than
accurate.  True, prolog is currently the logic programming
language of choice.  True, prolog provides a handy notion
of variable that takes painful effort to build in Lisp.
But prolog really is "Micro-Planner done right", as its
advocates claim, and that is enough to allow most AI people
to tune out prolog because we found out a long time ago that
Micro-Planner isn't what we want.

Prolog suffers from being both too old and too new.  It is
too old, in the sense that it has chronological backtracking
built into its roots and the standard versions don't take
into account new ideas (such as TMS').  Yes, of course one
can encode these things in logic and "run them" in Prolog
(If you really believe prolog is "programming in logic" I
have a bridge to sell you).  However, you may not get an
answer in the lifetime of the universe, and doing three or
four experiments will have you sitting fruitlessly at the
console long enough to write a fast lisp program to do the
same thing!  But prolog is too new, in that it has also not
kept pace with progress in programming environments and
language design.  Talking to prolog programmers is a bit
like talking to lisp programmers 10 years ago -- there are
many dialects, often changing rapidly, wild disagreements
on syntax, etc.  The days of CommonProlog are likely to be
pretty far off.

Judging the idea of logic programming by prolog is a bit like
judging lisp by Lisp 1.5.  I believe that only a tiny fraction
of the interesting ideas which should be encapsulated in a real
logic programming language have been developed.  If some venture
capitalist were to ask me into whose pocket should he put $1M to
get a signficant advance in the state of logic programming, I
wouldn't suggest spending it on the prolog community.  I'd
probably suggest splitting it between McAllester & de Kleer,
whose ideas on TMS', while rather different, are both far more
interesting and potentially productive ideas than, say, improving
the speed of unification or studying and-parallelism.  We simply
don't have enough experience writing and using reasoning languages
yet to either cast our ideas into stone (i.e., adopt prolog and
spend our time optimizing it) OR pass final judgement on the
merits of logic programming.

On logic and action:  Why shouldn't Shakey be considered a counter
example to Carl's claim about the impossibility of action in a
system based on logic?  Shakey used resolution theorem proving to
decide what to do and other kinds of routines (such as A* search)
to flesh out the details.  If the existence of these other routines
is considered sufficient to discount Shakey as a counterexample
then the claim seems rather uninterestingly obvious.

------------------------------

Date: Wed 14 Aug 85 12:07:05-EDT
From: Vijay <Vijay.Saraswat@CMU-CS-C.ARPA>
Subject: On logic programming as a foundation for AI.

This is a response to some of the issues recently raised
by Hewitt in a post to the PHIL-SCI digest. (Phil SCI
Digest, Aug 14).

For the sake of argument,  Prolog is probably a `more convenient'
language for writing compilers  than Lisp; on the other hand,
the current state of sophistication of Lisp implementations makes
them extremely attractive for developing rapid prototype
implementations.   For the near future I see the ideal programming
environment as being a Lisp Machine with a very fast Prolog on it
(If Symbolics Prolog lives up to rumours of 100K Lips that will
do, thank you.)  Why would anyone EVER want to write a CommonLisp
interpreter in  Prolog?  There are much better things to do ...
particularly when you already have CommonLisp and Prolog!

What does this have to do with "being a foundation for AI"?
If he means that "X is a foundation for AI" if X is used by
most people for writing their AI programs, then that is
irrelevant.  I would say that "X is a foundation for AI"
if most work being carried out in AI can fit in the theoretical
framework of X. With such a working definition it is rather
doubtful if any one paradigm can be a foundation for AI.  But I
would submit that the concept of logic programming is a step
towards reaching such a foundational understanding.

"Logic as a programming language will fail as the foundations
for AI because..."

Let's be careful how we bandy terms around.  Logic is not a
programming language. Logic is logic, a formal system. The
first axiom of logic programming (the LOGIC axiom) is that
the definite clause subset of logic has an appealing
interpretation as a programming language, if the process of
SLD-refutation (which is complete for this subset) is taken
as the inference mechansim.  In this programming language,
the user has DON'T KNOW CHOICE, i.e. the power of specifying
existential searches.  This power is rather unnerving.  The
second axiom of logic programming (the CONTROL axiom) is that
it is possible to provide general control mechanisms which can
be exploited by a programmer for controlling the search. This
led to the rise of logic programming languages, which allow,
in general, the programming of incomplete searches, (hence the
handling of non-montonic inference, defaults, inheritance ...).

The most ancient of these is PROLOG, which essentially provided
the control structure of SEQUENTIALITY. The language PROLOG has
something to do with logic of course, but, because of its
incomplete proof procedure, its semantics is best given via
denotational means, rather than logical means.  This is the
first corollary of logic programming: a formal understanding of
logic programming languages has to appeal to traditional computer
science techniques for giving semantics to programming languages.
The surprising discovery is that because of the simplicity of the
underlying execution mechanism such semantics are surprisingly
simple: far simpler than semantics for ALGOL, LISP (has anyone
ever attempted a semantics for something like COMMONLISP?),
ACTORS...  This has very powerful connotations with respect to
semantically based programming environments, program
transformations, and meta-programming.

There have been more recent languages which have provided other
control features: PARLOG, GHC, Concurrent Prolog and CP[!,|,&],
to name just a few in the main stream of Horn logic programming,
have attempted to also provide don't care choice and parallelism.
To frame it in the current parlance, these languages essentially
provide support for object-oriented programming.  While a formal
denotational semantics for the last is still being developed
(the others do not yet have formal semantics), it has already
been demonstrated that it has a surprisingly simple partial
correctness semantics.

More important, these languages demonstrate that while
keeping the LOGIC component more or less identical, it is
possible to achieve a great variety of operational
behaviours by changing the CONTROL component.  Hence logic
programming is not a LANGUAGE, it is a PROGRAMMING PARADIGM
encompassing all these operational behaviours.

To sum up, logic programming langauges are not just any
programming languages, at some level, most programs written
in such languages have a declarative interpretation
compatible with a logic system, typically universally quantified
definite clause logic. Logic programming languages are not just
logic systems because their control component plays an important
part.  The art of designing logic programming languages is
concerned with maintaining a delicate balance between these two
divergent themes. Even with their control components, logic
programming languages can generally be given simple semantics,
which reflects their underlying conceptual simplicity.

As far as supporting most of the current AI paradigms is concerned,
it should be clear that Definite clause logic supports naturally
the notions of  goal-driven backward chaining.

It is my contention (as yet unsupported) that the basic style of
computation provided by Concurrent Prolog and CP[!,|,&] naturally
supports data-driven, forward-chaining inference and also knowledge
structuring a la semantic network based languages, again via the
primary mechanism of message-passing.  ("reserach in progress",
though there is an article by Shapiro andd Takeuchi on object
oriented programming in Concurrent Prolog).

Hence I would conclude that, to the extent that logic programming
languages support such  programming paradigms, and to the extent
that they themselves have secure theoretical foundations, one can
at least claim that logic programming offers a step towards a
unified understanding of the foundations of (programming paradigms
used in) AI.

Note that I am not saying anything at all about the psychological
plausibility of controlled inference as the essential "problem
solving capability" in  intelligent agents.

Let me now discuss three specific poinst raised by Hewitt:

"1.  Logical inference cannot be used to infer the decisions that
need to be taken in open systems because the decisions are not
determined by system inputs.

2.   Logic does not cope well withthe contradictory knowledge
bases inherent in open systems.  It leaves out counterarguments
and debate.

3.   Taking action does not fit within the logic paradigm."

Some of these contentions have been made by Hewitt in other
contexts and I still remain as mystified by them as I was then.

Let us keep in mind that we are talking of programming langauges,
albeit peculiar programming languages. How does a LISP function
'take action'? Presumably by doing a (setq foo 'take-action).
How does an OPS-5 program take action? Presumably by executing a
(make task-312 ↑task take-action) on the right hand side of a
production. So also  logic programming languages take action by
instantiating a variable to some value. If that variable was
actually implemented as a particular memory cell which is being
monitored by ahard-wired coke-dispenser, then it would 'take
action': deliver a coke bottle.

Presumably what is confusing Hewitt is that in Prolog bindings
can be done on backtracking.  But all that is in the programmer's
control! If the programmer intended that the action to be taken
is irrevocable, he would write his axioms in such a fashion that
the binding would not be backtracked over. That is a control
problem!

About contradictory data bases.  First let me make the obvious
statement that a certain level of understanding, every set of
definite clause has a model, indeed an infinity of models. Hence
there is no scope for representing contradictory knowledge
directly via axioms in a Horn logic program.  So how do logic
programming systems do it?  Well you have to PROGRAM IN your
handling of such data.  You have to write a program, just as you
would write a program in LISP or ATOLIA which does to this data
what you want to do with it.  The axioms you write (the program
you write) will be executed in the fashion determined by the logic
programming language you choose and by the control you provide:
usually this approximates some kind of inferencing. BUt the action
that gets taken when you execute your program depends upon your
program.  If you had written your program such that when you
encounter an inconsistency it would print out "The moon is made of
green cheese" and stop, then, if an inconsitency is encountered,
believe it or not, it will print out "The moon is made of green
cheese" and stop. If you wanted to program in counterargumentsd
and debate then go ahead and do that.  Logic programming does not
provide "counterarguments and debate" as a primtive concept.
The advantage that you get by writing your program in a logic
programming language is that you can reason about it, you can
(hopefully) prove that when the program encounters an inconsistency
it will print out "The moon is made of greencheese" and terminate.
And because of the simplicity of the programming paradigm that you
are using, your proofs would be relatively simple.  If you were
careful in writing your program and in choosing your logic
programming langauge, the answer that you would get would be a
logical consequence of YOUR AXIOMS (program) which of course COULD
HAVE DONE whatever it wanted to to the data.

Hence, while on the face of it Hewitt's statement (2) is true,
it is totally irrelevant.

This also takes care of the argument that "logical inference
cannot be used to  infer decisions that need to be taken in open
systems ..." because logical inference is being used to execute
YOUR PROGRAM, which determines how to make those decisions.  If
your program is actually an OPS-5 interpreter which acts on the
data that it gets by using a knowledge base of productions, then,
by heavens, YOUR PROGRAM is NOT doing logical inference.  Hence
even if Hewitt's contention is correct, it is totally irrelevant.

-- Vijay A. Saraswat

------------------------------

End of PROLOG Digest
********************

∂26-Aug-85  0845	EMMA@SU-CSLI.ARPA 	Mailing lists   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 26 Aug 85  08:45:48 PDT
Date: Mon 26 Aug 85 08:39:52-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Mailing lists
To: folks@SU-CSLI.ARPA
Tel:  497-3479

  The old area mailing lists (f1, f2, c1, p1, nlinterest...) except
Pinterest have been discontinued.  I should have the new project 
mailing lists working soon.

-Emma
-------

∂26-Aug-85  1033	ACUFF@SUMEX-AIM.ARPA 	Explorer
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 26 Aug 85  10:33:16 PDT
Date: Mon 26 Aug 85 10:30:50-PDT
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorer
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12138244931.61.ACUFF@SUMEX-AIM.ARPA>

   We are now the proud owners of a TI Explorer, the first a many.  It
is currently in the vestibule of Whelan A-1105.  The manuals for it
are located above the DLion in the cubicle next to the Explorer.
Power it up by pressing the round button on the processor unit on the
floor to the right of the display.  When you are done, and if it seems
unlikely that anyone else will be using the machine for a while, run
(si:shutdown), and then turn off the processor.  Leave the display
powered up.  It currently has only Chaos network capability, so you
can get to files on the Whelan Symbolics machines, but not to other
Stanford hosts.  This should be remedied by mid September.

   Please send questions, problems, or bugs to me.  My office is right
next to the machine, so this should be particularly easy.  Note that
this is a fairly new machine, so there are likely to be problems -->
save your work frequently.

	-- Rich
-------

∂26-Aug-85  1141	STUCKY@SU-CSLI.ARPA 	D-Lion Users  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 26 Aug 85  11:41:42 PDT
Date: Mon 26 Aug 85 11:35:02-PDT
From: Susan Stucky <Stucky@SU-CSLI.ARPA>
Subject: D-Lion Users
To: folks@SU-CSLI.ARPA, ckiparsky@XEROX.ARPA

Dear All, 

This is a reminder of the D-Lion Users meeting tomorrow (that's Tuesday,
the 27th) at 1:30 in the Dandelion room in the trailers at Ventura.  A
number of people have volunteered to talk about (and demonstrate, when
useful) some of the software, both the standard stuff and programs users
have developed.  Hans, for instance, has offered to demonstrate
TEDITKEYPLUS.  If you have anything you think is useful and that you
would be willing to demonstrate, please send me a message about it; the
information will make it easier to plan the meeting.  Thanks. 

-Susan
-------

∂26-Aug-85  1434	ullman@diablo 	surprise visit 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 26 Aug 85  14:34:31 PDT
Date: Mon, 26 Aug 85 14:28:20 pdt
From: Jeff Ullman <ullman@diablo>
Subject: surprise visit
To: nail@diablo

Tomasz Imielinski just told me he missed his plane and
wants to kill some time by coming over here.
He has done some work in an area that sounds similar to
what Jeff Naughton is doing: when are recursions "real"
and when are they replaceable by finite interations.
Anyone interested may meet in my office about 4:30 today
(Monday).

By the way, let's meet at 11AM for the regular nail meeting.
I'd like to focus on some open questions in rule evaluation.
				---Jeff

∂30-Aug-85  1301	PATASHNIK@SU-SUSHI.ARPA 	[karp%ucbernie@Berkeley (Richard Karp):]
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  13:01:18 PDT
Date: Wed 28 Aug 85 13:22:42-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: [karp%ucbernie@Berkeley (Richard Karp):]
To: aflb.local@SU-SUSHI.ARPA

Date: Wed, 28 Aug 85 10:01:58 PDT
From: karp%ucbernie@Berkeley (Richard Karp)
Message-Id: <8508281701.AA03682@ucbernie.ARPA>
To: ashok@SU-SUSHI.ARPA, theory%ernie@Berkeley


SEMINAR ANNOUNCEMENT
MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley   642-0143

Thursday, August 29  2 p.m. MSRI Lecture Hall
Zvi Galil  Columbia University and Tel-Aviv University
Recent Results on the Complexity of the Minimum Spanning Tree Problem

Thursday, August 29  4 p.m.  MSRI Lecture Hall
Structure in Complexity Theory Seminar
Juris Hartmanis   MSRI
U:NP  -  Problems With Unique Solutions
-------

∂30-Aug-85  1324	ullman@diablo 	meeting today  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  12:55:19 PDT
Date: Wed, 28 Aug 85 09:49:24 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting today
To: nail@diablo

We'll meet at 11AM in 301.  I have no real agenda.
What I hope is that people will contribute their thoughts
on open questions.  I have a few, but not enough to fill up
an hour.
				---Jeff

∂30-Aug-85  1327	chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--September 3rd   
Received: from UCB-VAX.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  13:27:46 PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA00676; Thu, 29 Aug 85 14:09:44 pdt
Received: by ucbcogsci.ARPA (5.5/4.48)
	id AA27201; Thu, 29 Aug 85 14:12:49 PDT
Date: Thu, 29 Aug 85 14:12:49 PDT
From: chertok%ucbcogsci@Berkeley (Paula Chertok)
Message-Id: <8508292112.AA27201@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar--September 3rd
Cc: chertok%ucbcogsci@Berkeley

                    BERKELEY COGNITIVE SCIENCE PROGRAM
                                Fall 1985
                  Cognitive Science Seminar -- IDS 237A

      TIME:                Tuesday, September 3, 11:00 - 12:30
      PLACE:               240 Bechtel Engineering Center
      (followed by)
      DISCUSSION:          12:30 - 1:00 in 200 Building T-4

      SPEAKER:          Leonard Talmy, UCB
      TITLE:            ``Force Dynamics in Language and Thought''


      A semantic category that  has  previously  been  neglected  in
      linguistic  research  is that of ``force dynamics''--how enti-
      ties interact with respect to force.   Included  here  is  the
      exertion  of force, resistance to such a force, the overcoming
      of such a resistance, blockage of  the  expression  of  force,
      removal of such blockage, and the like.

           Though scarcely recognized before, force dynamics figures
      significantly  in  language structure.  It is, first of all, a
      generalization over the traditional notion  of  ``causative'':
      it  places naturally within a single framework not only `caus-
      ing', but also `letting,' as well as a set of notions not nor-
      mally considered in the same context.

           Force dynamics, furthermore,  plays  a  structuring  role
      across a range of language levels.  First, it has direct gram-
      matical  representation.   In  English,  such   representation
      appears not only in subsets of conjunctions, prepositions, and
      other closed-class elements but, most significantly,  also  as
      the  semantic  category  that  the  modal system as a whole is
      dedicated to expressing.   Force  dynamic  patterns  are  also
      incorporated in open-class lexical items, and bring numbers of
      these together into systematic relationships.   Lexical  items
      involved in this way refer not only to physical force interac-
      tions but, by metaphoric extension, also to psychological  and
      social  interactions,  conceived  in  terms  of  psycho-social
      ``pressures.''  In addition, force dynamic principles  can  be
      seen to operate in discourse that is involved with persuasion.
      Such rhetorical interchange (including efforts to exhort, con-
      vince,  or  logically  demonstrate) involves the deployment of
      points to argue for and against conflicting positions.

           Force dynamics is a major conceptual  organizing  system,
      constituting one of four major ``imaging'' systems that I have
      developed which provide an integrated semantic  schematization
      of  a referent scene.  Cognitively, it corresponds to concepts
      within ``naive  physics''  as  well  as  to  ones  in  ``naive
      (social)  psychology,''  and  can  be  contrasted  with modern
      scientific concepts in these domains.

      --------------------------------------------------------------------

      UPCOMING TALKS

      Sept 10:    Amos Tversky, Psychology, Stanford
      Sept 17:    Alan Schoenfeld, Education, UCB
      Sept 24:    Peter Pirolli, Education, UCB
      Oct 1:      David Rumelhart, UCSD
      Oct 8:      Terry Winograd, Computer Science, Stanford
      Oct 15:     Ron Kaplan, Xerox PARC

∂30-Aug-85  1342	HANS@SU-CSLI.ARPA 	talk by Uwe Reyle    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  13:38:57 PDT
Date: Tue 27 Aug 85 23:45:40-PDT
From: Hans Uszkoreit <Hans@SU-CSLI.ARPA>
Subject: talk by Uwe Reyle
To: folks@SU-CSLI.ARPA

Uwe Reyle of the University of Stuttgart (Germany) will give a talk
on "Grammatical Functions, Discourse Referents, and Quantification"
on Wednesday, August 28, at 15:15pm.  The talk is part of the activities
of the Working Group on Syntax and Semantics and will take place in 
the Ventura Hall Seminar Room.  
-------

∂30-Aug-85  1348	EMMA@SU-CSLI.ARPA 	parking stickers
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  13:47:47 PDT
Date: Thu 29 Aug 85 09:45:18-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: parking stickers
To: folks@SU-CSLI.ARPA
Tel:  497-3479


Parking stickers have to be renewed by September 15th.  We are doing a
group order to save people some hassle.  If you are interested, please
fill out the form and bring it to either Jamie Marks or Ingrid
Deiwicks before next Thursday (Sept. 5).  Stickers should arrive in
the following week.

You are allowed to buy parking stickers, if you are affiliated with
the university (staff, faculty, visiting scholar, student...).

Please send all queries to Jamie@csli.
-------

∂30-Aug-85  1355	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--September 3rd    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  13:52:07 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Thu 29 Aug 85 14:13:39-PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA00676; Thu, 29 Aug 85 14:09:44 pdt
Received: by ucbcogsci.ARPA (5.5/4.48)
	id AA27201; Thu, 29 Aug 85 14:12:49 PDT
Date: Thu, 29 Aug 85 14:12:49 PDT
From: chertok%ucbcogsci@Berkeley (Paula Chertok)
Message-Id: <8508292112.AA27201@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar--September 3rd
Cc: chertok%ucbcogsci@Berkeley

                    BERKELEY COGNITIVE SCIENCE PROGRAM
                                Fall 1985
                  Cognitive Science Seminar -- IDS 237A

      TIME:                Tuesday, September 3, 11:00 - 12:30
      PLACE:               240 Bechtel Engineering Center
      (followed by)
      DISCUSSION:          12:30 - 1:00 in 200 Building T-4

      SPEAKER:          Leonard Talmy, UCB
      TITLE:            ``Force Dynamics in Language and Thought''


      A semantic category that  has  previously  been  neglected  in
      linguistic  research  is that of ``force dynamics''--how enti-
      ties interact with respect to force.   Included  here  is  the
      exertion  of force, resistance to such a force, the overcoming
      of such a resistance, blockage of  the  expression  of  force,
      removal of such blockage, and the like.

           Though scarcely recognized before, force dynamics figures
      significantly  in  language structure.  It is, first of all, a
      generalization over the traditional notion  of  ``causative'':
      it  places naturally within a single framework not only `caus-
      ing', but also `letting,' as well as a set of notions not nor-
      mally considered in the same context.

           Force dynamics, furthermore,  plays  a  structuring  role
      across a range of language levels.  First, it has direct gram-
      matical  representation.   In  English,  such   representation
      appears not only in subsets of conjunctions, prepositions, and
      other closed-class elements but, most significantly,  also  as
      the  semantic  category  that  the  modal system as a whole is
      dedicated to expressing.   Force  dynamic  patterns  are  also
      incorporated in open-class lexical items, and bring numbers of
      these together into systematic relationships.   Lexical  items
      involved in this way refer not only to physical force interac-
      tions but, by metaphoric extension, also to psychological  and
      social  interactions,  conceived  in  terms  of  psycho-social
      ``pressures.''  In addition, force dynamic principles  can  be
      seen to operate in discourse that is involved with persuasion.
      Such rhetorical interchange (including efforts to exhort, con-
      vince,  or  logically  demonstrate) involves the deployment of
      points to argue for and against conflicting positions.

           Force dynamics is a major conceptual  organizing  system,
      constituting one of four major ``imaging'' systems that I have
      developed which provide an integrated semantic  schematization
      of  a referent scene.  Cognitively, it corresponds to concepts
      within ``naive  physics''  as  well  as  to  ones  in  ``naive
      (social)  psychology,''  and  can  be  contrasted  with modern
      scientific concepts in these domains.

      --------------------------------------------------------------------

      UPCOMING TALKS

      Sept 10:    Amos Tversky, Psychology, Stanford
      Sept 17:    Alan Schoenfeld, Education, UCB
      Sept 24:    Peter Pirolli, Education, UCB
      Oct 1:      David Rumelhart, UCSD
      Oct 8:      Terry Winograd, Computer Science, Stanford
      Oct 15:     Ron Kaplan, Xerox PARC

∂30-Aug-85  1407	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	moving    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  12:57:43 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 27 Aug 85 21:45:03-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Tue 27 Aug 85 21:45:05-PDT
Received: by wisc-rsch.arpa; Tue, 27 Aug 85 23:32:06 cdt
Message-Id: <8508280423.AA16622@wisc-rsch.arpa>
Received: from csnet-relay.arpa by wisc-rsch.arpa; Tue, 27 Aug 85 23:23:29 cdt
Received: from ibm-sj by csnet-relay.csnet id ad17724; 28 Aug 85 0:18 EDT
Date: Tue, 27 Aug 85 20:50:35 PDT
From: Peter van Emde Boas <pveb%ibm-sj.csnet@csnet-relay.arpa>
To: udi@wisc-rsch.ARPA
Subject: moving
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

Dear friends
We are heading back east to Holland. After August 30 we will no longer be
reachable at IBM San Jose.
Our Home address, effective Sep 19 1985 will be restored to be
 
Peter, Ghica, Donald, Caspar and Evert van Emde Boas
Beethovenlaan 44
2102 EW Heemstede
Netherlands
Phone: 011-31-23-281598
 
Peter's work address:
IIW/FVI, Univ. of Amsterdam
Roetersstraat 15
1018 WB  Amsterdam
Netherlands
Phone: 011-31-20-5223063 (secr) 5223065 (office)
Electronic address:  mcvax!uva!peter@SEISMO
 
Ghica's Work address:
IBM INS/DS
PO box 24
1420 AA Uithoorn
Netherlands
Phone: 011-31-2975-53528
VNET address:  EMDEBOAS AT UITVM1
 
Don't hesitate to continue sending electronic mail. I consider
it most helpful.
We enjoyed our stay in the US for the past eight months and all the
hospitality experienced. If you ever get to Holland please use the
above addresses and phone numbers for contacting us.
Greetings  Peter.


∂30-Aug-85  1409	EMMA@SU-CSLI.ARPA 	Re: parking stickers 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  14:08:01 PDT
Date: Thu 29 Aug 85 14:23:49-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Re: parking stickers
To: folks@SU-CSLI.ARPA
In-Reply-To: Message from "Michael Georgeff <georgeff@SRI-AI.ARPA>" of Thu 29 Aug 85 13:56:05-PDT
Tel:  497-3479


  Forms were issued to all employees of Stanford University.  If you 
did not get a parking sticker form, you can pick one up at the
Stanford Police Station.

ps. CSLI is not able to get any extra blank forms for the people
at SRI or Xerox who might need a parking sticker.

pps. Please address queries to Jamie; any queries addressed to me will
be passed along to him, but you will probably have to wait longer for
an answer.
-------

∂30-Aug-85  1427	EMMA@SU-CSLI.ARPA 	Newsletter August 29, No. 43   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  13:45:17 PDT
Date: Wed 28 Aug 85 17:10:02-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter August 29, No. 43
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
August 29, 1985                 Stanford                       Vol. 2, No. 43
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
           CSLI ACTIVITIES FOR *THIS* THURSDAY, August 29, 1985

   12 noon		TINLunch
     Ventura Hall       ``A Unified Indexical Analysis of ``same'' and
     Conference Room    ``different'': A Response to Stump and Carlson''
			by David Dowty
			Discussion led by Mats Rooth
		
   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week
     Conference Room	
			
   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←

         CSLI ACTIVITIES FOR *NEXT* THURSDAY, September 5, 1985

   12 noon		TINLunch
     Ventura Hall       ``Predication, Logical Syntax, and the Type-free 
     Conference Room    Conception of Properties, Relations, and Propositions''
			Discussion led by Chris Menzel
			(Abstract on page 2)

   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week
     Conference Room	
			
   3:30 p.m.		Tea
     Ventura Hall		

!
Page 2                     CSLI Newsletter                    August 29, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
       ``Predication, Logical Syntax, and the Type-free Conception
              of Properties, Relations, and Propositions''

      Since Frege, the (arguably) dominant conception of properties,
   relations, and propositions (PRPs) has been typed.  The first rigorous
   formal development of this view by Russell was motivated largely by
   his discovery of the justly famous paradox that bears his name, and
   was for many years considered the definitive solution to the problem.
   More recently, the view (albeit in extensionalist garb) has enjoyed a
   renewed respectability, and a renewed applicability, in linguistics
   and the philosophy of language prompted by Montague's assiduous
   investigations into the semantics of natural language.
      Recent years, however, have seen a growing dissatisfaction with the
   typed conception of PRPs.  I will begin this week's discussion by
   pointing out some of the problems lurking behind this dissatisfaction,
   and will then present the basics of a type-free alternative.  Despite
   the fact that Frege himself had a typed conception of PRPs, the
   type-free view I will present is largely Fregean in spirit: I will
   argue in particular that properties and relations are in some sense
   ``unsaturated.''  This insight, however, doesn't require typing, as
   Frege seemed to think, and I will show where, in my view, he went
   wrong.  I will then argue that the standard function/argument notation
   of first- and higher-order logic ought, as Frege intended, to be
   thought of as representing the ``completion'' of an unsaturated
   property or relation, and not, contra Bealer, the standing of two or
   more objects in a further relation of predication.  This suggests a
   number of important questions about the nature of logic that will be
   raised for discussion.				--Chris Menzel
                              ←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                        ``Swedish Ditransitives''
              Summary of the meeting on Thursday, August 22

      In most Swedish ditransitive sentences, either object may
   passivize.  This talk addressed the question of what special property
   of Swedish permits the second object to advance to subjecthood in the
   passive.  It was argued that the property in question is a disjunction
   between two syntactic functions, SUBJECT and TOPIC, on the
   sentence-initial position (the verb-second constraint).  Subjects can
   be distinguished from topics through various syntactic tests (e.g.,
   embedding under raising verbs), but this distinction is often opaque
   in simple clauses.  It was argued that this conflation of two
   alternative functions on a single constituent structure node has
   caused the topic to be reanalyzed as a subject.  Since topicalization
   does not alter grammatical relations, it applies freely to second
   objects in Swedish and English.  But only in Swedish can this
   topicalized second object be reanalyzed as a subject.  When the source
   (before topicalization) is an impersonal passive construction, the
   result is a passivized second object.
      Finally, a ``movement'' analysis, in which topicalization and
   passivization are two instances of a single process defined on
   c-structure (such as ``move alpha''), was rejected.  With verbs
   subcategorized for both a direct object and an oblique function, the
   prepositional object topicalizes but does not passivize, even in
   identical c-structures.  This suggests that the two processes belong
   to different sub-components of the syntax.		--Steve Wechsler
!
Page 3                     CSLI Newsletter                     August 29, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                            NEW CSLI REPORTS

     Report No. CSLI-85-29, ``Equations, Schemata and Situations: A
   framework for linguistic semantics'' by Jens Erik Fenstad,
   Per-Kristian Halvorsen, Tore Langholm, and Johan van Benthem, and
   Report No. CSLI-85-30, ``Institutions: Abstract Model Theory for
   Computer Science'' by J. A. Goguen and R. M. Burstall, have just been
   published.  These reports may be obtained by writing to David Brown,
   CSLI, Ventura Hall, Stanford, CA 94305 or Brown@SU-CSLI.
-------

∂30-Aug-85  1626	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.3, 30 Aug 85  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 30 Aug 85  16:25:56 PDT
Return-Path: <Neumann@SRI-CSL.ARPA>
Received: from SRI-CSL.ARPA by SRI-CSLA.ARPA with TCP; Fri 30 Aug 85 12:40:09-PDT
Received:  from SRI-CSLB by SRI-CSL via DDN;  30 Aug 85 12:35:20-PDT
Date: Fri 30 Aug 85 12:35:09-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.3, 30 Aug 85
To: RISKS: ;
ReSent-Date: Fri 30 Aug 85 16:25:13-PDT
ReSent-From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
ReSent-To: jmc-lists@SU-AI.ARPA

RISKS-FORUM Digest        Friday, 30 Aug 1985      Volume 1 : Issue 3

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator
          (Contributions to RISKS@SRI-CSL.ARPA)
          (Requests to RISKS-Request@SRI-CSL.ARPA)
          (This vol/no can be FTPed from SRI-CSL:<RISKS>RISKS-1.3)
          (Issue n of vol 1 is in SRI-CSL:<RISKS>RISKS-1.n)      

Contents:
  Miscellaneous comments on V1#2 (Dave Curry)
  Computer/hardship list (Jerome Rosenberg)
  Medical KBES --  Some AI systems may need FDA approval
  Health hazards of CRT use (Robin Cooper)

----------------------------------------------------------------------

Date: Thu, 29 Aug 85 21:22:44 EST
From: davy@purdue-ecn.ARPA (Dave Curry)
To: risks@sri-csl.arpa
Subject: Miscellaneous comments on V1#2

Just some miscellaneous comments on some of the things in RISKS V1#2.
Hope this isn't too long.

1) Fishermen.  This sounds like a crock to me.  I wonder whether the
broken buoy or the fact that the storm was not predicted was the deciding
factor in the case.  Since the NWS/NOAA is providing a service, which
nobody is *required* to use, I can't understand how they can be sued for
not predicting a storm.  What would happen if they predicted a storm which
never showed up?  Could all the fishermen who stayed home sue for their
lost profits?  I can see it now....  "Cloudy Thursday, rain Friday -- use
this information at your own risk."

2) Union Carbide.  I always wonder in cases like this whether the plant is
actually having more accidents than usual, or if because of Bhopal we're
just hearing about it more because the press has a new victim to pick on.
The number of accidents at that plant is disgraceful.  Does anyone think
the government will shut it down?

3) Bob Carter's comments.  I think I agree with PGN on these... I would
prefer to see RISKS cover more or less anything related to computer
"hazards", rather than centering on one or two things.  There are plenty
of other lists which already take certain parts of this (e.g. SOFT-ENG for
"who's responsible" type stuff, ARMS-D for SDI).  I also like the SEN
quotes -- I don't personally read SEN, and even if some of the stuff is
dumb (computer kills scientist), overall I think the brief summary PGN
provided in V1 #1 gives a nice broad range of topics to discuss.

4) Medical programs.  I'm not sure I trust these fully yet.  I'd have no
qualms about my doctor using one to *suggest* things to him, but I would
draw the line at his accepting the program's diagnosis unless he could
verify on his own that it was correct.  For example, a heart specialist
interpreting a heart-diagnosis program's output would be good; a general
practicioner's taking it as gospel would not be good.  We need to make sure
the doctor is capable of knowing when the program is wrong.  (I saw a
comment about MYCIN once - "if you brought MYCIN a bicycle with a flat
tire, it would try like hell to find you an antibiotic.")

5) SDI.  I'm going to leave this for the experts.  I personally lean
towards Parnas's "side", but I don't know enough about it.  I do
like reading the comments on it though.  (BTW, for those of you who
haven't yet read Herb Lin's paper, it's excellent.)

Great list so far... keep it coming.  As a (possibly) new topic, did
anyone go to this AI show in San Diego (?) or wherever?  I saw a blurb on
it somewhere... how about a review of what the current toys are and what
risks they may take?  I remember seeing something about a program to
interpret the dials and gauges of a nuclear power plant....

--Dave Curry
davy@purdue-ecn

------------------------------

Date: Thu, 29 Aug 85 14:00:58 cdt
From: jerome@wisc-rsch.arpa (Rosenberg Jerome)
To: risks@sri-csla
Subject: Computer/hardship list

    Peter: One basis for a focussed discussion of risks would be to try to
establish a list of those computer systems whose failure would cause great
hardship --economic, political, social --to a significant number of our
citizens. For example, the failure of our computer-controlled electric
power grid or the failure of the Reserve's check clearance system.
              Your readers/participants could be asked to suggest the systems 
to be included on the list. Your forum could then discuss probabilies
of failure,costs of failures vs failure time, etc. etc..    
                          Jerry

------------------------------

Date: Friday, 30 Aug 1985 05:37:48-PDT
From: goun%cadlac.DEC@decwrl.ARPA  (Ave decus virginum!)
To: risks@SRI-CSL.ARPA
Subject: Medical KBES

            Some AI systems may need FDA approval

    Expert systems come within the FDA ambit to the extent that
    they supplement doctor's work, according to Richard Beutal, a
    Washington D.C. attorney specializing in the legal aspects of
    technology. 

    An expert system may be defined as a computer program that
    embodies the expertise of one or more human experts in some
    domain and applies this knowledge to provide inferences and
    guidance to a user. some of the earliest and most
    sophisticated systems were developed for medical diagnosis:
    MCYIN, EMCYIN, CADUCEUS AND ATTENDING. [There are several
    more in use in Japan. --mjt] 

    Beutal called attention to proposed FDA regulations that, if
    implemented, would require medical expert systems to obtain
    FDA pre-marketing approval. Given that FDA approval for what
    are class 3 devices could take up to 10 years and that
    reclassifying such devices can take almost as long, these FDA
    regulations would virtually cause investment to dry up. 
      {Government Computer News Aug 16, 1985}

------------------------------

Date: Thu, 29 Aug 85 10:35:20 cdt
From: cooper@wisc-ai.arpa (Robin Cooper)
To: risks@sri-csl
Subject: health hazards of CRT use

With respect to the introduction of the topic of the health hazards of
using video terminals, I would be particularly interested in seeing
discussion of risks to pregnant women and their unborn children. Both
Sweden and Canada have apparently introduced legislation which gives
pregnant women the right to change job assignments, whereas the
official US line seems to be that there is not sufficient risk to
warrant this. 

Robin Cooper

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂01-Sep-85  2356	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	IJPP: new journal on parallel programming    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 1 Sep 85  23:56:43 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Sun 1 Sep 85 23:52:18-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Sun 1 Sep 85 23:52:21-PDT
Received: by wisc-rsch.arpa; Mon, 2 Sep 85 01:35:54 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Mon, 2 Sep 85 01:29:27 cdt
Message-Id: <8509020629.AA23525@wisc-crys.arpa>
Date: Sun 1 Sep 85 23:29:07-PDT
From: Gary Lindstrom <Lindstrom@UTAH-20.ARPA>
Subject: IJPP: new journal on parallel programming
To: potential-authors:;
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;


The following is an announcement of a new journal.  Contributions are now
actively being solicited, as explained below.  Please post this message
on local bulletin boards (hard and soft!).

Thanks,
	--Gary
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

		   INTERNATIONAL JOURNAL OF PARALLEL PROGRAMMING


			Editor-in-Chief:
				Gary Lindstrom
				Department of Computer Science
				University of Utah
				Salt Lake City, Utah 84112
				(801) 581-5586
				Lindstrom@Utah-20.ARPA

			Published by:
				Plenum Publishing Corporation
				233 Spring Street
				New York, NY  10013
				(212) 620-8000



Statement of Purpose:

	Parallel computer systems are characterized by the coexistence of
multiple coordinated activities.  Physically, such systems range from networked
independent computers to high performance single-chip processors.

	While parallelism has long been a physical reality in computer systems,
its impact on programming practice has in the past generally been felt only by
specialists in systems programming.  A new trend, however, is clearly emerging:
that of parallelism as a vital aspect of programming at all levels.

	The reasons for this trend are several, including the growing purview
of higher level languages, more elegant computing models relieved of sequential
computing ideology, and exciting new ideas in computer architecture motivated
by the advent of VLSI.

	The International Journal of Parallel Programming (IJPP) will offer a
window on these developments from a programmer's perspective.  Thus while no
pertinent aspect of parallel computing systems will be excluded a priori,
each contribution will observe the common theme of illuminating in some manner
the programming challenges presented by parallel computing systems.


Scope:

	Sample areas to be addressed in IJPP include (with a short suggestive
list of appropriate subareas mentioned in each case):

	* linguistic foundations (semantic formulations, abstract machine
		models)
	* conceptual frameworks for parallel computation (actors, guardians,
		synchronized tasks, functional and logic programming)
	* high-level languages for parallel programming (new formulations and
		extensions to existing languages)
	* evaluation methods (dataflow, reduction, eager vs. lazy evaluation)
	* implementation techniques (compilation, emulation, storage and
		name space management)
	* programming support systems (run-time libraries, debuggers,
		performance instruments)
	* pragmatic considerations (task granularity, load distribution and
		scheduling, error detection and recovery)
	* architectural characteristics (synchronous vs. asynchronous control,
		shared vs. partitioned memory, physical locality, reliability)
	* software engineering aspects (specification and design
		methodologies, rapid prototyping, migration of sequential code,
		verification and testing)
	* advances in parallel algorithms (prototypical problems, relationships
		to particular languages and evaluation methods)
	* performance studies (fundamental and empirical)
	* application studies (artificial intelligence, simulation, databases,
		numeric computation)

	In addition to full length refereed contributions, IJPP will offer a
department entitled "Nonpareil" (meaning, in French, both "nonparallel" and
"matchless").  Under this rubric will appear short contributions of a lighter
nature, including anecdotes, tongue-in-cheek commentary, and other entertaining
insights into the travails of parallel programming.


Editorial Board:

	A distinguished Editorial Board comprising approximately 20 experts in
the areas cited above is now being formed.


Commencement of the Journal:

	The first issue of IJPP will appear on or about July 1, 1986. Six
issues of approximately 90 pp. each will be issued per year.

	IJPP will be the successor to the International Journal of Computer and
Information Sciences, concluding fourteen volumes published under the
editorship of Julius T. Tou.  The first issue of IJPP will thus be designated
volume 15, number 1.


Call for Papers:

	Manuscripts for possible publication in IJPP should be sent to the
Editor-in-Chief.  Contributions must be in English, and previously unpublished.
Four single-sided copies are requested.

	To ensure eligibility for inclusion in the inaugural issue, manuscripts
should be received by October 15, 1985.

	Electronic mail will be used for correspondence with contributors
whenever feasible, and network addresses should be prominently indicated on
all submissions and inquiries.
-------
-------

∂02-Sep-85  1406	@SU-SUSHI.ARPA:PAPA@SU-SCORE.ARPA 	Ride for Berkeley   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 2 Sep 85  14:06:17 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Mon 2 Sep 85 14:01:59-PDT
Date: Mon 2 Sep 85 14:02:06-PDT
From: C. Papadimitriou <PAPA@SU-SCORE.ARPA>
Subject: Ride for Berkeley
To: aflb.local@SU-SCORE.ARPA
Message-ID: <12140118399.23.PAPA@SU-SCORE.ARPA>

Is anybody driving to Berkeley Tuesday before noon for the talks?
If so, I need a ride.
---Christos.
-------

∂02-Sep-85  2235	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.4, 02 Sep 85  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 2 Sep 85  22:35:39 PDT
Date: Mon 2 Sep 85 21:57:22-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.4, 02 Sep 85
To: RISKS: ;

RISKS-FORUM Digest        Monday, 2 Sept 1985      Volume 1 : Issue 4

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

          (Contributions to RISKS@SRI-CSL.ARPA)
          (Requests to RISKS-Request@SRI-CSL.ARPA)
          (Issue n of vol 1 is in SRI-CSL:<RISKS>RISKS-1.n)      

Contents:
  The Case of the Broken Buoy (Matt Bishop)
  Inaction; Buoys will be buoys; KAL 007; Malpractice (PGN)
  Health Hazards of CRT Use (Brint Cooper, Robin Cooper, PGN)
  Medical Software (Brint Cooper)
  Rolm's Hawk-32 (Doug Bryan)
----------------------------------------------------------------------

Date: 30 Aug 1985 1636-PDT (Friday)
From: Matt Bishop <mab@riacs.ARPA>
Organization: Research Institute for Advanced Computer Science
Address: Mail Stop 230-5, NASA Ames Research Center, Moffett Field, CA  94035
Phone: (415) 694-6363 [main office], (415) 694-6921 [my office]
Mythological-Animal: Unicorn
Pet-Peeve: Complaints about the number of header fields
Snack-Food: White-shelled Pistachio Nuts
To: risks@sri-csl.ARPA
Subject: The Case of the Broken Buoy

   Dave Curry's right. I remember reading a newspaper report which
said, in essence, that the NWS/NOAA lost because it had failed to
predict the storm.  I didn't believe it, so I read on, and the report
said that since they had known of a broken buoy, had failed to repair
it (I think it had been broken for several months), and therefore failed
to get the information needed to give a warning, they were guilty of
negligence and had to pay.  Quite a far cry from what the story had
begun as!

------------------------------

Date: Mon 2 Sep 85 14:05:15-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Inaction; Buoys will be buoys; KAL 007; Malpractice
To: RISKS@SRI-CSLA.ARPA

The issue of the lobstermen indeed rested on the negligence of not repairing
the buoy.  (As noted in RISKS-1.2, the weather buoy went unrepaired for
three months.)  

Negligence and inaction in the presence of informed knowledge are likely to
be the source of more lawsuits in the future.  For example, the NY Times of
1 September 85 had an article by Richard Witkin on KAL 007.

  Evidence introduced in lawsuits filed in connection with the Soviet downing
  of the Korean Air Lines Flight 007 suggests that American radar operators
  knew hours beforehand that the jetliner was off course and heading into
  Soviet airspace.

  The words, "We should warn him", presumably referring to the plane's pilot,
  were heard at the Government's civil air-traffic control station in Alaska
  as the Boeing 747 strayed off course toward its fatal encounter with a
  Soviet fighter plane two years ago today, according to the documents.
  
  The documents were submitted Friday as evidence in damage suits filed
  against the United States Government by relatives of the 269 people who
  died in the incident.

Medical malpractice suits have been on the upswing, and doctors are taking
extraordinary measures to compensate -- such as higher prices and otherwise
unnecessary tests and drugs.  But the question of what constitutes
computer-related malpractice is likely to emerge as a very sticky one, e.g.,
faulty computer system design, life-critical application programming, and
sloppy computer operation.  And what about a debugger or maintainer who
notices something fishy but does not carry through?  A remarkable case of a
casual observer playing a significant role took place on 1 Sept 85 when a
passenger on People Express Flight 183 from Dulles to Newark noticed minutes
after take-off that a cowling was missing on one of the engines.  (The plane
returned to Dulles.)  Imagine a lawsuit against a company, which in turn
sues the programmer.  The potential for legal confusion relating to computer
systems is really quite awesome, and the confusion has just begun.  Suppose
the windshear-warning system is finally installed (with the 31 May 84
near-disaster on take-off of a UA 727 and the recent crash providing an
impetus), and suppose that program has a bug?  Suppose the computer is not
working on landing?  There are some very serious questions that must be
raised.  The incidence of high-award law suits elsewhere is likely to
provide a strong forcing function.

------------------------------

Date:     Fri, 30 Aug 85 21:56:09 EDT
From:     Brint Cooper <abc@BRL.ARPA>
To:       cooper@WISC-AI.ARPA
cc:       risks@sri-csl.ARPA
Subject:  Re:  health hazards of CRT use

To balance this discussion, we need to include risks to pregnant women
and their born and unborn children of television sets that run 18 hours
a day in the home.  

Keep in mind:  X-radiation is generally produced by the very high
voltages traditionally used in color television sets and composite-video
color monitors.  Many of the monochrome monitors need no such voltages
and, so, produce no such radiation.  

Since most folks are now buying color TVs for their homes, we need to
examine that aspect of safety as well, especially since many of them are
used as monitors for home computers and video games.

Brint Cooper

------------------------------

Date: Sun, 1 Sep 85 12:13:49 cdt
From: cooper@wisc-ai.arpa (Robin Cooper)
To: abc@BRL.ARPA
Cc: risks@sri-csl.ARPA
Subject: Re:  health hazards of CRT use

Yes, that seems right, though I wonder what the facts are concerning
how close one sits to the device. People spend more time a few feet
away from their terminals than their TVs.

Robin Cooper

------------------------------

Date: Mon 2 Sep 85 21:10:33-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Re:  health hazards of CRT use
To: RISKS@SRI-CSLA.ARPA

There is also discussion in the literature on physical and psychological
problems resulting from sitting in front of your terminal for hours, most
notably back and neck problems, tension, stress, anxiety, and subsequent
depression.  This forum is not really the place to discuss another relevant
aspect of the problem, but let me just mention it anyway and then discourage
further commentary on it:  the standard American junk-food diet of coffee,
colas, and caffeine generally, orange juice, sugar, chocolate (containing
both sugar and caffeine), refined white flour, fried foods, and so on, is
now being linked with making many of those problems worse.

------------------------------

Date:     Fri, 30 Aug 85 22:00:55 EDT
From:     Brint Cooper <abc@BRL.ARPA>
To:       davy@EE.ARPA
cc:       risks@SRI-CSL.ARPA
Subject:  Medical Software

Actually, culpability for mistakes caused by medical diagnosis software
could be placed with the same person who is responsible for correct
interpretation of all diagnosis aids:  the physician him/herself.
Programmers, like authors of medical texts, are providing tools for the
physician, not replacing him or her.

What we CAN do as computer scientists, et al., is to educate the
medical profession to the limitations of these tools as well as to their
benefits.  For ourselves, the goals should include error and risk
reduction as we continue to discuss.

Brint

------------------------------

Date: Sat 31 Aug 85 22:58:00-PDT
From: Doug Bryan <BRYAN@SU-SIERRA.ARPA>
Subject: Rolm's Hawk-32
To: risks@SRI-CSL.ARPA

Speaking of possible hazards due to hardware failure, has anyone out there
had any experience with Rolm's 32 bit Mil Spec machine the Hawk-32?  Since 
the Hawk is a Mil Spec machine, I'm sure it will be used in situations where
failure could lead loss of life.

I would be interested in hearing about the Hawk's environment limitations,
mean time between failures and any other experiences people have had with
the machine.

doug

    [POSTSCRIPT: A few of you complained that the first issue had too much
     of a military flavor.  It is interesting that except for this last
     item, this issue and the previous issue had almost none!  On the
     other hand, the problems we are dealing with are universal, and
     we should be able to learn from all relevant discussions...  

     I had some complaints about the format breaking your dedigestifying
     programs.  I hope this is better, but if it really is, your programs 
     must be pretty stupid.  I did not change anything except the trailer.
     So maybe I don't have it right yet?

     Others complained that the issues were too big and did not come out
     often enough.  (I explained why -- I wasn't around.)  Now you will
     undoubtably complain that that they are too small and too frequent.
     But it really depends on what contributions are available.  PGN]

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂03-Sep-85  1004	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	call for papers - MIT conference on VLSI
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 3 Sep 85  10:03:49 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 3 Sep 85 10:01:32-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Tue 3 Sep 85 10:01:36-PDT
Received: by wisc-rsch.arpa; Tue, 3 Sep 85 11:37:33 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Tue, 3 Sep 85 11:06:22 cdt
Received: from MIT-MC.ARPA by wisc-crys.arpa; Tue, 3 Sep 85 11:06:15 cdt
Date: Tue,  3 Sep 85 12:04:51 EDT
From: E.Leiserson@wisc-rsch.arpa
To: theory@WISC-CRYS.ARPA
Subject: call for papers - MIT conference on VLSI
Message-Id: <[MIT-MC.ARPA].631653.850903.CEL>
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;


                               CALL FOR PAPERS

                           Fourth MIT Conference on

                          ADVANCED RESEARCH IN VLSI

                               April 7-9, 1986

                    Massachusetts Institute of Technology
                           Cambridge, Massachusetts


    The field of VLSI (Very Large Scale Integration) is an interrelated
    set of technical disciplines including semiconductor devices,
    processes, and patterning, and also circuit design, design automation,
    systems architecture, and complexity theory.  Successful advanced
    research in any of these areas requires understanding of the other
    areas--system architects and designers must understand the limitations
    imposed by the physical and chemical processes, and conversely,
    advances in device design occur best when circuit and architectural
    requirements are known.  This conference seeks to promote interaction
    among reseach workers in all disciplines that relate to integrated
    circuits.

    Call for Contributed Papers

    Original research papers may be submitted in any of the following
    areas:

      - Innovative Electrical Circuits     - Artificial Intelligence Applications
      - Process and Device Modeling        - Layout and Routing
      - Hardware Design Languages          - Silicon Compilation
      - VLSI Systems                       - Testing and Fault Tolerance
      - Special-Purpose Architectures      - Highly Parallel Architectures
      - Analysis and Simulation            - Highly Parallel Problem Domains
      - Sensory Systems                    - VLSI Theory

    DRAFT PAPERS (twelve copies) should be sent before October 1, 1985 to
    Charles E. Leiserson, Program Chairman, at the Conference mailing
    address:

        Charles E. Leiserson, Program Chairman
        Room 39-321
        Massachusetts Institute of Technology
        Cambridge, MA 02139

    Drafts should be sufficiently detailed to permit a review by the
    program committee.  Notification of acceptances will be sent to
    authors by November 1, 1985.  Camera-ready copy of papers for the 
    Proceedings is due January 1, 1986.

    Invited Speakers

            The invited speakers are leaders in the theory and practice of
    parallel computation.  They include:

             Kenneth E. Batcher (Goodyear) MPP
             Randall Rettberg (BBN) Butterfly
             Danny Hillis (TMC) Connection Machine
             Jack Dennis (MIT) Dataflow
             Charles L. Seitz (Caltech) Mosaic
             Leslie G. Valiant (Harvard) Interconnection Networks
             Allan Gottlieb (NYU) Ultracomputer
             Richard M. Karp (Berkeley) Parallel Complexity
             David Elliot Shaw (Columbia) NON-VON

    Attendance Information

    Arrangements for the Conference will be similar to those for the MIT
    conferences held in 1980, 1982, and 1984.  The conference will be
    relatively small and informal.  Attendance is by invitation only.  For
    Information, write to Paul Church, Registrations, Room 39-321, MIT,
    Cambridge, MA 02139, or telephone (617) 253-8138.

    Organization of the Conference

    The conference is organized by the Microsystems Program
    Office, MIT.  The program committee consists of H. Fuchs (UNC), 
    Kim Kokkonen, (Intel), H. T. Kung (CMU), F. T. Leighton (MIT), 
    C. E. Leiserson, Chairman (MIT), R. J. Lipton (Princeton), 
    R. F. Lyon (Schlumberger), F. P. Preparata (Illinois), 
    C. D. Thompson (Berkeley), J. D. Ullman (Stanford), 
    and R.E. Zippel (MIT).

    The MIT local committee consists of J. Allen, D. Antoniadis, 
    L. Glasser, T. Knight, H. Lee, B. Lory, P. Penfield, Chairman, 
    J. Raffel, H. Smith, C. Sodini, C. Terman and J. Wyatt.


    -------


∂03-Sep-85  1530	ullman@diablo 	paper received 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 3 Sep 85  15:30:05 PDT
Date: Tue, 3 Sep 85 15:12:56 pdt
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

I have a copy of "Query Optimization in Logical Databases"
by Kifer and Lozinskii.
They talk about how to push selections and projections through
Horn clauses.
Anybody want to see it?

PS: When I offer to distribute papers like this, I'm really
talking to local Stanford people.
I know that people outside Stanford get NAIL mail, which is
fine with me.  But please treat messages like this as an
announcement that the paper exists, and contact the author,
not me, for a copy.  Thanks.
				---Jeff

∂03-Sep-85  1654	RICE@SUMEX-AIM.ARPA 	Wednesday's Meeting
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 3 Sep 85  16:54:07 PDT
Date: Tue 3 Sep 85 16:54:12-PDT
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Wednesday's Meeting
To: aap@SUMEX-AIM.ARPA
Message-ID: <12140411872.66.RICE@SUMEX-AIM.ARPA>

Sorry about the late message everyone.

I haven't been able to organise anything for tomorrow so unless anyone
else can get something together I hereby declare tomorrow's meeting
to be null and void.

On a brighter note there WILL be a meeting next week.  This will be
as follows :-

Wed Sept 11
9 A.M. (ghasp)

Dr. Shimon Cohen

from Schlumberger P.A. res. will describe language development
work that he is doing as part of the Faim project.

(many thanks to our leader for lining this one up)


Rice
-------

∂03-Sep-85  1815	BRAD@SU-CSLI.ARPA 	vacation   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 3 Sep 85  18:15:14 PDT
Date: Tue 3 Sep 85 18:06:01-PDT
From: Brad Horak <BRAD@SU-CSLI.ARPA>
Subject: vacation
To: folks@SU-CSLI.ARPA

I'll be on vacation from September 4th until September 16th.  Please be aware
that I won't be able to answer any net mail until I return.  I would
appreciate a cc of any message concerning problems that would normally be
sent to me - it beats coming home to an empty mailbox!

--Brad
-------

∂04-Sep-85  1038	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Sep 85  10:38:21 PDT
Received: from ames-vmsb.ARPA by SU-SCORE.ARPA with TCP; Wed 4 Sep 85 10:33:32-PDT
Date: 4 Sep 85 09:49:00 PDT
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA


               ASSOCIATION FOR COMPUTING MACHINERY
                San Francisco Golden Gate Chapter
               "SIGBIG" Special Interest Committee
                 For Large High Speed Computers

 Meetings on  the first Wednesday of each month at 7:30 PM.   Speakers 
 who  can give insights to various aspects of  SUPERCOMPUTING are 
 featured each month.

 Next meeting:
     Wednesday, September 4,1985,  7:30 PM
     Speaker:   Doug Vaughan/Alliant

                The Atrium building
                5776 Stoneridge Mall Road
                Pleasanton

   Directions:  Take the Foothill offramp South from Hwy 580
                (Half a mile West from the intersection of 
                580 and 680 in Pleasanton)
                Turn left at the second light onto DEODAR
                Turn right at the first stop sign onto Stoneridge Mall Rd
                second building on your right is the Atrium.
                The conference room is on the bottom floor next to 
                Bagels and Lox deli

 ---------------------------------------------------------------
 Tape-recordings  of  most of the previous  may  be obtained
 in exchange for a tape cassette or $5.00 by contacting: 
                Mary Fowler (415)261-4058
                Supercomputing  #192, BOX 2787
                Alameda, CA. 94501-0787

 For information contact Mary Fowler, Chairperson (415) 261-4058
                     or  Mike Austin, Publ. Chair (415) 423-8446

------

∂04-Sep-85  1346	chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept 10    
Received: from UCB-VAX.ARPA by SU-AI.ARPA with TCP; 4 Sep 85  13:46:41 PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA00494; Wed, 4 Sep 85 13:40:04 pdt
Received: by ucbcogsci.ARPA (5.5/4.48)
	id AA11537; Wed, 4 Sep 85 13:43:19 PDT
Date: Wed, 4 Sep 85 13:43:19 PDT
From: chertok%ucbcogsci@Berkeley (Paula Chertok)
Message-Id: <8509042043.AA11537@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar--Sept 10

                          BERKELEY COGNITIVE SCIENCE PROGRAM
                                      Fall 1985
                        Cognitive Science Seminar -- IDS 237A

            TIME:                Tuesday, September 10, 11:00 - 12:30
            PLACE:               240 Bechtel Engineering Center
              (followed by)
            DISCUSSION:          12:30 - 1:30 in 200 Building T-4

            SPEAKER:          Amos  Tversky,  Department  of   Psychology,
                              Stanford University

            TITLE:            ``Misconception  of  Chance   Processes   in
                              Basketball''

            We investigate the origin and the validity of  common  beliefs
            regarding ``the hot hand'' and ``streak shooting'' in the game
            of basketball.  Basketball players  and  fans  alike  tend  to
            believe  that  a player's chance of hitting a shot are greater
            following a hit than following a miss on  the  previous  shot.
            However, detailed analyses of the shooting records of the Phi-
            ladelphia 76ers provided no evidence for a  positive  correla-
            tion  between the outcomes of successive shots.  The same con-
            clusions emerged from free-throw records of  the  Boston  Cel-
            tics,  and  from a controlled shooting experiment with the men
            and women of Cornell's varsity teams.  The outcomes of  previ-
            ous  shots  influenced  Cornell  players'  predictions but not
            their preformance.   The  belief  in  the  hot  hand  and  the
            ``detection''  of streaks in random sequences is attributed to
            a general misconception of  chance  according  to  which  even
            short random sequences are thought to be highly representative
            of their generating process.
            --------------------------------------------------------------
            UPCOMING TALKS

            Sept 17:    Alan Schoenfeld, Education, UCB
            Sept 24:    Peter Pirolli, Education, UCB
            Oct 1:      David Rumelhart, Cognitive Science, UCSD
            Oct 8:      Terry Winograd, Computer Science, Stanford
            Oct 15:     Ron Kaplan, Xerox PARC
            Oct 22:     Lotfi Zadeh, Computer Science, UCB
  

∂04-Sep-85  1354	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept 10
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Sep 85  13:53:39 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Wed 4 Sep 85 13:44:11-PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA00494; Wed, 4 Sep 85 13:40:04 pdt
Received: by ucbcogsci.ARPA (5.5/4.48)
	id AA11537; Wed, 4 Sep 85 13:43:19 PDT
Date: Wed, 4 Sep 85 13:43:19 PDT
From: chertok%ucbcogsci@Berkeley (Paula Chertok)
Message-Id: <8509042043.AA11537@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar--Sept 10

                          BERKELEY COGNITIVE SCIENCE PROGRAM
                                      Fall 1985
                        Cognitive Science Seminar -- IDS 237A

            TIME:                Tuesday, September 10, 11:00 - 12:30
            PLACE:               240 Bechtel Engineering Center
              (followed by)
            DISCUSSION:          12:30 - 1:30 in 200 Building T-4

            SPEAKER:          Amos  Tversky,  Department  of   Psychology,
                              Stanford University

            TITLE:            ``Misconception  of  Chance   Processes   in
                              Basketball''

            We investigate the origin and the validity of  common  beliefs
            regarding ``the hot hand'' and ``streak shooting'' in the game
            of basketball.  Basketball players  and  fans  alike  tend  to
            believe  that  a player's chance of hitting a shot are greater
            following a hit than following a miss on  the  previous  shot.
            However, detailed analyses of the shooting records of the Phi-
            ladelphia 76ers provided no evidence for a  positive  correla-
            tion  between the outcomes of successive shots.  The same con-
            clusions emerged from free-throw records of  the  Boston  Cel-
            tics,  and  from a controlled shooting experiment with the men
            and women of Cornell's varsity teams.  The outcomes of  previ-
            ous  shots  influenced  Cornell  players'  predictions but not
            their preformance.   The  belief  in  the  hot  hand  and  the
            ``detection''  of streaks in random sequences is attributed to
            a general misconception of  chance  according  to  which  even
            short random sequences are thought to be highly representative
            of their generating process.
            --------------------------------------------------------------
            UPCOMING TALKS

            Sept 17:    Alan Schoenfeld, Education, UCB
            Sept 24:    Peter Pirolli, Education, UCB
            Oct 1:      David Rumelhart, Cognitive Science, UCSD
            Oct 8:      Terry Winograd, Computer Science, Stanford
            Oct 15:     Ron Kaplan, Xerox PARC
            Oct 22:     Lotfi Zadeh, Computer Science, UCB
  

∂04-Sep-85  1408	STUCKY@SU-CSLI.ARPA 	Dlion-users   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Sep 85  14:08:06 PDT
Date: Wed 4 Sep 85 13:54:29-PDT
From: Susan Stucky <Stucky@SU-CSLI.ARPA>
Subject: Dlion-users
To: folks@SU-CSLI.ARPA, carol@SU-CSLI.ARPA, bboard@SU-CSLI.ARPA

Dear All, 

A week ago, a meeting for Dlion-users and support staff was held at
CSLI.  If you are interested, a report on that meeting can be found on
the Dandelion bboard (Heading: Memo).  Also, a mailing list,
dlion-users@CSLI, has been established.  If you would like to be on that
list, send a request (as usual) to Requests@CSLI.  All further
announcements of meetings will go to the users list only, so as to
minimize junk mail.

Cheers. 

-Susan
-------

∂04-Sep-85  1736	EMMA@SU-CSLI.ARPA 	Newsletter September 5, No. 44 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Sep 85  17:36:32 PDT
Date: Wed 4 Sep 85 17:17:57-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter September 5, No. 44
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
September 5, 1985               Stanford                       Vol. 2, No. 44
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
           CSLI ACTIVITIES FOR *THIS* THURSDAY, September 5, 1985

   12 noon		TINLunch
     Ventura Hall       ``Predication, Logical Syntax, and the Type-free 
     Conference Room    Conception of Properties, Relations, and Propositions''
			Discussion led by Chris Menzel

   2:15 p.m.		CSLI Talk
     Ventura Hall	``FORK: A Flavor-Based Environment for Object-oriented
     Seminar Room	Knowledge Representation''
			C. Beckstein, G. Goerz, 
			University Erlangen-Nuernberg, W. Germany 
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←

         CSLI ACTIVITIES FOR *NEXT* THURSDAY, September 12, 1985

   12 noon		TINLunch
     Ventura Hall       ``Free Word Order in GPSG'' 
     Conference Room    by Arnold Zwicky
			Discussion led by Hans Uszkoreit, SRI and CSLI
			(Abstract on page 2)

   2:15 p.m.		CSLI Talk
     Ventura Hall	``Arithmetical Truth and Hidden Higher-Order Concepts''
     Seminar Room	Daniel Isaacson, Oxford University
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

!
Page 2                     CSLI Newsletter                  September 5, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                   ABSTRACT FOR THIS WEEK'S CSLI TALK
                 ``FORK: A Flavor-Based Environment for
               Object-oriented Knowledge Representation''
   C. Beckstein, G. Goerz, University Erlangen-Nuernberg, West Germany
            2:15, Thursday, September 5, Ventura Seminar Room

      Most object-oriented extensions of LISP provide only marginal
   support for the purpose of knowledge representation. In particular,
   there are only poor means---if any---for specifying meta-information
   about attributes of objects such as typed domains, methods for
   determining values (demons), multiple-valued attributes and explicit
   control of inheritance.  Furthermore, they usually don't offer
   adequate utilities for handling multiple perspectives, retrieving
   objects through patterns of characteristic features, and maintaining
   structural relations (integrity constraints) in and between objects.
   FORK is an attempt to extend Flavors, an object-oriented extension of
   LISP, by adding features which are well known from frame-like systems
   with the advantage of keeping a systematic distinction between classes
   and instances. The procedural knowledge is attached to classes either
   in the usual sense of methods as functions or in the form of (forward
   chaining) rule sets. In addition, FORK offers a programming
   environment to support users in the construction and maintenance of
   large, hybrid knowledge bases.
                              ←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
                       ``Free Word Order in GPSG''

      The ID/LP version of GPSG provides an elegant scheme for describing
   certain syntactic phenomena that are usually subsumed under the terms
   ``free word order'' or ``free constituent order.''  The questions
   addressed in this paper are (i) whether (and how) all the variants and
   degrees of ordering freedom can be described in the framework and (ii)
   whether universal generalizations can be expressed.  In this connection, 
   a universal version of a Pullum-type liberation rule is discussed.
                              ←←←←←←←←←←←←
                   ABSTRACT FOR NEXT WEEK'S CSLI TALK
         ``Arithmetical Truth and Hidden Higher-Order Concepts''
                   Daniel Isaacson, Oxford University
          2:15, Thursday, September 12, Ventura Conference Room

      The natural numbers can be characterized within a given domain by
   the second-order condition that they constitute a minimal collection
   closed under a given one to one mapping (succession) and containing an
   element (zero) not in the range of that mapping.  Peano Arithmetic is
   what can be expressed of this characterization by first-order
   axiomatization, and is in this sense a natural, conceptually intrinsic
   formal system.  On the other hand, Godel showed that the truths of
   arithmetic are not recursively enumerable, so that any true axiomatic
   formal system for arithmetic has proper first-order extensions.  It
   might seem in this way that no one first-order axiomatic system could
   be of intrinsic conceptual importance.  This talk will explore
   considerations that might resolve the tension between these two
   observations, in particular to suggest that possibly Peano Arithmetic
   can be considered as complete with respect to genuinely arithmetical
   truth, in the sense that perceiving the truth of first-order truths in
   the language of arithmetic that are not provable in Peano Arithmetic
   must be in terms of higher-order concepts which they code.
!
Page 3                     CSLI Newsletter                  September 5, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                  TALK
            ``Unification-Based Speech Parsing with a Chart''
     C. Beckstein, G. Goerz, Univ. Erlangen-Nuernberg, West Germany
            2:15, Tuesday, September 10, Ventura Seminar Room

      We describe GuLP, a chart parser to be used as a syntactic module
   of the Erlangen Speech Understanding System EVAR. Operating with a
   unification grammar, GuLP realizes an agenda-based multiprocessing
   scheme, which allows the application of various parsing strategies to
   fragments of the same utterance in a transparent way.  The overall
   control mechanism is realized through a general interrupt system.  In
   order to process speech data, a variety of new features has been
   incorporated: in particular the ability to perform incremental
   analysis, to do direction independent island parsing, to process gaps
   in utterances, and to handle hypothesis scores. Finally, a complexity
   estimate and a few experimental results are discussed.
                              ←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                    ``Discourse-based Interactions of
           Four Morphosyntactic Subsystems in Northern Pomo''
              Summary of the meeting on Thursday, August 29

      Formal theories of discourse representation face the task of
   modelling both text-external deixis (expressions anchored to the
   context of utterance) and text-internal deixis (e.g., Partee's (1984)
   treatment of Reichenbach's `reference time' or Banfield's (1982)
   discourse primitive SELF.)  Cathy O'Connor discussed these dimensions
   in the light of complex interactions of four morphosyntactic
   subsystems of Northern Pomo.
   (1) Third person, non-clause-bounded reflexives function
       logophorically; they establish that the SELF is anchored
       to 3rd person.
   (2) A possessive prefix found on kinship terms that is
       necessarily a bound anaphor is optional outside the
       minimal clause containing its antecedent.  If SELF is 3rd
       person, the anaphoric prefix is obligatory, deictically
       linked to this text-internal discourse entity.
   (3) An alternation in case-marking for subjects of
       unaccusative verbs conveys subjective expression of
       internal experience (`I'm feeling really sick') versus
       objective reporting (`I'm sick today'). This is normally
       limited to 1st person subjects. In discourse contexts
       where SELF is 3rd person, the alternation is sanctioned
       for 3rd person subjects.
   (4) A set of `evidential' verbal inflections, which
       indicate utterance-speaker's evidence for the assertion,
       display pragmatically motivated co-occurrence restrictions
       with respect to the above phenomena.
      In light of these findings the relation of logophoricity and
   subjunctive mood was discussed. The proposal was made that mood
   subordination in a discourse representation is the appropriate domain
   for an account of these complex facts.  Finally, the problem of
   representing (1) through (4) above was discussed, and the notions of
   text-internal and text-external deixis were suggested to be necessary
   components in a representation of the discourse structures
   involved.					--Cathy O'Connor
-------

∂04-Sep-85  2123	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.5, 04 Sep 85  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 4 Sep 85  21:23:44 PDT
Date: Wed 4 Sep 85 20:32:56-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.5, 04 Sep 85
To: RISKS: ;

RISKS-FORUM Digest       Wednesday, 4 Sep 1985      Volume 1 : Issue 5

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

          (Contributions to RISKS@SRI-CSL.ARPA)
          (Requests to RISKS-Request@SRI-CSL.ARPA)
          (FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

Contents:
  The Strategic Defense Initiative (Joseph Weizenbaum)
  1.5 million Ford engines need recall? (Hal Murray)
  Risks in CAD, etc. (Eugene Miya)
  crt & non-crt risks (Mike McLaughlin)
  Computerworld... on Union Carbide and NJ false arrests (Charlie Spitzer)
  More on false arrests (PGN)
----------------------------------------------------------------------

Date: Wed 4 Sep 85 14:19:15-EDT
From: Joseph Weizenbaum <JOSEPH@MIT-XX.ARPA>
Subject: The Strategic Defense Initiative
To: risks@SRI-CSL.ARPA

Greetings !

	I've just been introduced to the RISKS b-board you have undertaken
to maintain.  It is a good idea.  What stimulates this particular outburst
is John McCarthy's observation that many of the computer people who
maintain that Star Wars can't be made to work for technical reasons, e.g.,
those brought forth by Parnas, are people who have opposed many other
government initiatives.  I imagine he means among others the war on Viet
Nam, the MX missile, the ABM initiative of years ago, the administration's
policies vis-a-vis Nicaragua, Cuba and El Salvador, and so on and on.

I confess I've been on what from John's point of view must be characterized
as the wrong side on each of the above controversies.  But then John has
been on what I would have to see as the wrong side on all these questions.
Does that relieve me of the obligation to guard my intellectual honesty by
studying his actual arguments ?  If so, then the very possibility of dialog
between holders of opposing views becomes impossible.

It is, however, important for people who indeed have a point of view to make
their positions clear, at least not to hide them.  Their doing so ought not
to disqualify what they then have to say.  For myself, I find it even more
important to actually draw on my quasi pacifist position in arguing about
Star Wars and similar issues and to explicate the connections I make.  I do
believe (with Parnas and many others) that the software required simply
cannot be produced to the degree of confidence without which it would be a
meaningless exercise.  I don't want to rehash the various technical
arguments here, however.  Let me just rest on the well publicized statements
of CPSR and of Parnas.  I want to say in addition, however, that I would not
support the SDI initiative even if I thought it to be technically feasible.
In that, John is quite right.  I'm afraid that many of the computer people
who rest on the technical arguments alone leave the impression that these
alone constitute their objections to the program.  Perhaps that is the
position of many objectors in the computer community.  I think, however,
there are many who would join me in resisting the program even if it could
be shown to be technically feasible.  I think John is quite right in asking
that that be made explicit where it applies.

This is not the place to air political and ideological positions.  For
clarity's sake, I just want to add to the above that I believe it to be
necessary to the survival of us all that we come to some social, political,
cultural accommodation with the rest of the peoples of the world even when,
or especially when, they organize their societies differently than we do
ours.  SDI is in the tradition of the great technological fixes which
appear to their authors to relieve them of the responsibility to confront
underlying human problems.

Besides, SDI is a giant step to the further militarization of space and of
our society.  I oppose that as well.

Joseph Weizenbaum.

  [For those of you just returning from the London 8th International
   Conference on Software Engineering, we eagerly await reports on
   the panel session of Fred Brooks and Dave Parnas on the feasibility
   of SDI from the software engineering point of view alone.

   You will find a lengthy special report on "Star Wars" in the 
   IEEE Spectrum, September 1985.  My copy arrived today.

       SDI: The Grand Experiment
       Part 1 -- Mind-boggling complexity
       Part 2 -- Exotic weaponry
       Part 3 -- Debating the issues

   PGN]

------------------------------

Date: Tue, 3 Sep 85 12:18:36 PDT
From: Murray.pa@Xerox.ARPA (Hal Murray)
Subject: 1.5 Million Ford Engines Need Recall?
To: RISKS@SRI-CSL.ARPA

This morning, my radio said something about a consumer group wanting
Ford to recall 1.5 million engines. Nobody knows what's wrong, but they
are blaming it on the computer. (I didn't get the fine print. I wasn't
awake.) Anybody know if that's the real problem or just a convenient
scapegoat?

  [The 4 Sept 85 NY Times has a note on the recall of 454,000 
   Chevy/Pontiac compacts for corroding pollution control equipment, and
   105,000 VW and Audi cars for faulty brake hoses, but nothing on Ford.  I
   would love to follow up on the 1.5 million Fords, but haven't found 
   anything yet.  PGN]

------------------------------

From: eugene@AMES-NAS.ARPA (Eugene Miya) 
Date:  3 Sep 1985 0859-PDT (Tuesday)
To: risks@sri-csl.ARPA
Subject: Risks in CAD, etc.

Something, I have been wondering about, perhaps for future discussion might
concern liabilities of CAD products.  It seems more merchandise I purchase
is shoddy, and I am beginning to wonder what some of the consequences of
"making the metal a bit thinner to save.."  could be.  I realize we are
using CAD and simulation tools to make things more efficient, perhaps the
case of the over efficient engine which flamed out when it flew through rain
[as reposted in SEN, I believe] might be a case in point.  What were our
margins of safety in the over-engineering we did in the past?  Any studies yet?

Lastly, regarding mail formats:  I have run on a gamut of different mailers
[my current one, mh, is not bad], but I can sympathize with those having
problems.  It seems Peter's comment about programs was a bit harsh.  I used
to read netmail on an IBM machine which concatentated all letters and was
destructive [read once mail].

--eugene miya
  NASA Ames Research Center

------------------------------

Date: Wed, 4 Sep 85 18:22:22 edt
From: mikemcl@nrl-csr (Mike McLaughlin)
To: RISKS@SRI-CSL.ARPA
Subject: crt & non-crt risks

It is important we separate crt from non-crt risks.  X-rays, color-perception, 
& possibly eye fatigue I see as crt related.  Posture may be, for a person 
tied to the tube for extended periods.  Junk food is a non-crt risk, but may 
be a denial-of-service risk, if introduced into certain apertures around the 
crt.  Might also be a hazard to your health, if conductive. 

X- & like radiation:  I am done producing children, I hope.  So does my
wife.  Unless radiation reaches carcinogenic levels, I am not concerned for
me.  My children all use/will use crts, unless some other display becomes
more economical in the near future.  We have 5 children, all of age.  I am
concerned about them.

Posture:  As an occasional, voluntary, crt user, my posture is my problem. 
Take the paper out of my office; or give me a clerical/data entry type job; 
then I will see posture as a crt/computer risk.  Any obstetrician will worry 
about any woman who sits in any one position for long periods.  At one point 
in one pregnancy my wife had to fly home while I drove alone - solely so she 
would not have to sit still too long.  (Many years ago I was told that the 
blood supply to the brain passed through the peri-anal region.  This accounts 
for the number of dumb comments and sleeping attendees at various conferences 
with inadequate breaks.) 

Color-perception:  When I go home after dark tonight, the white line will be 
pink.  No, I'm not on anything.  If the screen were pink, the line would be 
green.  If color sensitivity mattered to me... say, if I performed color-
matching titrations in a hospital, or put color-coded resistors & capicators
into non-ICs, I would worry about color perception.  

	Considering the liability discussion in V1 #4, perhaps we all should. 
In 1956 or 1957 I ran across the proceedings of something-or-other on human 
factors in submarine design.  Book was pretty beat up, so it had been around
for a while.  It cited some *railroad safety* research on color perception. 
I think the RR stuff was pre WW-II.  Said red & green were neat colors for 
signal lights.  Also said *yellow symbols on a black background* were the 
best combination for a symbolic display... and that the reverse was the next 
best.  Hence, road signs.  Amber screens... ?  

Eye-fatigue:  Not crt-unique, but... look at anything long enough, your eyes 
will tire.  Look at anything slightly fuzzy, & your eyes will tire quickly, as
they try to focus on the un-focusable.  

Summary:  If a tired terminal operator hits a tree on the way home, it might
be due to poor color perception, fatigue due to poor posture (read:
furniture), eye fatigue due to poor colors, poor contrast, fuzzy images.  It
might be a financial disaster for the firm that employed said deceased.
Some attorney might look closely into the work situation, and computers
would get a bad name when we are really talking about bad management of the
computer-workplace.

	- Mike

------------------------------

Date:  Tue, 3 Sep 85 13:07 MST
From:  Charlie Spitzer <Spitzer%pco@CISL-SERVICE-MULTICS.ARPA>
Subject:  Computerworld Aug 19 articles
To:  Risks@SRI-CSL.ARPA

Readers may be interested in 2 articles from Aug 19 Computerworld.

page 6. Union Carbide modeling program given wrong data.

  Discusses wrong information about gases input to a program that was
  supposed to model gas cloud dispersal.  Notable quote: "These programs
  have been sold to safety people as opposed to engineers, because [the
  systems] provide good [public relations], are attractive visually and
  can provide a fairly inexpensive way of dealing with a problem you hope
  you'll never have."

page 12. On-line crime suspect system implicated in false arrest.

  Discusses case of a NJ woman arrested, strip searched and jailed on two
  separate occasions because of inadequate information stored in the NCIC
  computer.

charlie

------------------------------

Date: Tue 3 Sep 85 13:56:12-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: More on False Arrests
To: RISKS@SRI-CSLA.ARPA

  [For those of you who do NOT read the ACM SIGSOFT Software Engineering
  Notes, here are three items taken from my RISKS column in the July 1985
  issue on the subject of false arrests.  For those of you who have already
  read these items, you have come to the last message in this issue and need
  read no further.]

In the past quarter year, there were two different stories of people winding
up in jail due to computer-related snafus, and an earlier story that
serendipitously goes along with them.

1. The AnchoRages Of Sin

An article on page 17 of the 15 April 1985 issue of ComputerWorld  entitled

        ``DMV computer error puts innocent motorist in jail''

provides us with another software glitch with harmful side-effects.  (Brad
Davis [b-davis@@utah-cs.ARPA] was the first to point this one out to me.)

The article (by Kathleen Burton) details how a mistake in programming the
new Alaskan Department of Motor Vehicles computer system resulted in
a motorist spending a night in a Fairbanks jail.  The computer indicated 
(erroneously) that C. R. Griffin was driving with a suspended license.
The article also said that only by human intervention were 400 additional 
driver's licenses not erroneously suspended.  Apparently the database
kept records of all violations in the past @i[five] years, but was supposed
to search only over the last @i[two] years for motorists who should be
suspended.  A programmer was quoted as saying that ``the cost of correcting
the mistake [in the program] was insignificant.''

2. Shirley There Must Be a Way Out

And then, on 25 April 1985, the Associated Press ran a story about
congressional hearings on the FBI national crime computer.  Two incidents were
included.  The first involved an airline flight attendant who was falsely
arrested and detained because of incorrect information in the FBI's national
crime computer.  Sheila Jackson Stossier was arrested on 28 October 1983 at the
New Orleans airport, because a woman named Shirley Jackson was wanted by Texas
authorities.  She wound up in jail for the night and detained in Louisiana for
five days.  She now has an arrest record, and her married name Stossier is
listed in the computer as an alias.  Coincidentally, another Shirley (Jones)
was also wrongly arrested because another woman alias Shirley Jones was listed
in the computer -- despite the facts that they had different birthdays, were
six inches apart in height, and 70 pounds in weight.  ``Despite this, the
Sheriff's office refused to drop the charges.''  (To make matters worse, it was
later determined that the wanted Shirley was already in jail at the time!)

3. One in Five Warrant Records Were Wrong -- Poor Odds

David Burnham (NY Times News Service) reported the following 
related story on 12 Feb 1985.

  A Michigan man filed suit charging that he was wrongfully arrested five
  times in 14 months after an arrest warrant for a man using his name was
  placed in the national computer system of the FBI.  The man, Terry Dean
  Rogan, charged that four of the arrests occurred after Michigan police had
  made an unsuccessful effort to get the warrant changed.  Rogan contends and
  the police confirm that the man actually being sought was another person
  using Rogan's stolen identity and credit cards.  Rogan, who is 27 years old,
  is not wanted for any crime.  

[The rest of the last story (which goes on for another page) is in the July
issue of Software Engineering Notes.  It was also BBOARDed earlier, so I
did not think it should be recyled again!]

------------------------------

End of RISKS-FORUM Digest
************************

-------

∂05-Sep-85  1640	EMMA@SU-CSLI.ARPA 	New project mailing lists 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 5 Sep 85  16:40:49 PDT
Date: Thu 5 Sep 85 16:33:31-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: New project mailing lists
To: folks@SU-CSLI.ARPA
Tel:  497-3479

                  PROJECT MAILING LISTS (1985-1986)

The following is a list of the new projects with the name of the
project, the person who is in charge, and the mailing address.  If you
are in charge of a project, please check the correctness of the
mailing list by either visiting the appropriate file or by using
mmailbox.  Please send changes to requests@csli.  If you are NOT in
charge of a particular project and you wish to be added, please send
your requests to the appropriate project coordinators.

Linguistic Approaches to Computer Languages. Hans Uszkoreit (done)
	LACL=@PS:<CSLI-LISTS>LACL.DIS

Embedded Computation Group.  Brian Smith (3 sub groups)
	ECG=@PS:<CSLI-LISTS>ECG.DIS

 	sub 1: Research on Situated Automata.  Stan Rosenschein 
	       ROSA=@PS:<CSLI-LISTS>ROSA.DIS

 	sub 2: Semantically Rational Computer Languages. Curtis Abbott
	       SRCL=@PS:<CSLI-LISTS>SRCL.DIS

 	sub 3: Representation and Reasoning. Brian Smith (RRR ?)
	       (not named yet)

Discourse, Intention, and Action.  P. Cohen. done
	DIA=@PS:<CSLI-LISTS>DIA.DIS

Rational Agency. Michael Bratman (done)
	RATAG=@PS:<CSLI-LISTS>RATAG.DIS
	RATIONAL-AGENCY=RATAG

Head-driven Phrase Structure Grammar. I. Sag and T. Wasow (done)
	HPSG=@PS:<CSLI-LISTS>HPSG.DIS

Grammatical Theory and Discourse Structures. Joan Bresnan
	LFG=@PS:<CSLI-LISTS>LFG.DIS

Phonology and Phonetics. ?Kiparsky
	PINTEREST=@PS:<CSLI-LISTS>PINTEREST.DIS

Computational Models of Spoken Language. Withgott
	CMOSL=@PS:<CSLI-LISTS>CMOSL.DIS

Finite State Morphology.  Karttunen
	FSM=@PS:<CSLI-LISTS>FSM.DIS

Visual Communication. Pentland@sri-ai
	VISCOM=@PS:<CSLI-LISTS>VISCOM.DIS

A Lexical Initiative. Annie Zaenen  (done)
	LEXINIT=@PS:<CSLI-LISTS>LEXINIT.DIS

AFT lexical representation Theory. Julius Moravcsik
	AFT=@PS:<CSLI-LISTS>AFT.DIS

Foundations of Document Preparation.  David Levy.
	DOCPREP=@PS:<CSLI-LISTS>DOCPREP.DIS

Foundations of Grammar. Karttunen (done)
	FOG=@PS:<CSLI-LISTS>FOG.DIS

Situation Theory and Situation Semantics (STASS) Jon Barwise
	STASS

System Development Languages (SDLG) Terry Winograd
	SDLG

-------

∂06-Sep-85  0048	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.6, 06 Sep 85  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 6 Sep 85  00:48:39 PDT
Date: Fri 6 Sep 85 00:04:39-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.6, 06 Sep 85
To: RISKS: ;

RISKS-FORUM Digest       Friday, 6 Sep 1985      Volume 1 : Issue 6

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Joseph Weizenbaum's comments (Dave Parnas)
  Good Risks and Bad Risks (Dave Brandin, PGN)
  Hot rodding you AT (Dan Bower)
  Hazards of VDTs and CRTs (Al Friend)
  crt & non-crt risks (Brint Cooper)
  The Case of the Broken Buoy (Herb Lin, Matt Bishop)

          (Contributions to RISKS@SRI-CSL.ARPA)
          (Requests to RISKS-Request@SRI-CSL.ARPA)
          (FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Thu, 5 Sep 85 07:28:56 pdt
From: vax-populi!dparnas@nrl-css (Dave Parnas)
To: nrl-css!Neumann@SRI-CSLA.ARPA
Subject: Joseph Weizenbaum's comments <JOSEPH@MIT-XX.ARPA>: sdi]

	Although there is a great deal of truth and wisdom in Weizenbaum's 
message, I believe that he overlooks the reason that SDI would be
destabilizing and another step in the Arms race.  It is not because
of the stated goals of the program (Reagan's March 1983 speech) but because
those goals are not achievable.  There would be nothing wrong with rendering
ICBMs and other weapons obsolete.  On the contrary, everyone should want to 
see every country, city, and town protected by an impenetrable shield that
would free it from the fear of the indiscriminate horror that rained down on
Nagasaki and Hiroshima.  It is because the SDIO efforts will not lead to 
technology of that sort, that SDI is the things that Weizenbaum says it is.

	 I agree with Weizenbaum that we need to seek non-technological
solutions.  Technology is not likely to provide solutions in a situation 
where we oppose a power with equally sophisticated technology.  

	I believe that SDI is one issue where both disarmament and armament
supporters could agree.  Both sides seek peace through different mechanisms,
but neither will find their goals advanced by an untrustworthy "shield".


Dave

------------------------------

Date: Thu 5 Sep 85 11:40:30-PDT
From: Dave Brandin <BRANDIN@SRI-AI.ARPA>
Subject: Good Risks and Bad Risks
To: Neumann@SRI-CSL.ARPA

Peter: I love your material that's being generated and produced, but I note
that it seems to weigh overwhelmingly against the computer.  Aren't people
sending you any GOOD stuff?  Like with the aid of a computer, 27 lives were
saved, etc.?  Like using the new NEC fingerprint computer, they were able to
match the Stalker's finger-prints in 3 minutes, etc?  Maybe you need a Call
for Good News?  

Dave

------------------------------

Date: Thu 5 Sep 85 23:32:45-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Good Risks and Bad Risks
To: RISKS@SRI-CSLA.ARPA
Cc: BRANDIN@SRI-AI.ARPA

Today's SF Chronicle had a nice article on "Computer Holds Promise in
Diagnosing Heart Disease", in greatly reducing the number of false
negatives.  But even there are significant risks.  Suppose you or your
doctor trusts the computer program more because it indeed has fewer false
negatives, and now you produce a false negative.  We are back to the case of
the woman who killed her daughter and tried to kill herself and her son
because the computer program had falsely produced an "incurable"
diagnosis.  (See the July 85 issue of Software Engineering Notes.)

Well, in the first issue of RISKS I recall saying there has got to be more
to this forum than just pointing out negative things.  I noted hope from the
research community, although one of the agonizing things that we have
observed in the ACM Special Interest Group on Software Engineering (SIGSOFT)
is the enormous gap between the research community and what is actually
being done in practice.  For critical systems, the ordinary software
development techniques are simply not good enough.

Yes, we should of course point out successes.  For example, the Shuttle
project has had many -- along with its much more visible problems.  

Peter

------------------------------ 

Date: Wed, 4 Sep 85 14:41:38 EDT
From: Dan←Bower%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
To: RISKS@SRI-CSLA.ARPA
Subject: Hot rodding you AT

In a recent issue of PC Magazine, Peter Norton espoused the idea of
substituting a faster clock chip to enhance performance.  Now, according
to the folk on the Info-IBM PC digest, this may create problems.  An
off the shelf PC AT is composed of components guaranteed to work to
IBM spec, e.g. 6 Mhz.  If I increase the clock rate, then the whole
rest of the machine has to be up to snuff.  If not, a part dies and
I pay a nasty repair bill.

Now if I took Mr. Norton's word as gospel, swapped chips and set
my PC AT on fire, would he be liable?  How about the publisher?

------------------------------

Date: Thu, 5 Sep 85 15:23:05 edt
From: friend@nrl-csr (Al Friend, Space and Naval Warfare Systems Command)
To: risks@sri-csl
Subject: Hazards of VDTs and CRTs

When evaluating the risks associated with various forms of technology
it is sometimes useful to have in hand the available data.

The Food and Drug Administration published a study in 1981:

               An Evaluation of Radiation Emission
                              from
                     Video Display Terminals

                   HHS Publication FDA 81-8153

The ionizing, optical, RF and acoustic radiation from a number of
terminals was measured.  I will briefly quote some of the conclusions
of this study.

For ionizing radiation:

  3.5  DISCUSSION

  Sufficient research information is available to estimate a range of
  risks of injury from ionizing radiation exposure.  Delayed disease,
  such as heritable mutation or cancer, usually forms a basis for the
  estimation, expressed in terms of the instances of the effect per
  person per unit of radiation (rad,rem, or R).  The risk estimates
  form a basis for radiation protection guidelines.

  For a VDT operator, the radiation protection guideline for 
  individuals in the general population is appropriate.  The guideline
  -- 500 millirem per year -- is for man-made radiation exposures
  excluding medical use.  For both normal and Phase III operating
  conditions, the likely emission from a VDT is 0.1 mR per hour or less.
  Terminals capable of exceeding the 0.5 mR per hour regulatory limit
  receive special attention (see Section 3.2, above).  With assumptions
  of 6 hours of viewing per day, 5 days per week for 50 weeks per year,
  the annual radiation dose to an individual 2 inches from the front
  surface of a screen emitting 0.1 mR per hour would be 150 millirem.
  Note that 2 inches is an unrealistically short viewing distance; as
  one moves further away from the screen, the radiation exposure
  decreases correspondingly.
  
For RF radiation:

  4.5 DISCUSSION

  Research information on bioeffects for the frequency range 15kHz to
  125 kHz is lacking, so empirical estimates of injury are not possible.
  However, the radiation in this frequency region interacts only
  slightly with the human body, so that significant biological effects
  are unlikely.  At the present time, no standard or guideline has been
  adopted in the U.S. for grequencies below 10 MHz.

For ultrasound radiation:

  5.4 DISCUSSION

  When airborne ultrasound impinges on human skin less than 1 percent
  is absorbed, the remainder being reflected.  The ear, however, is an
  efficient coupler of acoustic energy from air into the human body.
  Therefore, investigations of the biological effects of ultrasound
  levels much higher than those found in the VDT survey have included
  temporary threshold shifts in hearing (6).  So-called subjective
  effects have also been associated with high levels of ultrasound
  exposure, and include fatigue, headache, tinnitis, instability, a
  "fullness" in the ear, and nausea.  One report (7) tentatively
  associates the subjective effects with audible high frequency
  components of sonic radiation.  The studies were performed in the
  exposure range 70-120 dB in an industrial setting, and at 150 dB.
  No long term effects or delayed injuries are known.

  No formal standard for ultrasound exposure presently exists in the
  U.S.  Among several voluntary guidelines available, the 
  recommendations of W.I. Acton of the United Kingdom were used to
  compare the VDT results, because they are the most conservative in
  this frequency range.  The highest acoustic measurement obtained
  from a VDT in this study was 68 dB, well below Acton's guideline
  of 75 dB, and well below the energies associated with biological
  effects.

For "ergonomic" factors:

  7. CONCLUSIONS AND RECOMMENDATIONS

  *****
  *****

  The word processing field has expanded much faster than has the
  understanding of its impact on people who use VDTs.  The impact
  may be felt in areas such as employee morale, compensation,
  work hours, and work conditions.  We suggest that work conditions
  be given serious conisideration as the primary cause of VDT-user
  complaints.  The problem is not simple, however.  An extensive
  review of stress factors in the word processing work area (10)
  identified five separate factors that contribute to fatigue:
  vision, posture, environment, task organization, and higher
  order items such as disease susceptibility.  As early as 1976,
  it was recognized that glare (room lighting reflecting from the
  VDT face plate), work position, ambient noise, and work duration
  (absence of breaks) could be the most important factors
  influencing the VDT worker's health (11).  The parallel
  between the 1976 and 1979 studies is sufficiently strong for
  us to suggest that efforts expended to reduce stress caused
  by these factors would also reduce the adverse impact on
  health.

The above quotes from the FDA document are some of its most important
conclusions.  References to additional work are provided in the
document.  There has been further work since the time of this report
(1981).  I do not have immediate access to these later references.  I
believe they tend to bear out the conclusions of this report.

From conversations with those closer to this field than I, I get the
impression that one of the major stress factors in commercial word
processing operations is the highly regimented work situation, and the
possibility of being fired, if the operator does not turn out a certain
minimum amount of mistake free work per hour.

------------------------------

Date:     Thu, 5 Sep 85 11:55:42 EDT
From:     Brint Cooper <abc@BRL.ARPA>
To:       mikemcl@nrl-csr.ARPA
cc:       RISKS@SRI-CSL.ARPA
Subject:  Re:  crt & non-crt risks

Many of the crt/workplace issues you raise are shared by another group
whose members are quite diverse in their use of crt terminals:
secretaries. 

I know this is not quite the correct forum, but workplace rules and
legislation designed to "protect" users of terminals from problems of
posture, vision, and stress should consider this forgotten group of
workers as well.  Their problems are nearly the same.

Brint

------------------------------

Date: Thu,  5 Sep 85 17:06:21 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  The Case of the Broken Buoy
To: mab@RIACS.ARPA
cc: LIN@MIT-MC.ARPA, risks@SRI-CSL.ARPA

In response to:

       Dave Curry's right. I remember reading a newspaper report which
    said, in essence, that the NWS/NOAA lost because it had failed to
    predict the storm.  I didn't believe it, so I read on, and the report
    said that since they had known of a broken buoy, had failed to repair
    it (I think it had been broken for several months), and therefore failed
    to get the information needed to give a warning, they were guilty of
    negligence and had to pay.  Quite a far cry from what the story had
    begun as!

On the other hand, the NWS also said that even if the buoy had been
alive at the time, they would not have predicted the storm.  This
isn't to defend sloppy journalism, just to point out that the
newspaper was in essence correct in this instance.

------------------------------

Date:  5 Sep 1985 2049-PDT (Thursday)
From: Matt Bishop <mab@riacs.ARPA>
Organization: Research Institute for Advanced Computer Science
Address: Mail Stop 230-5, NASA Ames Research Center, Moffett Field, CA  94035
Phone: (415) 694-6363 [main office], (415) 694-6921 [my office]
To: Herb Lin <LIN@MIT-MC.ARPA>
Cc: risks@SRI-CSL.ARPA
Subject: Re: The Case of the Broken Buoy

Did the NWS say that (i.e., even if the buoy had been alive at the time,
they could not have predicted the storm) in testimony, or after the verdict?
If after the verdict, no comment.  But if as testimony, Herb, the jury (or
judge) apparently didn't believe the NWS testimony.  If you believe the NWS
claim, the headline was correct, but it's unfair to say the court ruled that
way when it explicitly based its ruling on negligence.

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂08-Sep-85  0149	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.7, 08 Sep 85  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 8 Sep 85  01:49:26 PDT
Date: Sun 8 Sep 85 00:37:04-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.7, 08 Sep 85
To: RISKS: ;

RISKS-FORUM Digest       Sunday, 8 Sep 1985      Volume 1 : Issue 7

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  The risks of not using some technology (John McCarthy)
  More on SDI (Joseph Weizenbaum)
  SDI reliability (Martin Moore)
  Re: Hazards of VDTs and CRTs (Bernie Elspas)
  Viruses, Trojan horses, and worms (Fred Hapgood, PGN)
  Re: The Case of the Broken Buoy (Herb Lin, Matt Bishop) 
  Re: Hot rodding you AT (Keith F. Lynch)

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      
----------------------------------------------------------------------

Date: 07 Sep 85  1329 PDT
From: John McCarthy <JMC@SU-AI.ARPA>
Subject: The risks of not using some technology 
To:   risks@SRI-CSL.ARPA    

	The problem with a forum on the risks of technology is that
while the risks of not using some technology, e.g. computers, are
real, it takes imagination to think of them.  A further problem
with newspaper, magazine and TV discussion of technology is that
journalists and free-lance writers tend to run in intellectual
mobs.  This biases the discussion for everyone, especially when
the same journalists read each others writings and call it public
opinion.  Here are some illustrations.

1. Suppose some organization manages to delay interconnecting
police data systems on some specious civil liberty grounds.
Suppose some wanted murderer is stopped for a traffic offense
but not arrested, because he is wanted in a jurisdiction not
connected to the computer system used by the police officer.
He later kills several more people.  The non-use of computers
will not be considered as a cause, and no-one will sue the
police for not interconnecting the computers - nor will anyone
sue the ACLU.  The connection will not even be mentioned in
the news stories.

2. No relative of someone killed on U.S. 101 during the 10 years
the Sierra Club delayed making it a freeway sued the Sierra Club.

3. No non-smoker who dies of lung cancer in an area newly polluted by
wood smoke will sue the makers of "Split wood not atoms" bumper
stickers.

***

Based on past experience, I expect this question to be ignored, but here's
one for the risk-of-computers collectors.  Is a risk-of-computers
organization that successfully sues to delay a use of computers either
MORALLY or LEGALLY LIABLE if the delay causes someone's death?  Is there
any moral or legal requirement that such an organization prove that they
have formally investigated whether their lawsuit will result in killing
people?  As the above examples indicate, the present legal situation
and the present publicity situation are entirely unsymmetric.

***

	Here's another issue of the social responsibility of computer
professionals that has been ignored every time I have raised it.

The harm caused by tape-to-tape batch processing as opposed to on-line
systems.

From the earliest days of commercial computing people have complained
about seemingly uncorrectable errors in their bills.  The writers
don't know enough to connect this with the use of tape-to-tape
batch processing.  Under such a system when a customer complains,
the person who takes the complaint fills out a form.  A key puncher
punches the form on a card.  At the next file-update, this card
goes to tape, and a tape-to-tape operation makes the correction.
If there is any error in the form or in the key punching, the
correction is rejected, and the customer gets the wrong bill again.
On-line systems permit the person who takes the complaint to make
the correction immediately.  Any errors in making the correction
show up immediately, and the person can keep trying until he gets
it right or ask for help from a supervisor.  Not only is the customer
better off, but the complaint-taker has a less frustrating job.

My own experience with the difference occurred in 1979 when my
wallet was stolen, and I had to tell American Express and Visa.
American Express had an on-line system, and the person who took
the call was even able to give me a new card number on the spot.
The Visa complaint-taker had to look it up on a micro-fiche file
and call back, and still they got it wrong.  They gave me a new
account number without cancelling the old one.

Perhaps this issue is moot now, but I suspect there are still
many tape-to-tape systems or systems using modern equipment that
still emulate the old systems.  Shouldn't computer professionals
who pretend to social responsibility take an interest in an
area where their knowledge might actually be relevant?

Once upon a time, beginning perhaps in the middle nineteenth
century, scientific organizations were active in pressuring
government and business to use new technology capable of
reducing risk and promoting the general welfare.  I have in
mind the campaigns for safe water supplies and proper sewage
disposal.  Here's a new one that involves computer technology.

Theft can be reduced by introducing the notion of registered
property.  When you buy a television, say, you have the option
of buying a registered model, and the fact that it is registered
is stamped on it.  Whenever someone buys a piece of used registered
property he has the obligation of telephoning the registry to
check whether the property with that serial number has been reported
stolen and recording his ownership.  Repairmen are also obliged
to telephone either by voice or by keyboard.

Unfortunately, too many computer people imagine their social
responsibility to consist solely of imagining risks.

------------------------------

Date: Sat 7 Sep 85 16:30:11-EDT
From: Joseph Weizenbaum <JOSEPH@MIT-XX.ARPA>
Subject: More on SDI (reply to comments on RISKS-1.5 statement)
To: Neumann@SRI-CSL.ARPA

I've received a number of responses to the remark I made that I would not
support the SDI program even if I thought it could be made to work.  I have
the feeling that,  if I try to respond globally, a full blown debate
may ensue.  That I really don't want to conduct with the bboard as the
medium of expression.  Nevertheless, I feel obligated to say just a few
words in an attempt to clarify some ideas that have probably been 
misunderstood.

I said that my attitude derives from what I called a "quasi pacifist"
position.  One writer thought that pacifists are opposed to all forms of
self defense.  Actually pacifists are often the first to come to the
defense of justice being trampled.  But the form of their resistance to
wrongs is non-violent.  It ought also not to be confused with "passive"
resistance - Gandhi often pointed out, usually by his own example, that there
is nothing passive about non-violent resistance.  My use of the term "quasi
pacifist" also elicited comment:  Am or am I not a pacifist?  Let me say I
strive to become a pacifist, to grow up to be one.  One isn't an adult by
virtue of merely wishing or claiming to be one.  Just so with being a
pacifist.  I am still far from the goal.

People apparently believe that, were the SDI technically feasible there
could be no reasonable objections to its development and deployment.
Wouldn't it be comforting if every region, every city and village in
America had, so to speak, an invisible shield over it which guarded against
the invasion of hostile missiles, they ask.  Speaking entirely in practical
terms, I would remind them that every year tons (perhaps kilotons) of
marijuana are smuggled past the U.S. Coastguard and the custom service.
Now that technical progress allows the construction of nuclear "devices"
smaller than a moderately sized overnight bag, a determined enemy could
destroy American cities without "delivering" war heads by air mail at all ! 
If I were responsible for national security, I would worry if, a few days
before the President's traditional State of the Union message, usually
delivered to the assembled leadership of all three branches of our
government, some foreign embassy evacuated all its personel.  Perhaps a
nuclear device of moderate size had made its way to Washington and is about
to decapitate the government.  We can no more bring peace to this globe by
putting impenetrable domes over nations than we can halt the violence in
our cities by providing everyone with bulletproof clothing.  Human problems
transcend technical problems and their solutions.

But suppose we could solve the smuggled bomb problem.

I would still oppose SDI.

SDI is an attempt at a technological solution to problems which have
their roots in and are social, political, economic, cultural, in other
words, human problems.  It is an attempt to find solutions under the the
light provided by technology when in fact we know them to reside only in
the human spirit.  That is what guarantees the failure of SDI more surely
than its complexity or the impossibility of its being tested.

Beyond all that is the fact that we live in a world of finite resources.
The scarcest resource of all is human talent and creativity.  The military
already commands the time and energy of most American scientists and
engineers.  Money is another scarce resource on which the military has first
call. On the other hand, social services of all kinds
are being cut back.  Meanwhile  the country faces social problems
of horrendous dimensions:  There is massive, deep poverty in the land.
Adequate health care is beyond the reach of millions of citizens and
ruinously expensive for many more millions.  The schools are spewing out "a
rising tide of mediocrity" while a huge fraction of our youth is functionally
illiterate.  The conditions that brought on the riots in American cities,
for example in Watts, have never been attended to - they silently tick
away, time bombs waiting to go off.

When resources are limited they must be distributed on the basis of a
widely based consensus on priorities.  To silently consent to lowering
still further the priorities our society assigns to the people's health and
education in favor of spending the billions of dollars required above and
beyond the already huge military budget for only the first stages of SDI,
is, it seems to me, to condone the continuing impoverishment and
militarization of not only America, but of the whole world. Ever more
scientists and engineers will be occupied with military work.  Ever more
industrial workers of many different kinds will be enmeshed in the
militarized sectors of society by, for example, being required to have
military security clearances.  There is a danger that, in the process of the
growing militarization of society, a certain threshold, hard to define but
terribly real, will be crossed and that, once crossed, there will be no ready
road back to a civilian society.  

	Joseph Weizenbaum

------------------------------

Date: Fri, 06 Sep 85 14:54:52 CDT
From: mooremj@EGLIN-VAX
Subject: SDI reliability
To: risks@sri-csl

[Peter, I have also posted this to SOFT-ENG.  If you think the duplication
is reasonable, please include it in RISKS as well. -- mjm]

I've been thinking about the SDI system and how it will be implemented.  
Specifically, I've been looking at a system composed of N independent 
platforms, each of which performs its own detection, decision making, and
response.  Given this type of system, we can reach a few conclusions about
the reliability of the whole system, based on the reliability of a single
platform.  I've crunched a few numbers: nothing profound, just some basic
statistics.

Definitions:

1. A "false positive" is an attack response by a platform when such a response
   is not justified.
2. A "false negative" is failure of a platform to attack when an attack 
   response is justified.

Let's look at the false positive case first.  How likely is the system to
experience a false positive, based on the probability for each platform?

             N:    50        100        200        500       1000       2000
Pp:         +------------------------------------------------------------------
  1.000E-12 |  5.000E-11  1.000E-10  2.000E-10  5.000E-10  1.000E-09  2.000E-09
  1.000E-11 |  5.000E-10  1.000E-09  2.000E-09  5.000E-09  1.000E-08  2.000E-08
  1.000E-10 |  5.000E-09  1.000E-08  2.000E-08  5.000E-08  1.000E-07  2.000E-07
  1.000E-09 |  5.000E-08  1.000E-07  2.000E-07  5.000E-07  1.000E-06  2.000E-06
  1.000E-08 |  5.000E-07  1.000E-06  2.000E-06  5.000E-06  1.000E-05  2.000E-05
  1.000E-07 |  5.000E-06  1.000E-05  2.000E-05  5.000E-05  1.000E-04  2.000E-04
  1.000E-06 |  5.000E-05  1.000E-04  2.000E-04  4.999E-04  9.995E-04  1.998E-03
  1.000E-05 |  4.999E-04  9.995E-04  1.998E-03  4.988E-03  9.950E-03  1.980E-02
  1.000E-04 |  4.988E-03  9.951E-03  1.980E-02  4.877E-02  9.517E-02  1.813E-01
  1.000E-03 |  4.879E-02  9.521E-02  1.814E-01  3.936E-01  6.323E-01  8.648E-01

Pp is the probability that a given weapons platform will experience a false 
positive.  N is the number of platforms in the system.  The entries in the
table give the probability that a false positive will occur on at least one
platform (and one may be enough to start a war.)  For example, if there are
1000 platforms, and each one has a one-millionth (1.000E-6) probability of
experiencing a false positive, then the cumulative probability that some 
platform will do so is 9.995E-4, or .09995%.  Looking at the table, I'd say
the numbers in the lower right corner are rather disquieting, to say the least.

Now let's look at the false negative case.  The table is structured a little
differently here.  In the false positive case, a single failure is disastrous;
in the false negative case, it's not.  The probability of a false negative
should be many orders higher than that of a false positive, simply because the
protections against a false positive will actually enhance the chances of a
false negative.  This table deals with a 100-platform system (that being the
most my binomial coefficient routine can handle). 

   Pn:  .001    .01     .05     .1      .2      .3      .4      .5
N:  +---------------------------------------------------------------
 30 | 1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  1.0000
 35 | 1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  0.9991
 40 | 1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  0.9824
 45 | 1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  0.9991  0.8644
 50 | 1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  0.9832  0.5398
 55 | 1.0000  1.0000  1.0000  1.0000  1.0000  0.9995  0.8689  0.1841
 60 | 1.0000  1.0000  1.0000  1.0000  1.0000  0.9875  0.5433  0.0284
 65 | 1.0000  1.0000  1.0000  1.0000  0.9999  0.8839  0.1795  0.0018
 70 | 1.0000  1.0000  1.0000  1.0000  0.9939  0.5491  0.0248  0.0000
 75 | 1.0000  1.0000  1.0000  1.0000  0.9125  0.1631  0.0012  0.0000
 80 | 1.0000  1.0000  1.0000  0.9992  0.5595  0.0165  0.0000  0.0000
 85 | 1.0000  1.0000  1.0000  0.9601  0.1285  0.0004  0.0000  0.0000
 90 | 1.0000  1.0000  0.9885  0.5832  0.0057  0.0000  0.0000  0.0000
 95 | 1.0000  0.9995  0.6160  0.0576  0.0000  0.0000  0.0000  0.0000
100 | 0.9048  0.3660  0.0059  0.0000  0.0000  0.0000  0.0000  0.0000

Pn is the probability that a given platform will experience a false negative.
N is the minimum number of platforms (out of 100) which respond correctly.
The table entries give the probability that at least N platforms respond
correctly.  For example, if the probability of a given platform experiencing
a false negative is 0.1 (10%), then the probability is 99.92% that at least
80 out of 100 platforms respond correctly, 58.32% that at least 90 respond
correctly, and so on.

Some of the Pn's and Pp's may strike you as much too high.  I don't think so. 
The two tables were constructed on the simplifying assumption that Pn and Pp
are constants; actually, they are reliability functions.  The longer a platform
is in service, the more likely it is to malfunction.  If we assume that the
time-to-failure rate of a platform is some form of Weibull distribution
[a*B*t**(B-1) * e**(-a*t**B)], then the reliability function is given by Z(t)
= a*B*t**(B-1).  I did not use this in constructing the tables in order to
keep from drowning in figures, and because I don't really know how to choose
a, B, and unit t, until we get a history of actual performance (and by then it
may be too late...)  Suggestions are welcome. 

                                  Martin Moore (mooremj@eglin-vax.arpa)

------------------------------

Date: Fri 6 Sep 85 15:45:16-PDT
From: Bernie <ELSPAS@SRI-CSLA.ARPA>
Subject: Hazards of VDTs and CRTs 
To: RISKS@SRI-CSLA.ARPA
RE:  RISKS contribution from friend@nrl-csr (Al Friend, Space and 
     Naval Warfare Systems Command); RISKS-1.6

The 1981 FDA study cited by Friend probably contains much useful (albeit
rather "soothing") information about VDT *radiation* hazards (ionizing,
RF, and acoustic).  One should observe carefully, however, that the 
quoted material fails to mention other kinds of hazards, nor does its 
title reflect any others.  One should, therefore, not assume that 
*radiation* hazards are the whole story for VDTs.  I would have felt 
more relieved at the data presented had it included some other, more 
obvious, risk factors such as visual effects.

In particular, recent studies show that at least two visual effects may
be quite important as factors producing severe eye fatigue.  The first,
visual flicker (resulting from the screen refresh rate), is probably
well understood (from extensive psychovisual experimentation in
connection with TV viewing).  The higher screen refresh rates used on 
some computer graphic displays seem to minimize this problem.  However,
60 fields/sec (50, in Europe) is standard for most personal computers.

[Flicker depends on many factors: rate, ambient light, screen contrast, 
brightness, subject motion, color, etc.  More seems to be known about the 
conditions for minimal *perceptible* flicker than about those that can
produce visual fatigue, eyestrain, headaches, etc.  Also, there is a
fairly large variation among different subjects even for minimal
perceptible flicker, and flicker may be noticeable (and annoying) in 
the "fringe visual field" (off to the side) even when it is not detectable 
for the object directly ahead.]

The second factor is connected with the fact that the human eye is not 
chromatically corrected, i.e., its focal accommodation is different for
different colored objects.  The result is that when the eye is focused
correctly on a blue object, a nearby red object will be slightly out of
focus.  One study [1] indicates that the discrepancy is about 0.5
diopters (for a viewing distance of 50 cm).  According to one report
I've seen (sorry, I can't find the reference!), this means that in a
multicolored display the eye will automatically be making rapid
focus adjustments in scanning the screen.  Even worse, the effect can
also exist in some monochrome displays, i.e., where the character 
color (white, say) is achieved by a mixture of two differently colored 
phosphors separated substantially in wavelength.  In the latter situation 
it appears that (at least for some people) the eye may undergo extremely 
rapid focus oscillations in the futile attempt to bring both component 
colors into focus.   Quite understandably this may result in severe eye 
fatigue, even though the subject may not be consciously aware of what 
is happening.  This occurs mostly when the two phosphors radiate nearly 
pure spectral lines.  Single-phosphor displays and those where the 
component pure colors are close enough in wavelength seem not be prone 
to this disturbing effect.  I recall seeing the statement that AMBER 
displays are not objectionable for this reason, and that one nation
(Sweden or West Germany, I think) has specified amber displays for 
extended-time industrial use.

It seems to me that the chromatic refocusing effect is probably the more
serious of the two cited, especially on high resolution displays.  The
fact that it seems not to have been noticed on conventional (analog)
color TV displays may be accounted for by their relatively poor
resolution (low bandwidth).  Thus, the brain expects to see a sharper image
on a high-resolution (RGB) display than on a conventional TV (where
everything--especially the reds and oranges--is pretty blurred anyway).

In summary, in concentrating on the "serious" potential hazards of 
X-rays, etc., from VDTs, we should not thereby overlook the more obvious
factors concerned with the visual process itself.


1. G.M. Murch, "Good color graphics are a lot more than another pretty
   display," Industrial Research and Development, pp. 105-108 (November
   1983).

Bernie Elspas

[Material inside [...] may be deleted at editor's option.  Bernie]
                      [The editor decided to leave it in.  PGN]

------------------------------


Date: Fri 6 Sep 85 22:55:13-EDT
From: "Fred Hapgood" <SIDNEY.G.HAPGOOD%MIT-OZ@MIT-MC.ARPA>
Subject: Viruses, Trojan horses, and worms
To: RISKS@SRI-CSL.ARPA

	I would like to see a discussion by the members of this list
of the degree to which computer users, whether individuals or
organizations, are vulnerable to worms and Trojan Horses. These
terms, which first appeared in this list in #3, refer to programs
designed to inflict some form of unpleasantness on the user, up to
and including the destruction of the system. Typically they erase
all files in reach. I have read discussion, in Dvorak's column in
*Infoworld*, of the possibility that such programs might modify the
operating system as well such that when the unfortunate user tries
to restore the destroyed files from backup disks, those too would be
erased.  One can also imagine, vaguely, programs that are insidious
rather than calamitious, that introduce certain categories of error,
perhaps when strings of numbers are recognized. These might be able
to do even more damage over the long run.

	There are two issues with these programs. The first is what
they might do, once resident. The second is the nature of the
vector, to borrow a medical term. Worms can be introduced directly,
by 'crackers', or surreptiously, by hiding them inside a legitimate
program and waiting for an unsuspecting user to run that program on
his system, thus activating the 'Trojan Horse'. The article cited in
#3 had to do with a program camouflaged as a disk directory that was
circulated on the download BBSs. One could imagine a spy novel
devoted to the theme: perhaps it was the KGB, and not Ben Rosen, who
provided the money to launch Lotus. Inside every copy of 1-2-3 and
Symphony is a worm which, every time it is run, checks the system
clock to see if it was later than, say, October 1, 1985. On that
date the commercial and industrial memory of the United States dies.
The CIA suspects something is up, but they don't know what.
Unfortunately the director of the team working on the problem is a
KGB mole.  Fortunately there is this beautiful and brilliant female
computer genius ...

	Anyway, I have a specific question: can anyone imagine a
circumstance in which a program appended to a piece of text in a
system could get hold of the processor? It would appear not, which
is a good thing, because if such circumstances did exist, then it
would become possible to spread worms by pigyybacking them on a
telecommunicated piece of text. The right piece of text -- some
specialized newsletter, or even a crazily attractive offer from
a 'Computer Mall'-- might find itself copied into thousands of
systems. But I am not a technical person, and cannot establish
to my satisfaction that such an eventuality is truly impossible.

	Is it? 

------------------------------

Date: Sat 7 Sep 85 23:59:24-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Re: Viruses, Trojan horses, and worms
To: SIDNEY.G.HAPGOOD%MIT-OZ@MIT-MC.ARPA
Cc: RISKS@SRI-CSLA.ARPA 

Absolutely not.  It is quite possible.  However, I can assure you that
this issue does not now include a virus -- although some message systems 
tend to permit you to edit a message before resending it, with no
indication that it has been altered.  Thus, even in the presence of all
of those routing headers, you can never be sure you really have picked
up or been forwarded the original message.  The example of squirreled
control characters and escape characters that do not print but cause
all sorts of wonderful actions was popular several years ago, and
provides a very simple example of how a message can have horrible
side-effects when it is read.

Worms, viruses, and Trojan horses from their technical aspects are probably
best discussed elsewhere -- e.g., in SECURITY@RUTGERS.  (See also Fred
Cohen's paper in the 7th DoD/NBS Computer Security Conference in 1984.)
From the RISKS point of view, they are definitely important to this forum
-- and they present a very serious risk to the public.  PGN]

------------------------------

Date: Fri,  6 Sep 85 16:01:38 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  The Case of the Broken Buoy
To: mab@RIACS.ARPA
cc: risks@SRI-CSL.ARPA

    Did the NWS say that (ie, even if the buoy had been alive at
    the time, they could not have predicted the storm) in testimony,
    or after the verdict?  If after the verdict, no comment.

I believe it was during testimony, but I am not certain.

    But
    if as testimony, Herb, the jury (or judge) apparently didn't
    believe the NWS testimony.  If you believe the NWS claim, the
    headline was correct, but it's unfair to say the court ruled
    that way when it explicitly based its ruling on negligence.

But it is not clear that the court understands that the significance
of "missing data" is context-dependent.  Sometimes it matters, and
sometimes it doesn't.  This is a point that non-scientists have a very
hard time understanding.

I am not defending the NWS; they should have repaired the buoy.  But
given limited resources, how are they to set priorities in deciding
what to repair first?  The implications of the verdict are to me
frightening, placing NWS and all other similar organizations in a
double bind: all equipment must be functional even when they don't
have sufficient dollars to keep it that way.

------------------------------

Date:  6 Sep 1985 1359-PDT (Friday)
From: Matt Bishop <mab@riacs.ARPA>
To: Herb Lin <LIN@MIT-MC.ARPA>
Cc: risks@SRI-CSL.ARPA
Subject: Re: The Case of the Broken Buoy

    But it is not clear that the court understands that the significance
    of "missing data" is context-dependent.  Sometimes it matters, and
    sometimes it doesn't.  This is a point that non-scientists have a very
    hard time understanding.

At this point I'm going to bow out of the discussion, since I am not
familiar enough with the decision to know if the court understood that
point.  The NWS certainly should have made its position very clear, so
the court could make an informed decision (about whether or not
negligence was involved.)

------------------------------

Date: Fri,  6 Sep 85 09:24:39 EDT
From: Keith F. Lynch <KFL@MIT-MC.ARPA>
Subject: Re: Hot rodding you AT
To: RISKS@MIT-MC.ARPA

    Date: Wed, 4 Sep 85 14:41:38 EDT
    From: Dan←Bower%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
    Subject: Hot rodding you AT

    In a recent issue of PC Magazine, Peter Norton espoused the idea of
    substituting a faster clock chip to enhance performance.  Now, according
    to the folk on the Info-IBM PC digest, this may create problems.  An
    off the shelf PC AT is composed of components guaranteed to work to
    IBM spec, e.g. 6 Mhz.  If I increase the clock rate, then the whole
    rest of the machine has to be up to snuff.  If not, a part dies and
    I pay a nasty repair bill.

    Now if I took Mr. Norton's word as gospel, swapped chips and set
    my PC AT on fire, would he be liable?  How about the publisher?

I doubt this would break anything.  The machine would simply cease
working above a certain speed, and resume working below that speed.

I know of a couple people who have done this on APPLE computers,
tried various speeds so as to run their machine at the highest speed
it will go.

Also, I once did the same thing with a synchronous link, i.e. hooked
up an external clock and cranked it up to the highest speed it would
work reliably at.

Also, I have done this with my Hayes modem.  The standard duration for
touchtone pulses is 70 ms.  The phone system here will accept as short
at 38 ms.
								...Keith
------------------------------

End of RISKS-FORUM Digest
************************

-------

∂08-Sep-85  1840	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.8, 08 Sep 85, 2nd issue today!    
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 8 Sep 85  18:40:18 PDT
Date: Sun 8 Sep 85 16:21:08-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.8, 08 Sep 85, 2nd issue today!
To: RISKS: ;

RISKS-FORUM Digest      Sunday, 8 Sep 1985      Volume 1 : Issue 8

               ***** NOTE: Second Issue Today *****

          FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                   Peter G. Neumann, moderator

Contents:
  Risks of omission (Nancy Leveson, Nicholas Spies, Herb Lin, Dave Parnas)
  Hot rodding you AT and the weather (John McCarthy)
  Re:  Good Risks and Bad Risks (Brint Cooper)
  SDI reliability (Herb Lin)
  Viruses, Trojan horses, and worms (Lin and Neumann, 2 each -- his own?)

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      
----------------------------------------------------------------------

Date: 08 Sep 85 14:58:56 PDT (Sun)
From: Nancy Leveson <nancy@uci-icsd>
To: RISKS@SRI-CSLA.ARPA
Subject: Risks of omissions

I had not intended to get involved in the RISKS Forum discussions, but
despite my great respect for John McCarthy's accomplishments, I just cannot
let his latest message (RISKS-1.8, Sept. 8) pass without comment.

Some important points that need to be considered:

1) Nothing is completely safe.  All activities and technologies involve
risk.  Getting out of bed is risky -- so is staying there.  Nitrates have
been shown to cause cancer -- not using them may mean that more people will
die of food poisoning.

2) Technology is often introduced for the mere sake of using the "latest,"
sometimes without considering the fact that the situation may not really be
improved.  For example, everybody seems to be assuming lately that machines
will make fewer mistakes than humans and there is a frantic rush to include
computers and "artificial intelligence" in every new product.  Where speed
is the determining factor, then they may be right.  Where intelligent
decision making in the face of unforeseen situations and factors is
foremost, then it may not be true.  Some electro-mechanical devices may be
more reliable than computers.  Since I am identified with the area of
"software safety," I am often consulted by those building safety-critical
software systems.  It is appalling how many engineers insist that computers
do not make "mistakes" and are therefore safer than any other human or
electro-mechanical system.  We (as computer scientists) have often been
guilty of condoning or even promoting this misconception.  Often it seems
that the introduction and use of non-scientific and misleading terminology
(e.g. "intelligent," "expert", "proved correct") has far outstripped the
introduction of new ideas.

3)  Technology introduced to decrease risk does not always result in 
increased safety.  For example, devices which have been introduced into
aircraft to prevent collisions have allowed reduced aircraft separation
with perhaps no net gain in safety (although there is a net gain in efficiency 
and profitability).  There may be certain risk levels that people are
willing to live with and introducing technological improvements to reduce
risks below these levels may merely allow other changes in the system which
will bring the risks up to these levels again.

4) Safety may conflict with other goals, e.g. productivity and efficiency.
Technology that focuses on these other goals may increase risk.

John McCarthy suggests that people ignore the risks of not using technology.
I would suggest that it is not that these risks are ignored, but that they
are known and we have learned to live with them while the risks of using new
technology are often unknown and may involve great societal upheaval in
learning to adapt to them.  Yes, wood smoke may cause lung cancer, but note
that recent studies in Great Britain show that the incidence of prostate
cancer in men who work in atomic power plants is many times that of the
general population.  To delay introducing technology in order to assure that
greater risks are not incurred than are currently borne by the population
seems justifiable.  Yes delay may cause someone's death, but the
introduction may cause even more deaths and disruption in the long-run.

The solution is to develop ways to assess the risks accurately (so that
intelligent, well-informed decision-making is possible) and to develop ways
to reduce the risk as much as possible.  Returning to the topic of computer
risk, citizens and government agencies need to be able to make informed
decisions about such things as the safety of fly-by-wire computer-controlled
commercial aircraft or between computer-controlled Air Traffic Control with
human-assistance vs. human-controlled Air Traffic Control with computer
assistance.  To do this, we need to be able to assess the risks and to
accurately state what computers can and cannot do.

Forums like this one help to disseminate important information and promote
the exchange of ideas, But we also need to start new initiatives in computer
science research and practice.  I have been writing and lecturing about this
for some time.  For example,

  1) we need to stop considering software reliability as a matter of
     counting bugs.  If we could eliminate all bugs, this would work.  But
     since we cannot at this time, we need to differentiate between the
     consequences of "software failures."

  2) Once you start to consider consequences of failures, then it is possible
     to develop techniques which will assess risk.  

  3) Considering consequences may affect more aspects of software than just
     assessment.  Some known techniques, such as formal verification and
     on-line monitoring, which are not practical to detect all faults may be
     applied in a cost-effective manner to subsets of faults.  Decisions may
     be able to be made about the use of competing methodologies in terms of
     the classes of faults that they are able to detect, remove, or tolerate.
     But most important, by stating the "software problem" in a different way 
     (in terms of consequences), it may be possible to discover new approaches
     to it.  My students and I have been working on some of these.  Most
     software methodologies involve a "forward" approach which attempts to
     locate, remove, or tolerate all software faults.  An alternative is to
     take a backward approach which considers the most serious failures and
     attempts to determine if and how they could occur and to protect the
     software from taking these actions.

If using some of these techniques (or despite their use), it is determined
that the software would be more risky than conventional systems or is above
a minimum level of acceptable risk, then we can present decision makers with
these facts and force them to consider them in the decision-making process.
Just citing horror stories or past mistakes is not enough.  We need ways of
assessing the risk of our systems (which may involve historical data
presented in a statistically proper manner) and ways to decrease that risk
as much as possible.  Then society can make intelligent decisions about
which systems should and should not be built for reasons of acceptable or
unacceptable risk.

------------------------------

Date: 8 Sep 1985 12:00-EST
From: Nicholas.Spies@CMU-CS-H.ARPA
Subject: Risks of omissions
To: JMC@SU-AI
Cc: risks@sri-csl

The question of responsibilities for non-use of computers are largely
meaningless in terms of law unless the dangers of non-use were known to
substantially increase the probability of greater harm. In the case of your
three short examples:

(1) If the ACLU had acted in good faith in seeking to limit sharing of
police information and a court had looked favorably on their argument after
weighing the possible risks, then the court is responsible because only the
judge had the ability to decide between two courses of action. To make the
ACLU responsible would be to deny it and its point of view access to due
legal process. To make it necessary for the ACLU to anticipate the court's
response to its bringing suit would have the same chilling effect on our
legal system.

(2) The same argument applies to the Sierra Club and US 101.  If US 101 had
been built and then some people were killed, one could as easily conclude
that the Sierra Club (or anyone else) might be sued for NOT obstructing the
highway!

(3) The "Split Wood not Atoms" poster-vendor might be sued if it could be
conclusively proven that he was a knowing party to a conspiracy to give
people lung cancer. But we might assume that his motivation was actually to
prevent a devastating nuclear accident that might have given 10,000 people
lung cancer...

Again, a risks-of-computers organization can only present its case to court
and people and, so long as no malfeasance is involved, cannot be held
responsible for its failure to predict future consequences. There are far
more important "unsymmetric" relationships than that of the press vs. the
legal system that pertain to issues of responsibility, namely, that of past
vs. future and known vs. unknown. I feel that you are correct in pointing
out how computer people would do well to apply their expertise to solving
problems of society. In this case the moral imperitives are quite clear.

------------------------------

Date: Sun,  8 Sep 85 15:51:44 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  Risks of omissions (not using some technology)
To: JMC@SU-AI.ARPA
cc: RISKS-FORUM@MIT-MC.ARPA, risks@SRI-CSL.ARPA

    	The problem with a forum on the risks of technology is that
    while the risks of not using some technology, e.g. computers, are
    real, it takes imagination to think of them....

You raise an interesting point that deserves more discussion.
However, as one who is concerned that the major problem arises from an
uncritical acceptance of technology, I strongly disagree with your
suggestion that the scales are stacked AGAINST those who are
"pro-technology".  The reason that anyone adopts any given technology
is that it provides benefits that he can see; the problem is getting
those individuals to see that there are costs as well.  In other
words, the bias in the system is towards acceptance of technology, not
rejection of it.  It is this general bias that "risks" people are
trying to correct.

    ...Is a risk-of-computers
    organization that successfully sues to delay a use of computers either
    MORALLY or LEGALLY LIABLE if the delay causes someone's death?  Is there
    any moral or legal requirement that such an organization prove that they
    have formally investigated whether their lawsuit will result in killing
    people?  As the above examples indicate, the present legal situation
    and the present publicity situation are entirely unsymmetric.

There are such precedents; organizations can be held liable for using
technology that that is not as up to date as is "generally
applicable".  On the more general point, it is much harder to
establish liability as the result of someone's inaction as compared to
the result of someone's action; this does happen, but it is harder to
prove. 

    The harm caused by tape-to-tape batch processing as opposed to on-line
    systems.

I like this example; let's have more discussion of it.  There was a
time when on-line systems were totally unreliable, but I think that
time has past.

    Shouldn't computer professionals
    who pretend to social responsibility take an interest in an
    area where their knowledge might actually be relevant?

This is a cheap shot unworthy of pioneers in computer science.  More
than anyone else, computer professionals are the ones who are in the
best position to assess the limits as well as the promises of
technology.  You once said that the responsibility of the scientist is
to take his science whereever it may lead.  I agreed with you then,
and I agree with that now.  But science is part of a social and
cultural context, and cannot be separated as cleanly as one might
imagine.  If it is proper for the developers of science and technology
to propose new applications of their work (because they see these
potential applications as beneficial to society), it is also
appropriate for them to suggest possible consequences of their use as
well. 

------------------------------

Date: Sun, 8 Sep 85 09:06:10 pdt
From: vax-populi!dparnas@nrl-css.arpa (Dave Parnas)
To: Neumann@SRI-CSLA.ARPA
Subject: Risks of omissions; Weizenbaum's message
ReSent-To: risks@SRI-CSLA.ARPA

1.  McCarthy's contribution can be summarized simply,  "Life is 
full of tough decisions about when to use technology; we have to
consider both sides."  Those who have been concerned with the risks
of computer technology have been saying, "There are risks to using
computer technology; we have to consider both sides".  I sense 
agreement on the obvious conclusion.

2.  I can disagree only with one aspect of Weizenbaum's contribution.
He says that he would be against SDI even if it would work, but his arguments
mainly show even more reasons why it won't "make nucelar weapons impotent
and obsolete."  It is probably useless to argue about how we would feel 
about the system if it would work, but I feel the decision would be much 
harder to make than it is now.

    While I would prefer not to waste technological resources on such a thing,
I would see some truth in the argument that the non-technological solutions 
also have a clear risk of failure.  

 Dave

------------------------------

Date: 08 Sep 85  1108 PDT
From: John McCarthy <JMC@SU-AI.ARPA>
Subject: Hot rodding you AT and the weather
To:   risks@SRI-CSL.ARPA    

Assuming, contrary to fact, that putting a faster clock
in your AT would cause it to catch fire, the RISKS contributors are
quite wrong about who would be sued.  According to the new legal
doctrine of  bursae profundae,  it isn't the poor BBOARD operator
that would have to pay but rich IBM.  It wouldn't take much of a lawyer
to figure out that IBM should have anticipated that someone might
do this and warned against it or even somehow made it impossible.

Changing the subject, consider the effect of the court decision that
NOAA was liable for not fixing the buoy.  Up to now weather predictions
have been offered as a non-guaranteed service with the user taking
responsibility for the extent to which he relies on it.  The court
has said that this is no longer possible.  Any institution with
"deep pockets" cannot safely offer information on a user responsibility
basis.  What if Stanford University has negligently failed to
replace a stolen book from its medical library, and someone dies
who would have been saved had his doctor found the book?  Stanford's
lawyers should advise Stanford to deny access to its medical library
to practicing physicians.

------------------------------

Date:     Sun, 8 Sep 85 11:23:16 EDT
From:     Brint Cooper <abc@BRL.ARPA>
To:       RISKS@SRI-CSL.ARPA
cc:       BRANDIN@SRI-AI.ARPA
Subject:  Re:  Good Risks and Bad Risks

Another example of "good news/bad news" is the use of
computerized axial tomography (CAT scans) as a substitute for
exploratory surgery in the head and body.

	Advantages:  elimination of much surgical risk;
		     negative diagnoses (disease NOT present) without surgery
		       much lower cost

	Disadvantages:  because of the advantages, CAT scans may be over-used;
		        increased exposure to X-rays which, itself, can be 
                          disease-enhancing

Brint

------------------------------

Date: Sun,  8 Sep 85 15:58:27 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  SDI reliability
To: mooremj@EGLIN-VAX.ARPA
cc: RISKS-FORUM@MIT-MC.ARPA, risks@SRI-CSL.ARPA

My primary complaint about your otherwise interesting table is that it
assumes independent failure modes.  I think it is much more likely
that the effects of coupled failures are larger.  In particular, given
the failure of one platform, it is more likely that more than one will
fail.

------------------------------

Date: Sun,  8 Sep 85 16:02:40 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  Viruses, Trojan horses, and worms
To: Neumann@SRI-CSL.ARPA
cc: LIN@MIT-MC.ARPA, RISKS-FORUM@MIT-MC.ARPA, RISKS@SRI-CSL.ARPA,
    SIDNEY.G.HAPGOOD@MIT-OZ

    .....  The example of squirreled
    control characters and escape characters that do not print but cause
    all sorts of wonderful actions was popular several years ago, and
    provides a very simple example of how a message can have horrible
    side-effects when it is read.

But if *I* have designed my operating system to display only what it
receives, how is this possible? 

------------------------------

Date: Sun 8 Sep 85 13:35:05-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Re: Viruses, Trojan horses, and worms
To: LIN@MIT-MC.ARPA

The example of squirreled non-printing characters is of course just 
one example; that one was relatively easy to fix once it was recognized. 
But until you have discovered a flaw, you are vulnerable.  And you never
know how many flaws remain undiscovered.

Sure, *you* may be smart, but large operating systems generally have many
security flaws.  *You* probably aren't smart enough to design and build a
system that has none.  (In fact, your solution is not quite good enough
by itself -- without some other assumptions on the rest of your system.)

Peter

    [By the way, let me add before someone comments, a message that causes
     grief when read would be a Trojan horse.  A message that propagates itself
     -- as in the case of the ARPANET collapse on 27 Oct 1980 -- would be
     a virus.  A BBOARD message that is innocently moved to another system 
     by third parties for others to get clobbered by has aspects of both --
     a virus-propagated Trojan horse .  PGN]

------------------------------

Date: Sun,  8 Sep 85 16:40:44 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  Viruses, Trojan horses, and worms
To: Neumann@SRI-CSL.ARPA
cc: LIN@MIT-MC.ARPA
In-reply-to: Msg of Sun 8 Sep 85 13:35:05-PDT from Peter G. Neumann <Neumann at SRI-CSLA.ARPA>

    Sure, *you* may be smart, but large operating systems generally have many
    security flaws.  *You* probably aren't smart enough to design and build a
    system that has none.  (In fact, your solution is not quite good enough
    by itself -- without some other assumptions on the rest of your system.)

I certainly recognize that operating systems have security flaws in
them, having been an OS hacker myself at one time.  But for the
problem of messages (as opposed to programs) doing bad things to my
system, I guess I never figured out a way of doing that.

My naive model is that I have a special program that intercepts the
raw bit stream that comes in from my communications port.  It then
translates this into ASCII, and then prints it on my screen.  

If this is all that my program does, I can't see what harm can be
done.

Now, if my program writes the message to disk, then I can see
potential problems.  So I simply count the characters that are sent to
me over the line, and save exactly that many characters on disk.

What am I missing?

------------------------------

Date: Sun 8 Sep 85 14:47:21-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Re: Viruses, Trojan horses, and worms
To: LIN@MIT-MC.ARPA

Herb, Indeed your model is a bit naive, in that you are looking at the
problem much to narrowly, and assuming that *everything* *else* works fine.
Suppose that your special program would work as you wish in intercepting the
raw bit stream.  Suppose, however, that there is an operating system flaw
that lets me change your special program to do what I wish! (There are many
examples of how this might happen.)  Now your program can be effectively
bypassed.  The point is NOT whether you can seal off one hole, but rather
that you are dealing with the tip of an iceberg and there may be titanic
holes you don't even know about.  Besides, as I said earlier, the message
squirreling is only one example -- and hopefully completely cured.  So I
hope you don't reply that you could use seals to guarantee that your program
is unchanged.  That would still miss the broader point.  PGN

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂09-Sep-85  1257	YAMARONE@SU-CSLI.ARPA 	ONE EXTRA SANDWICH:TURKEY/LT.RYE
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Sep 85  12:57:24 PDT
Date: Mon 9 Sep 85 12:52:06-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: ONE EXTRA SANDWICH:TURKEY/LT.RYE
To: folks@SU-CSLI.ARPA



there happens to be one extra sandwich available at the desk:

turkey on light rye: 2.50

ignore this message if you receive it later than 1:15.

Thanks, the Ventura Sandwich Corp.

-------

∂09-Sep-85  1337	SUSI@SU-CSLI.ARPA   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Sep 85  13:37:04 PDT
Date: Mon 9 Sep 85 13:28:04-PDT
From: Susi Parker <SUSI@SU-CSLI.ARPA>
To: folks@SU-CSLI.ARPA


 THE turkey on light rye is still available and edible...until further notice
-------

∂09-Sep-85  1417	CAROL@SU-CSLI.ARPA 	DANDELION ASSISTANCE
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Sep 85  14:17:15 PDT
Date: Mon 9 Sep 85 14:10:19-PDT
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: DANDELION ASSISTANCE
To: FOLKS@SU-CSLI.ARPA

Hi.  I'm Carol Kiparsky.  As promised in the long message summarizing the first Dandelion Users' Group meeting, I am here to help people at CSLI use the dandelions.  

First of all, I am assembling a package of documentation concerning such areas as: the basics of the Interlisp-D world  (windows, mouse, etc.), the editing system, useful packages, troubleshooting, and using CHAT to simulate a terminal so you can, for instance, use the mail system.  

Secondly, I am on hand to give tutorials to new users or answer questions that arise while you use the machines.  

Third, I will attempt to keep you informed of relevant developments from PARC. 


By the end of the month I hope to have the package of documentation ready. The material will be available in hardcopy, and also as files on the directory:

	  {csli}<dandelion.doc>

which is open to  all to read and print out.  At this point it contains material on the TEdit editing system.  Browse  this directory, and if you don't see what you want, contact me.  Hopefully I can find it for you.

Address questions and requests for documentation, appointments,  or other assistance to Carol@csli, or call me at 497-0939, or drop in weekdays at F2. 

-------

∂09-Sep-85  1517	@SU-CSLI.ARPA:GEORGEFF@SRI-AI.ARPA 	PLANLUNCH resumes next week  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Sep 85  15:17:12 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 9 Sep 85 15:09:51-PDT
Date: Mon 9 Sep 85 09:08:11-PDT
From: Michael Georgeff <georgeff@SRI-AI.ARPA>
Subject: PLANLUNCH resumes next week
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanLehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA

The PLANLUNCH seminars will resume next Monday (September 16).
The speaker next week will be Dave Wilkins.

Mike.
-------

∂09-Sep-85  2134	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.9, 09 Sep 85  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 9 Sep 85  21:34:04 PDT
Date: Mon 9 Sep 85 20:40:51-PDT
From: RISKS FORUM (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.9, 09 Sep 85
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS: ;

RISKS-FORUM Digest       Monday, 9 Sep 1985      Volume 1 : Issue 9

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  McCarthy, Weizenbaum on SDI  (Douglas Schuler)
  Why I'm against even a reliable SDI (Jeffrey Mogul)
  Risk Assessment and Risk Management (Edward V. Berard)
  Risks in displaying a file containing control characters (Keith F. Lynch)

[*** Apologies to those of you who have been suffering from INDIGESTIFICATION.
The blank line before the following 70 hyphens makes the difference, and your
undigestification programs should now work on RISKS!  PGN ***]

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Mon, 9 Sep 85 09:34:15 pdt
From: bcsaic!douglas@uw-june (douglas schuler)
To: RISKS@SRI-CSL
Subject: McCarthy, Weizenbaum on SDI

Joseph Weizenbaum states that he would be against SDI "even if it worked."
I agree.  The premise that "IF the SDI could work, then we must have it (at
any price)" is naive.  It seems that many people are willing to accept that
premise hoping that it will anchor the discussion in the technical area.

One factor which is rarely addressed is that of intermediate systems which
the SDI will spawn.  There is a tendency to think of the SDI system as being
one big system that one day appears overhead as a whole, integrated system.
I have little doubt that the SDI plan includes many deliverables along the
way.  These intermediate systems will both be arguably non-defensive and
pose a large problem for integration.  I.e., the system must not only be
trustworthy when "fully deployed" but at a multitude of intermediate steps.
Thus risks will exist in advance of the full delivery (if there ever is to
be one).

Another factor is the so-called defensiveness of the system.  If two people
are armed with guns and one suddenly dons a bullet proof vest this act will
be perceived as an offensive act.

Pro-SDI people almost always accuse SDI critics of being politically
motivated.  Given the immense (and possibly impossible) technical task of
getting the system to work, and the guarenteed proliferation of offensive
weapons (designed to penetrate the system), it is very, very difficult for
me to believe that the pro SDI folk are not motivated primarily from
political (and economic!!) grounds.

  - Doug Schuler

------------------------------

Date: Mon 9 Sep 85 16:40:02-PDT
From: Jeffrey Mogul <MOGUL@SU-SCORE.ARPA>
Subject: Why I'm against even a reliable SDI
To: risks@SRI-CSL.ARPA

To quote from RISKS Vol 1 #8:
    2.  I can disagree only with one aspect of Weizenbaum's contribution.
    He says that he would be against SDI even if it would work, but his
    arguments mainly show even more reasons why it won't "make nucelar weapons 
    impotent and obsolete."  It is probably useless to argue about how we
    would feel about the system if it would work, but I feel the decision
    would be much harder to make than it is now. [Dave Parnas]

I think this touches on the crux of the matter: what problem is SDI meant to
solve?  If we could guarantee that SDI would not only "make nuclear weapons
impotent and obsolete", but would in fact reduce the risks associated with
war (not necessarily the same thing) then I would not be against SDI.
However, I argue (and I suspect this is Weizenbaum's point, too) that an SDI
that worked according to the current specification would actually increase
risks, even though the system performed "flawlessly".

This is not the place to discuss the strategic implications of SDI, but I
think it's important to realize that there are those of us who believe both
that SDI is not likely to meet its current specification, nor that it would
be a good idea even if it did.

    [I] would see some truth in the argument that the non-technological
    solutions also have a clear risk of failure.  [Parnas]

I am afraid that there is no failure-proof solution, technological or not,
to the problem of "war".  John McCarthy is right that we must compare
the risks of the technological solution (e.g., SDI) to its non-use.  My
fear is that, in this case, the problem is not that the use of technology
might fail to solve the problem, but that it might actually make things
worse.

------------------------------

Date: Mon 9 Sep 85 08:16:07-PDT
From: Edward V. Berard <EBERARD@USC-ECLB.ARPA>
Subject: Risk Assessment and Risk Management
To: risks@SRI-CSL.ARPA

There has been some discussion of comparing alternative risks on the
RISKS mailing list lately. For example, what is the risk associated
with the introduction of a new technology versus not introducing the
technology? Risk assessment and risk management need not be
"guesstimates" nor "a number picked out of the air."

The insurance industry has had to assess and manage risks for years.
In fact, they have made quite a science out of these two areas. I
would recommend that those who wish to find out more about risk
management and risk assessment read:

   RISK MANAGEMENT AND INSURANCE, Fourth Edition, by C. Arthur
   Williams, Jr. and Richard M. Heins, McGraw-Hill, 1981.

Don't let the title put you off. Virtually the entire book is
dedicated to risk management, with only a few pages on insurance. You
will also find that there are entire professional societies dedicated
to managing and assessing risk, e.g., the American Risk and Insurance
Association and the Risk and Insurance Management Society.	

  		-- Ed Berard
				   EBerard at ECLB
				   (301) 251 - 1626

------------------------------

Date: Mon,  9 Sep 85 00:26:04 EDT
From: Keith F. Lynch <KFL@MIT-MC.ARPA>
Subject: Risks in displaying a file containing control characters
To: LIN@MIT-MC.ARPA
cc: Risks@SRI-CSL.ARPA, Security@RED.RUTGERS.EDU

    Date: Sun,  8 Sep 85 16:40:44 EDT
    From: Herb Lin <LIN@MIT-MC.ARPA>

    My naive model is that I have a special program that intercepts the
    raw bit stream that comes in from my communications port.  It then
    translates this into ASCII, and then prints it on my screen.  

    If this is all that my program does, I can't see what harm can be done.

Several kinds of terminals are programmable from the host, in that
certain escape sequences can be sent to them to get them to perform
actions such as defining the terminal's function keys.

If a user inserts the appropriate escape sequences in a mail message
to his system manager, or into a file which will be displayed by the
manager, when the manager reads that mail message it will reprogram
a function key on the manager's terminal, which the manager may have
programmed to do some common harmless function, to instead do some
other command such as give the user unauthorized privileges.

This is a fairly well known bug, and many mail systems are now
protected against it, in that they will not transmit any control
characters or escape sequences to the terminal.

The moral is that there are many subtle ways to break security,
and even things that seem to be quite safe may not really be.
								...Keith

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂09-Sep-85  2347	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #14  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Sep 85  23:47:36 PDT
Date:  9 Sep 85 2335-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #14
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Tuesday, 10 Sep 1985      Volume 1 : Issue 14

Today's Topics:

          Recent books and the symbolic computing discussion
                          RIACS job openings
              Programming language theory seminar (MIT)
              IJPP: new journal on parallel programming


----------------------------------------------------------------------

From: eugene@AMES-NAS.ARPA (Eugene Miya)
Date: 3 Sep 1985 1644-PDT (Tuesday)
Subject: Recent books and the symbolic computing discussion

The following titles have just appeared at Computer Literacy Book store.
I have not seen any of them at Stanford or Staceys yet:

%A John A. Sharp
%T Data Flow Computing
%I John Wiley & Sons, Inc.
%D 1985
%X Appears to have a slight English bias. Very little about SISAL,
some VAL.

%A William W. Wadge
%A Edward A. Ashcroft
%T Lucid, the Data Flow Programming Language
%I Academic Press
%D 1985

%A W. Daniel Hillis
%T The Connection Machine
%I MIT Press
%D 1985
%X A collection of papers on the ideals of this architecture.

%A J. L. Potter, ed.
%T The Massively Parallel Processor
%I MIT Press
%C Cambridge, MA
%D 1985

Sorry, they are in refer format.  I don't have Scribe.

I think the discussion on the distinction between numeric and
symbolic computing was way too one sided.  I was not
a very good devil's advocate in the beginning.  I have been
following a discussion on Scientific Programming Languages in
the net.lang newsgroup [much more lively, one chap really defended
using Ada].  The supporters of LISP were alive and a sub group
has spun off discussing why LISP isn't used more in scientific
environments.  It has not occurred to them that a lot of
engineers and scientists prefer "X = 1 + 2" to "(ADD 1 2)" semantics
aside.  I fear we have to get FORTRAN classes out of the physics
departments or else the physicists are not going to see the
advantages of tools like macsyma.

--eugene miya
  NASA Ames Research Center

------------------------------

[Forwarded by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

From: Nancy Blachman <nancy@RIACS.ARPA>
Date: 26 Aug 1985 1405-PDT (Monday)
To: arpanet-bboards@mit-mc.ARPA
Subject: Openings at the forefront of Computer Science

We, the Research Institute for Advanced Computer Science (RIACS), are
looking for a systems programmer to provide support for computer science
research in distributed computing, concurrent processing and programming,
among other things. We are a private non-profit contractor to NASA Ames
Research Center.  We are committed to UNIX (primarily BSD) shop and an
Internet site (riacs.ARPA).

We have the following equipment:
	Sequent Balance 8000 (12 processors, 16MB main memory, 1.5GB disk,
	    32 ttys, 16 users, 4.2BSD),
	VAX 11/730 (3MB main memory, 110MB disk, 4 dialin/out, 0 users
	    [uucp/usenet/internet gateway], 4.2BSD),
	Ridge 32C (4MB main memory, 370MB disk, SysV & 4.2BSD layered over
	    ROS 3.3, 0 users [compute server]),
	SGI IRIS 1500 Color Graphics Workstation (68k, geometry engine,
	    SysV with 4.2 enhancements and TCP/IP),
	Intel iPSC hypercube (32 nodes, delivery 8/19/85),
	BBN and Forward Technology bitgraph terminals,

as well as plans and a budget for:
	color graphics system (unspecified),
	workstations (unspecified).

All machines are networked together (TCP), and all have full access to
MILNet, ARPANet, and UUCP.

The systems programmer would work on problems such as:
	Simulators for new memory architectures that will be used as
	   components of autonomous robotic systems in the Space Station
	   Project.
	Creation of high performance workstations for use in the Space
	   Sciences and Human Factors branches.
	Support for multimedia and private network telecommunications
	   as part of scientific collaboration.
	Computational fluid dynamics (CFD) research, using an IRIS
	   workstation as tool to develop flow simulations and also as
	   front end for a Cray-2 and other supercomputers in the Numerical
	   Aerodynamic Simulation (NAS) project.

Currently there are 19 people at RIACS, including five summer students and
10 research scientists. Our environment is university-like, however, our
salaries are on par with industry.  We are located at the NASA Ames Research
Center in Mountain View, California. (We are an EOE employer.)

We are looking for someone with  a solid UNIX systems programming
background. If you are interested in the position, you have at least two
years experience with UNIX, and you are a U.S. citizen or you hold a
permanent resident visa, then to apply for the position, send a cover
letter, resume, and names of three references to:

	Peter J. Denning, Director
	RIACS
	Mail Stop 230-5
	NASA Ames Research Center
	Moffett Field, CA  94035
or
	nancy@riacs.ARPA

If you simply want more information contact me directly.

	Nancy Blachman
	nancy@riacs.ARPA
	415-694-6144

------------------------------

[Forwarded by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

Date: 09/08/85 17:59:32
From: MEYER at MIT-XX.ARPA
Re:   Programmng language theory seminar
To:   *bboard@MIT-XX
Subject: Programming language theory seminar
cc:   types@mc, toc@mc


                    SEMINAR IN PROGRAMMING LANGUAGE THEORY
                      TYPES; CONCURRENCY SEMANTICS; ETC.

                                ORGANIZING CALL

Reading seminar for presentation of recent papers.  Main foci on
generalized types in programming and concurrency semantics.  Tentative
time Thursdays at 3:00; first official meeting Thursday, Sept.19 in
3rd floor conference room.  If interested, send mail to MEYER@MC
indicating your schedule and papers you'd like to present or see
presented.  Credit can be arranged.

Papers currently proposed are

   - Semantics  of  concurrent  programs  based  on observed computational
     behavior [Pnueli  85], [Hennessy  85], [Larsen  85], [Stirling   85],
     [Graf, Sifakis 85],

   - Theory   of   representation  independence  of  abstract  data  types
     [Mitchell, Meyer 85], [Donahue, Demers 85], [Reynolds  83], [Plotkin 80],
     [Statman 85], [Fokkinga 81],

   - Simple   types:   semantics [Coppo  85], [Breazu-Tannen,  Meyer  85],
     continuations [Meyer, Wand 85],

   - type  checking [Mishra,  Reddy  85]  and   extensible   type   system
     facilities [Lucassen, Gifford 85].

                                  REFERENCES

Breazu-Tannen, V. and A.R. Meyer.  Lambda calculus with constrained types
(Extended abstract).  In Logics of Programs, Lect. Notes in Comp. Sci. 193,
Parikh, Rohit, Ed., Springer-Verlag, 1985, 23-40.

Coppo, M.  A completeness theorem for recursively defined types.  In Int'l
Conf. Automata, Languages and Programming, Lect. Notes in Comp. Sci. 194,
Brauer, Wilfried, Ed., Springer-Verlag, 1985, 120-129.

Donahue, James and Alan Demers.  Data types are values.  ACM Trans. on
Programming Lang. and Systems 7 (1985), 426-445.

Fokkinga, Maarten M.   On  the  notion  of  strong  typing.    In
Algorithmic  Languages,  DeBakker, J. and van Vliet, Eds., IFIP, North-Holland,
1981, 305-320.

Graf, S. and J. Sifakis.  From synchronisation tree logic to acceptance.  In
Logics of Programs, Lect. Notes in Comp. Sci. 193, Parikh, Rohit, Ed.,
Springer-Verlag, 1985, 128-142.

Hennessy, M.  An algebraic theory of fair asynchronous communicating processes.
In Int'l Conf. Automata, Languages and Programming, Lect. Notes in Comp. Sci.
194, Brauer, Wilfried, Ed., Springer-Verlag, 1985, 260-269.

Larsen, K.G.  A concept dependent equivalence between processes.  In Int'l
Conf. Automata, Languages and Programming, Lect. Notes in Comp.  Sci.  194,
Brauer, Wilfried, Ed., Springer-Verlag, 1985, 373-382.

Lucassen, John M.  and David K. Gifford.  A strongly typed language with an
extensible type system.  1985, MIT, extended abstract submitted to POPL 86.

Meyer, A.R.  and M. Wand.  Continuation semantics in typed lambda-calculi
(Summary).  In Logics of Programs, Lect.  Notes in Comp.  Sci.  193, Parikh,
Rohit, Ed., Springer-Verlag, 1985, 219-224.

Mishra, Prateek and Uday S. Reddy.  Declaration-free type checking.  In 12th
ACM Conf. Principles of Programming Languages, 1985, 7-21.

Mitchell, J.C.  and A.R.  Meyer.  Second-order logical relations (Extended
Abstract).  In Logics of Programs, Lect. Notes in Comp.  Sci. 193, Parikh,
Rohit, Ed., Springer-Verlag, 1985, 225-236.

Plotkin, Gordon D.  Lambda-definability in the full type hierarchy.  In To H.B.
Curry: Essays on Combinatory Logic, Lambda Calculus and Formalism, Seldin, J.P.
and J.R. Hindley, Eds., Academic Press, 1980, 363-373.

Pnueli, A.  Linear and branching structures in the semantics and logics of
reactive systems.  In Int'l Conf.  Automata, Languages and Programming, Lect.
Notes in Comp.  Sci.  194, Brauer, Wilfried, Ed., Springer-Verlag, 1985, 15-32.

Reynolds, John C.  Types, Abstraction, and Parametric Polymorphism.  In
Information Processing 83, R.E.A. Mason, Ed., 1983, 513-523.  North-Holland
Publishers.

Statman, R.  Logical relations in the typed lambda-calculus.  1985, To appear,
Information and Control.

Stirling, C.  A complete compositional modal proof system for a subset of CCS.
In Int'l Conf. Automata, Languages and Programming, Lect. Notes in Comp. Sci.
194, Brauer, Wilfried, Ed., Springer-Verlag, 1985, 475-486.

------------------------------

[Forwarded by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

Date: Sun 1 Sep 85 23:29:07-PDT
From: Gary Lindstrom <Lindstrom at UTAH-20.ARPA>
Sender: udi at wisc-rsch.arpa
To:   potential-authors:; at UTAH-20.ARPA
Re:   IJPP: new journal on parallel programming

The following is an announcement of a new journal.  Contributions are now
actively being solicited, as explained below.  Please post this message
on local bulletin boards (hard and soft!).

Thanks,
	--Gary
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

		   INTERNATIONAL JOURNAL OF PARALLEL PROGRAMMING


			Editor-in-Chief:
				Gary Lindstrom
				Department of Computer Science
				University of Utah
				Salt Lake City, Utah 84112
				(801) 581-5586
				Lindstrom@Utah-20.ARPA

			Published by:
				Plenum Publishing Corporation
				233 Spring Street
				New York, NY  10013
				(212) 620-8000



Statement of Purpose:

	Parallel computer systems are characterized by the coexistence of
multiple coordinated activities.  Physically, such systems range from networked
independent computers to high performance single-chip processors.

	While parallelism has long been a physical reality in computer systems,
its impact on programming practice has in the past generally been felt only by
specialists in systems programming.  A new trend, however, is clearly emerging:
that of parallelism as a vital aspect of programming at all levels.

	The reasons for this trend are several, including the growing purview
of higher level languages, more elegant computing models relieved of sequential
computing ideology, and exciting new ideas in computer architecture motivated
by the advent of VLSI.

	The International Journal of Parallel Programming (IJPP) will offer a
window on these developments from a programmer's perspective.  Thus while no
pertinent aspect of parallel computing systems will be excluded a priori,
each contribution will observe the common theme of illuminating in some manner
the programming challenges presented by parallel computing systems.


Scope:

	Sample areas to be addressed in IJPP include (with a short suggestive
list of appropriate subareas mentioned in each case):

	* linguistic foundations (semantic formulations, abstract machine
		models)
	* conceptual frameworks for parallel computation (actors, guardians,
		synchronized tasks, functional and logic programming)
	* high-level languages for parallel programming (new formulations and
		extensions to existing languages)
	* evaluation methods (dataflow, reduction, eager vs. lazy evaluation)
	* implementation techniques (compilation, emulation, storage and
		name space management)
	* programming support systems (run-time libraries, debuggers,
		performance instruments)
	* pragmatic considerations (task granularity, load distribution and
		scheduling, error detection and recovery)
	* architectural characteristics (synchronous vs. asynchronous control,
		shared vs. partitioned memory, physical locality, reliability)
	* software engineering aspects (specification and design
		methodologies, rapid prototyping, migration of sequential code,
		verification and testing)
	* advances in parallel algorithms (prototypical problems, relationships
		to particular languages and evaluation methods)
	* performance studies (fundamental and empirical)
	* application studies (artificial intelligence, simulation, databases,
		numeric computation)

	In addition to full length refereed contributions, IJPP will offer a
department entitled "Nonpareil" (meaning, in French, both "nonparallel" and
"matchless").  Under this rubric will appear short contributions of a lighter
nature, including anecdotes, tongue-in-cheek commentary, and other entertaining
insights into the travails of parallel programming.


Editorial Board:

	A distinguished Editorial Board comprising approximately 20 experts in
the areas cited above is now being formed.


Commencement of the Journal:

	The first issue of IJPP will appear on or about July 1, 1986. Six
issues of approximately 90 pp. each will be issued per year.

	IJPP will be the successor to the International Journal of Computer and
Information Sciences, concluding fourteen volumes published under the
editorship of Julius T. Tou.  The first issue of IJPP will thus be designated
volume 15, number 1.


Call for Papers:

	Manuscripts for possible publication in IJPP should be sent to the
Editor-in-Chief.  Contributions must be in English, and previously unpublished.
Four single-sided copies are requested.

	To ensure eligibility for inclusion in the inaugural issue, manuscripts
should be received by October 15, 1985.

	Electronic mail will be used for correspondence with contributors
whenever feasible, and network addresses should be prominently indicated on
all submissions and inquiries.

------------------------------

End of PARSYM Digest
********************

∂10-Sep-85  1014	RICE@SUMEX-AIM.ARPA 	Tomorrow's Meeting.
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 10 Sep 85  10:14:11 PDT
Date: Tue 10 Sep 85 10:14:34-PDT
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Tomorrow's Meeting.
To: aap@SUMEX-AIM.ARPA
Message-ID: <12142174130.25.RICE@SUMEX-AIM.ARPA>


Ok, This is your last warning.

9 am.  Wed Sept 11

Dr Shimon Cohen (Schlumberger P.A. res.)
will describe language development work that
he is doing as part of the Faim project.

Rice
-------

∂10-Sep-85  1402	RICHARDSON@SU-SCORE.ARPA 	Sr. Faculty Meeting
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Sep 85  14:02:25 PDT
Date: Tue 10 Sep 85 13:58:39-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Sr. Faculty Meeting
To: tenured@SU-SCORE.ARPA
Message-ID: <12142214922.31.RICHARDSON@SU-SCORE.ARPA>

There will be a sr. faculty meeting on September 24, 1985 following the
general faculty meeting scheduled at 2:30 in Conference Room 146 of
Margaret Jacks Hall. Should you have any agenda items that you would like
to be included, please indicate to me.

Thanks,
Anne Richardson
-------

∂10-Sep-85  2130	ullman@diablo 	weird message  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 10 Sep 85  21:30:29 PDT
Date: Tue, 10 Sep 85 21:27:13 pdt
From: Jeff Ullman <ullman@diablo>
Subject: weird message
To: nail@diablo

There was a message from outer space yesterday, announcing last week's
meeting.  This week's meeting will be at our usual time, 11AM
wednesday in 301.  Jeff Naughton is the speaker.

∂11-Sep-85  1239	BETSY@SU-CSLI.ARPA 	CSLI Activities in 1985-86    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Sep 85  12:39:27 PDT
Date: Wed 11 Sep 85 12:30:26-PDT
From: Betsy Macken <BETSY@SU-CSLI.ARPA>
Subject: CSLI Activities in 1985-86
To: folks@SU-CSLI.ARPA



The following is information about our plans for the structure of CSLI
activities in 1985-86:

As you know, we have organized the CSLI budget and our remaining SL
funds around the projects described in your proposals.  We stated in
the second year report that these projects are the ones CSLI is
committed to for the next three or four years.  We'd like you to
commit to regular meetings for your project, and we want to organize
our seminars and colloquia around them as well.
  

Project Meetings  

To facilitate regular meetings, we are trying to make the Ventura site
a more attractive place to hold them, particularly during the hours
around lunch time.  We have plans to improve and expand our common
areas and our lunch service.  Regular transportation from SRI and PARC
could also be arranged if this would be helpful.  We'd like you to
consider saving the hours between 11 and 2 for project meetings; this
would give you 14 one-hour slots (excluding the TINlunch slot) per
week to choose from, and we thought that by everyone thinking of these
times as CSLI times, it might be easier to arrange meetings and also
to find people available for informal meetings.  In addition, it would
eliminate the need for people to go back and forth between sites two
or three times in a day.


TINLunch

TINLunch will have the same format as in previous years.  Chris
Menzel and Mats Rooth will be our leaders this year. 


Thursday Seminars

Besides individual presentations from the postdocs, we'd like each
project to present its goals and progress in at least one seminar
meeting during the coming year.  I have attached a tentative list of
dates for projects and postdocs to give seminars.  You are free to
trade with each other if the dates are not appropriate, but you do
need to let me know of whatever trades you make so that we always have
an updated list.

Thursday Colloquia

We'd like to make CSLI's colloquium series a major part of our outreach
program where each meeting is of exceptionally high interest to the
majority of CSLI researchers and, as much as possible, to the wider
academic community.  At the same time, we'd like to have fewer and less
regular meetings to allow more time and flexibility for your project
meetings. 

To accomplish this we decided to:

	1) Have three colloquia meetings each quarter

	2) Ask projects to be responsible for, or share the
	   the responsibility for, one meeting

	3) Be flexible about the meeting times and places; that
	   is, meetings can be held in the evenings and in lecture
	   rooms which are larger than G-19 whenever necessary

	4) Provide up to $1000 per meeting from general funds 
	   for guest speakers

I have tentatively partitioned the academic year into 9, 2- to 3-week
periods and assigned one or more project to each period.  I've asked
the projects to be responsible for one colloquium meeting within their
assigned period of time or for making alternative arrangements by
combining forces or trading with other projects.


Exceptions

If you have ideas for additional seminars or colloquia that don't seem
to fit in these lists, please let me know and I'll work out a way to
schedule them.  We have already had some requests, and we don't want
to turn down interesting events.  

						- Betsy


Tentitive Seminar and Colloquium Assignments

Seminars:

10-3	Situation Theory and Situation Semantics 
10-10 	Rational Agency
10-17	Menzel
10-24	Discourse, Intention and Action
10-31	Foundations of Document Preparation
11-7	Phonology and Phonetics 
11-14	Finite State Morphology
11-21	Computational Models of Spoken Language
1-9	Blair
1-16	Embedded Computation:  Research on Situated Automata
1-23	Embedded Computation:  Semantically Rational Computer Languages
1-30	Kirchner
2-6	Embedded Computation:  Representation and Reasoning
2-13	Semantics of Computational Languages
2-20	Cleland
2-27	Linguistic Approaches to Computer Languages
3-6	Rooth
3-27	Grammatical Theory of Discourse Structure
4-3	AFT Lexical Representation Theory
4-10	Gawron
4-17	Lexical Initiative
4-24	Head-Driven Phrase Structure Grammar
5-1	Sells
5-8	Foundations of Grammar
5-15	Zalta
5-22	Visual Communication

Colloquia:

10-3 to 10-17:	Situation Theory and Situation Semantics
10-24 to 11-7:	Discourse, Intention and Action
11-14 to 11-21:	Phonology & Phonetics, Finite State Morphology, & 
                Computational Models of Spoken Language
1-9 to 1-23:	Embedded Computation & Document Preparation
1-30 to 2-13:	Embedded Computation, Computational Language
	        Semantics, & Linguistic Approaches to Computer Languages
2-20 to 3-6:	Rational Agency
3-27 to 4-10:	Grammatical Theory of Discourse Structure, Lexical 
		Initiative, & AFT Lexical Representation Theory
4-17 to 4-24:	Head-Driven Phrase Structure Grammar & Foundations of
		Grammar
5-1 to 5-22:	Visual Communication







-------

∂11-Sep-85  1741	EMMA@SU-CSLI.ARPA 	Newsletter September 12, No. 45
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Sep 85  17:41:24 PDT
Date: Wed 11 Sep 85 16:56:02-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter September 12, No. 45
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
September 12, 1985              Stanford                       Vol. 2, No. 45
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *THIS* THURSDAY, September 12, 1985

   12 noon		TINLunch
     Ventura Hall       ``Free Word Order in GPSG'' 
     Conference Room    by Arnold Zwicky
			Discussion led by Hans Uszkoreit, SRI and CSLI

   2:15 p.m.		CSLI Talk
     Ventura Hall	``Arithmetical Truth and Hidden Higher-Order Concepts''
     Seminar Room	Daniel Isaacson, Oxford University

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *NEXT* THURSDAY, September 19, 1985

   12 noon		TINLunch
     Ventura Hall       ``Some Remarks on the Relationship of Mind to 
     Conference Room    Meaning and Language''
			Discussion led by Daniel Isaacson, Oxford University
			(Abstract on page 1)

   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week
     Seminar Room	

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
  ``Some Remarks on the Relationship of Mind to Meaning and Language''

     These remarks will be attempting to point toward a mind-based
   approach to meaning and language, and to say something as to why one
   might not be persuaded by considerations which have been taken as
   disallowing such an approach and as requiring rather that
   understanding of the phenomenon of language must underlie our
   understanding of mind.  This perspective is partially motivated by
   focusing attention on progression of the infant from pre-verbal states
   of mind to linguistic expression.  Access to pre-verbal mental states
   as required on this approach may be provided by psychoanalysis, in
   particular by the work of Melanie Klein.  In these terms, basic
   cognitive and emotional development constitute two aspects of a single
   process.					--Daniel Isaacson
!
Page 2                     CSLI Newsletter                  September 12, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                 TALK
    ``Crossing the Rubicon: From a Physics of Dead Coordinate Spaces
               to a Physics of Living Coordinate Spaces''
   Dr. Peter Kugler, The Crump Institute for Medical Engineering, UCLA
                   Monday, September 23, 1985, 2:15pm

      This talk will be about the Ecological (Gibsonian) view of
   language, and will concentrate upon the conceptual tools which the
   Ecological approach considers necessary for the study of language.  A
   more complete abstract will appear in next week's newsletter.
                              ←←←←←←←←←←←←
                         ARTICULATORY PHONETICS
                             Osamu Fujimura
                         AT&T Bell Laboratories

   A one-week course sponsored by the Linguistics Department and CSLI.

	Place: Seminar Room, CSLI, Stanford University
	Dates: Tuesday, September 17 - Monday, September 23
	Hours: Tuesday 2-5, WThFM 10-12, 2-5

   Course description:
      The current status of articulatory studies will be reviewed with an
   emphasis on recent findings using a computer-controlled x-ray
   microbeam system.  The movement patterns of articulatory organs such
   as the tongue, the lower lip, the mandible and the velum reveal strong
   prosodic effects on what usually are considered segmental gestures,
   e. g. vowel gestures.  The temporal organization of the
   multidimensional articulatory patterns cannot be explained by the
   conventional segment concatenation models.  Some new principles of
   phonetic organization will be examined, and their implications
   concerning phonological representations of speech will be discussed.
      A relatively large sample of microbeam (pellet movement) data will
   be provided for student exercise, using the phonetics laboratory's
   interactive computer facility and specially prepared analysis tools.
   A two-hour lecture in the morning will be normally followed by
   laboratory work using the graphics terminals during the afternoon
   (and, if there is demand, during the evening as well).
      No particular background knowledge will be presupposed, and there
   is no registration fee. If you expect to attend, please contact Paul
   Kiparsky at CSLI, Ventura Hall, Stanford University, Stanford, CA
   94305 (net address: Kiparsky@CSLI.ARPA).

   Course Outline

   Tuesday	2-5	Anatomy: Articulatory organs, observation and
   			  measurement methods
   Wednesday	10-12	Physics: Linear systems and acoustics of speech
   			  production; perception
		2-5	Lab setup and basic demo
   Thursday	10-12	Some observations from X-ray microbeam data
		2-5	Exercise (Lab)
   Friday	10-12	Temporal organization of speech
		2-5	Exercise (Lab)
   Monday	10-12	Models of speech production in relation to phonology
		2-5	Overall discussion
-------

∂11-Sep-85  1742	BRESNAN@SU-CSLI.ARPA 	Valentina Zavarin 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Sep 85  17:41:53 PDT
Date: Wed 11 Sep 85 17:38:11-PDT
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: Valentina Zavarin
To: researchers@SU-CSLI.ARPA
cc: zavarin@SU-CSLI.ARPA

Valentina Zavarin has returned from Paris, where she worked with
Greimas' group on a project comparing the semiotic and Schankian
models of story telling.  She has worked on point of view,
discourse, and language from literary, linguistic, and
psychological perspectives, and would like to participate in
ongoing CSLI projects in these areas this year.  Would anyone who
knows of relevant projects please send a message to Zavarin@csli,
with a cc to Susi?

Thanks--

Joan
-------

∂12-Sep-85  0323	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:klawe.sjrlvm1%ibm-sj.csnet@csnet-relay.arpa 	talk on razborov's theorem    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  03:23:11 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 12 Sep 85 03:21:45-PDT
Received: from csnet-relay by SU-SCORE.ARPA with TCP; Thu 12 Sep 85 03:21:36-PDT
Received: from ibm-sj by csnet-relay.csnet id ad03900; 12 Sep 85 6:18 EDT
Date: Wed, 11 Sep 85 14:47:28 PDT
From: Maria Klawe <klawe%ibm-sj.csnet@csnet-relay.arpa>
To: aflb.all@su-score.ARPA
Subject: talk on razborov's theorem

SPECIAL TALK: A complete proof of Razborov's theorem on clique functions
 
TIME: 10:30 - 12:30, Thursday, September 19
 
PLACE: MSRI Lecture Hall, MSRI, Berkeley
 
SPEAKER: Noga Alon, Tel-Aviv University (visiting IBM San Jose)
 
ABSTRACT:
A complete and self-contained proof of an improved version of Razborov's
theorem on the monotone circuit complexity of clique functions
will be presented.  The improved lower bound is slightly less than
exponential, whereas Razborov's bound is just slightly more than
polynomial.  This is joint work with Ravi Boppana.
 

∂12-Sep-85  0905	chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept. 17   
Received: from UCB-VAX.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  09:05:08 PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA15013; Thu, 12 Sep 85 08:59:03 pdt
Received: by ucbcogsci.ARPA (5.5/4.48)
	id AA15273; Thu, 12 Sep 85 09:03:50 PDT
Date: Thu, 12 Sep 85 09:03:50 PDT
From: chertok%ucbcogsci@Berkeley (Paula Chertok)
Message-Id: <8509121603.AA15273@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar--Sept. 17

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                    Cognitive Science Seminar -- IDS 237A

          TIME:                Tuesday, September 17, 11:00 - 12:30
          PLACE:               240 Bechtel Engineering Center
          (followed by)
          DISCUSSION:          12:30 - 1:30 in 200 Building T-4

        SPEAKER:          Alan H. Schoenfeld, Education & Mathematics, UCB

        TITLE:            ``Obstacles To Making Sense of  Mathematical
                          Notions,  or, The Transfer Problem Rears its
                          Ugly Head Once Again''

        It can be argued that the fundamental difficulty in  mathemat-
        ics  learning  is the transfer problem.  That is, the power of
        mathematics lies in the potential applicability of  mathemati-
        cal  ideas  to  new situations.  It doesn't matter whether the
        idea is, for example, function, group,  number,  or  triangle.
        Once  any  particular  mathematical  entity  is  recognized as
        belonging to an identified class of objects, everything  known
        about  that  class of objects applies to that entity.  In such
        abstaction resides much of the power and utility of  mathemat-
        ics.   This paper explores some theoretical and some pragmatic
        obstacles to students' abstraction  of  mathematical  notions.
        We  look  at  two  domains,  whole number arithmetic and plane
        geometry.  Some parallels between the two domains  are  drawn,
        to  indicate  that the processes of abstraction are similar in
        both.  In the case of number, we examine a  theoretical  para-
        dox:  the  use of good ``hands on'' manipulatives to help stu-
        dents make sense of base 10 addition and subtraction may  make
        it  harder to understand the nature of ``number.'' In the case
        of geometry, we discuss an empirical  obstacle.   When  things
        are  compartmentalized  in the curriculum, connections that we
        would hope are ``natural'' turn out to be very hard to make.
        --------------------------------------------------------------
        UPCOMING TALKS
        Sept. 24:   Peter Pirolli, Education, UCB
        Oct. 1:     David Rumelhart, Cognitive Science, UCSD
        Oct. 8:     Terry Winograd, Computer Science, Stanford
        Oct. 15:    Ron Kaplan, Xerox PARC
        Oct. 22:    Lotfi Zadeh, Computer Science, UCB
        Nov. 19:    Richard Alterman, Computer Science, UCB
 

∂12-Sep-85  0908	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar--Sept. 17    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  09:08:10 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Thu 12 Sep 85 09:03:36-PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA15013; Thu, 12 Sep 85 08:59:03 pdt
Received: by ucbcogsci.ARPA (5.5/4.48)
	id AA15273; Thu, 12 Sep 85 09:03:50 PDT
Date: Thu, 12 Sep 85 09:03:50 PDT
From: chertok%ucbcogsci@Berkeley (Paula Chertok)
Message-Id: <8509121603.AA15273@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar--Sept. 17

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                    Cognitive Science Seminar -- IDS 237A

          TIME:                Tuesday, September 17, 11:00 - 12:30
          PLACE:               240 Bechtel Engineering Center
          (followed by)
          DISCUSSION:          12:30 - 1:30 in 200 Building T-4

        SPEAKER:          Alan H. Schoenfeld, Education & Mathematics, UCB

        TITLE:            ``Obstacles To Making Sense of  Mathematical
                          Notions,  or, The Transfer Problem Rears its
                          Ugly Head Once Again''

        It can be argued that the fundamental difficulty in  mathemat-
        ics  learning  is the transfer problem.  That is, the power of
        mathematics lies in the potential applicability of  mathemati-
        cal  ideas  to  new situations.  It doesn't matter whether the
        idea is, for example, function, group,  number,  or  triangle.
        Once  any  particular  mathematical  entity  is  recognized as
        belonging to an identified class of objects, everything  known
        about  that  class of objects applies to that entity.  In such
        abstaction resides much of the power and utility of  mathemat-
        ics.   This paper explores some theoretical and some pragmatic
        obstacles to students' abstraction  of  mathematical  notions.
        We  look  at  two  domains,  whole number arithmetic and plane
        geometry.  Some parallels between the two domains  are  drawn,
        to  indicate  that the processes of abstraction are similar in
        both.  In the case of number, we examine a  theoretical  para-
        dox:  the  use of good ``hands on'' manipulatives to help stu-
        dents make sense of base 10 addition and subtraction may  make
        it  harder to understand the nature of ``number.'' In the case
        of geometry, we discuss an empirical  obstacle.   When  things
        are  compartmentalized  in the curriculum, connections that we
        would hope are ``natural'' turn out to be very hard to make.
        --------------------------------------------------------------
        UPCOMING TALKS
        Sept. 24:   Peter Pirolli, Education, UCB
        Oct. 1:     David Rumelhart, Cognitive Science, UCSD
        Oct. 8:     Terry Winograd, Computer Science, Stanford
        Oct. 15:    Ron Kaplan, Xerox PARC
        Oct. 22:    Lotfi Zadeh, Computer Science, UCB
        Nov. 19:    Richard Alterman, Computer Science, UCB
 

∂12-Sep-85  0932	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	FOCS reminders 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  09:32:53 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 12 Sep 85 09:26:31-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Thu 12 Sep 85 09:26:23-PDT
Received: by wisc-rsch.arpa; Thu, 12 Sep 85 10:57:54 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Thu, 12 Sep 85 02:16:17 cdt
Message-Id: <8509120716.AA02792@wisc-crys.arpa>
Received: from csnet-relay.arpa by wisc-crys.arpa; Thu, 12 Sep 85 02:16:11 cdt
Received: from uoregon by csnet-relay.csnet id aa02798; 12 Sep 85 3:00 EDT
Received: by uo-vax3.UONET (4.12/1.0.uoregon)
	id AA10054; Wed, 11 Sep 85 17:35:51 pdt
Date: Wed, 11 Sep 85 17:35:51 pdt
From: Fall Conference Mail <focs%uoregon.csnet@csnet-relay.arpa>
Posted-Date: Wed, 11 Sep 85 17:35:51 pdt
To: theory@WISC-CRYS.ARPA
Subject: FOCS reminders
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;


        Some Reminders for Those Planning to Attend FOCS
                October 21-23, Portland, OR.

1.  The very reasonable room rates at the Marriott reflect a FOCS
    discount of about 45%.  HOWEVER, they are guaranteed only if
    reservations are made 21 days in advance of arrival.  Expect
    to spend a lot more if you have not made such a reservation.

2.  Up to 50 rooms are available at the even more dramatic student
    discount.  It is possible that these may be committed even
    before the 21 day cutoff.  Note that all room occupants must be
    full-time students.

3.  We are maintaining and distributing a roommate request list.  If
    you are interested, fill out and return the following to  
    focs@uoregon.CSNET

       Name ..................................................
       Institution ...........................................
       Phone number ..........................................
       Electronic address ....................................
       Student [ ]    Non-student [ ]
       Desired number of room occupants (2,3,4) .......

4.  Don't forget to compare, or ask your travel agent to compare,
    any quoted airfare with that obtainable with the special FOCS
    discount offered by United Airlines.  Call 800-521-4041 and
    give the FOCS account number 562T.

5.  The DEADLINE for advance registration is September 27th.



∂12-Sep-85  1208	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Next week's PLANLUNCH -- notice change in date, place   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  12:08:38 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Thu 12 Sep 85 12:01:32-PDT
Date: Thu 12 Sep 85 11:57:32-PDT
From: LANSKY@SRI-AI.ARPA
Subject: Next week's PLANLUNCH -- notice change in date, place
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanLehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA

Next week's planlunch will take place on WEDNESDAY at 11AM rather than
the usual Monday.  We will resume Monday seminars after that.  Also
notice the change in room -- EK242 is the old SRI-AI conference room.
We will return to the new conference room after next week as well.
See you there!   -Amy Lansky

------------------------------------------------------------------------

			     SCRUFFY PLANNING

			       David Wilkins
                         SRI International AI Center

		        11:00 AM, WEDNESDAY, September 18
	         SRI International, Building E, Room EK242


Most of the talks in the planlunch series have been about "neat"
research.  While neat researchers know they have a more 
expressive formalism than do planners like SIPE, they have rarely
sat in front of a machine and watched their systems be overwhelmed
by the computational loads even simple representations can impose.

I'll defend scruffy planning (not to replace but to supplement neat planning),
show how SIPE represents a simple 6-room indoor world for Flakey, show 
its solutions and use of computation, describe the kind of hacks needed to
make things run fast, describe pitfalls, and other scruffy stuff like that.
Only a small amount of time will be spent explaining SIPE, depending on 
audience's desire.

-------
-------

∂12-Sep-85  1353	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp%ucbernie@Berkeley   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  13:53:17 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 12 Sep 85 13:49:07-PDT
Received: from UCB-VAX.ARPA by SU-SCORE.ARPA with TCP; Thu 12 Sep 85 13:48:59-PDT
Received: from ucbernie.ARPA by UCB-VAX.ARPA (4.24/5.3)
	id AA21130; Thu, 12 Sep 85 13:47:00 pdt
Received: by ucbernie.ARPA (5.5/5.3)
	id AA08887; Thu, 12 Sep 85 13:51:11 PDT
Date: Thu, 12 Sep 85 13:51:11 PDT
From: karp%ucbernie@Berkeley (Richard Karp)
Message-Id: <8509122051.AA08887@ucbernie.ARPA>
To: aflb.all@su-score.ARPA

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley   642-0143

Tuesday, September 17 2:00
Vijaya Ramachandran
Finding a Minimum Feedback Arc Set in Reducible Graphs Using Maxflows and
Mincuts

Tuesday, September 17  4:00
Avi Wigderson
Ramsey-theoretic Lower Bounds in Parallel Computation

Wednesday, September 18  4:15  60 Evans
Joe Traub
An Introduction to Information-Based Complexity

Thursday, September 19
10:30
Noga Alon
The Monotone Circuit Complexity of Clique Functions
Tuesday, September 24  2:00
Umesh Vazirani
Sampling a Population with a Slightly Random Source

Tuesday, September 24  4:00
Adrian Ocneanu
Volumes of Tubes and Average Loss of Precision

Tuesday, October 1    2:00
Christos Papadimitriou
To Be Announced

Tuesday, October 1   4:00
Gary Miller
A New Method for Parallel Expression Evaluation

Joe Traub's lecture is on campus. All the others will be held in the MSRI
Lecture Hall.

∂12-Sep-85  1422	BARWISE@SU-CSLI.ARPA 	TGBBQ   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  14:21:53 PDT
Date: Thu 12 Sep 85 14:14:25-PDT
From: Jon Barwise <BARWISE@SU-CSLI.ARPA>
Subject: TGBBQ
To: Folks@SU-CSLI.ARPA

CSLI is having a late afternoon TGIF Bar-B-Que tomorrow.  Ask Susi for
more details.  Or just come.  Jon
-------

∂12-Sep-85  1424	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.10  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  14:24:37 PDT
Date: Thu 12 Sep 85 13:14:53-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.10
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS: ;


(Insert file: risks-1.10

-------

∂12-Sep-85  1540	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.10  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  15:40:45 PDT
Date: Thu 12 Sep 85 13:26:05-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: RISKS-1.10
To: RISKS: ;

RISKS-FORUM Digest       Thursday, 12 Sep 1985      Volume 1 : Issue 10

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Weizenbaum, etc.; even if SDI worked.... (John Shore)
  SDI (John McCarthy)
  More on SDI reliability (Martin Moore)

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

To: risks@sri-csl.ARPA
Subject: Weizenbaum, etc.; even if SDI worked....
Date: 12 Sep 85 11:11:13 EDT (Thu)
From: John Shore <epi-dc!shore@nrl-css.arpa>

It's tempting to respond to Weizenbaum by arguing against the general
proposition, "Don't do it if there's a way around it".  After all, should we
refuse to develop bullet proof vests and to equip police officers with them
just because a criminal might approach from behind and stab them in the ass?

Assuming that a proposed defensive system will work, the relevant question
is what is the cost of developing it compared to the cost of getting around
it?

In the case of SDI, one should distinguish between defense against a few
missiles vs. defense against a massive attack.  Either defense would be
enormously expensive to develop.  If the goal of the attacker is to detonate
a few bombs (or threaten to do so), then it is obviously easier and cheaper
to get around SDI than through SDI.  Here, Weizenbaum is probably right.  If
the goal is massive or total destruction (including destruction of our
missile forces), then getting around SDI (assuming SDI works) does not
appear to be either easy or inexpensive.  Here, Weizenbaum is probably
wrong.  In this case, however, the premise is most likely also wrong.

Moreover, suppose that the premise is right -- i.e. SDI works perfectly.  As
Parnas has pointed out, there's no way for anyone to establish this fact,
which shows the absurdity of arguments like "give us SDI and we will
dismantle our missiles".

------------------------------

Date: 12 Sep 85  0057 PDT
From: John McCarthy <JMC@SU-AI.ARPA>
Subject: SDI 
To:   risks@SRI-CSL.ARPA    

	Some remarks of mine about SDI on Stanford BBOARD have been referred
to.  For the benefit of non-readers of that BBOARD, they mainly concerned
whether I, like Chris Stuart, should use the IJCAI platform to say something
about it.  I said nothing in my lecture, but in my press conference, added
to my remarks on AI, the remark that there was no principle of computer
science that says that programs of any particular task cannot be written and
debugged.  Not much interest was shown by the assembled press; there was
exactly one question on that point.

	At the suggestion of Robert Jastrow, who is one of the main
scientific defenders of SDI, I made the same point in letters to three
Congressmen, said to be influential in the matter of SDI appropriations.

	Now I shall say my opinion about SDI.

	1. If it can be done, it should.  If it affords complete protect,
that's great, and if it affords partial protection, that's good.  The
balance of terror is a bad thing.  Here are answers to some counter
arguments to its desirability.  (a) Joe Weizenbaum says that it attempts a
technological solution to a problem that should be solved morally.  Alas,
moral progress has been so slow that almost the only moral problems to be
even partially solved are those that can at least partially been turned into
technological problems.  For example, the technology of contraception has
greatly reduced human unhappiness.  (b) It is argued that the Soviets would
have to attack at the first sign of deployment.  Every past imminent advance
by either side has in principle given the other side some temptation to
strike before it can be deployed.  So far as we know, neither side has even
come close to giving in to such temptation.  One reason is that the effect
of any advance is always subject to a probabilistic estimate, so temporizing
has always looked better than attacking.  Even if SDI works very well, it
may be that no-one will be able to be sure that it is that good.

	However, most likely the main reason has been is that neither side
ascribes the very worst intentions to the other with certainty.  Each side
has always said, "Perhaps they don't actually mean to attack us.  Why have a
nuclear war for sure instead of only a certain probability?"  Anyway the
Soviets have experienced a period in which we had complete nuclear
superiority and didn't attack them.

2. My opinion is that if the physics of the problem permits a good
anti-missile defense the programs can be written and verified.  However, it
will be quite difficult and will require dedicated work.  It won't be done
by people who are against the whole project.  Computer checked proofs of
program correctness will probably play some role.  So will anticipating what
kind of bugs would be most serious and putting the biggest effort into
avoiding them.  Having many people go over and discuss all the critical
parts of the program will also be important.

------------------------------

Date: Tue, 10 Sep 85 13:56:45 CDT
From: mooremj@EGLIN-VAX
Subject: More on SDI reliability
To: risks@sri-csl.arpa
Cc: soft-eng@mit-xx, lin@mit-mc, mooremj@eglin-vax

> From: Herb Lin <LIN@MIT-MC.ARPA>
> My primary complaint about your otherwise interesting table is that it
> assumes independent failure modes.  I think it is much more likely
> that the effects of coupled failures are larger.  In particular, given
> the failure of one platform, it is more likely that more than one will
> fail.

Good point.  My original post did concern only statistically independent
failures.  If I can be forgiven one more table, I'll address coupled failures.

Independent failures are caused by events isolated to a single platform,
e.g., electrical component failures.  The occurrence of such a failure in
platform J does not affect the probability of a similar failure in platform K,
i.e., P(K|J) = P(K|~J) = P(K).

Coupled failures are failures such that the probability of failure is low in
any platform, but is greatly increased in all platforms when it occurs in
any one of them.  For example, consider that a hostile power might develop a 
new method for its missiles to escape detection.  The probability that it will 
fool any one platform may be low; but if it fools one platform it is likely to
fool more than one, perhaps all.  For arbitrary platforms J and K, P(K|J) >>
P(K|~J). 

The original false positive table is not affected by this, since it showed 
the probability that at least one platform would fail.  Coupled failures
do not change that probability, only the probability that if one fails, others 
will (although it is true that while this country might be able to explain
away a single false positive, explaining a whole bunch of them could be a lot 
tougher!)

The false negative case is where the kicker really comes in.  The original
false negative table applies to independent failures.  The following table
is structured similarly, but instead of using the probability of failure (Pn),
it uses the degree of coupling, Pn(K|J).  This table shows, for a 100-platform
system, the probability of various numbers of successful responses, given
that at least one system has experienced a coupled failure.

Pn(K|J):  .5      .6      .7      .8      .9      .95     .99
      +-------------------------------------------------------
N:  0 | 1.0000  1.0000  1.0000  1.0000  1.0000  1.0000  1.0000
    5 | 1.0000  1.0000  1.0000  1.0000  0.9746  0.5550  0.0033
   10 | 1.0000  1.0000  1.0000  0.9973  0.5355  0.0265  0.0000
   15 | 1.0000  1.0000  0.9998  0.9123  0.0677  0.0001  0.0000
   20 | 1.0000  1.0000  0.9896  0.5200  0.0017  0.0000  0.0000
   25 | 1.0000  0.9993  0.8740  0.1204  0.0000  0.0000  0.0000
   30 | 1.0000  0.9822  0.5116  0.0097  0.0000  0.0000  0.0000
   35 | 0.9988  0.8525  0.1465  0.0003  0.0000  0.0000  0.0000
   40 | 0.9781  0.5054  0.0176  0.0000  0.0000  0.0000  0.0000
   45 | 0.8426  0.1574  0.0008  0.0000  0.0000  0.0000  0.0000
   50 | 0.5000  0.0219  0.0000  0.0000  0.0000  0.0000  0.0000
   55 | 0.1574  0.0013  0.0000  0.0000  0.0000  0.0000  0.0000
   60 | 0.0219  0.0000  0.0000  0.0000  0.0000  0.0000  0.0000
   65 | 0.0012  0.0000  0.0000  0.0000  0.0000  0.0000  0.0000
   70 | 0.0000  0.0000  0.0000  0.0000  0.0000  0.0000  0.0000

For example, if the degree of coupling is 0.7 -- that is, if something that
causes failure in one platform has a 70% chance of causing failure in any
other platform -- then the probability is 51.16% that at least 30 of 100
platforms will respond correctly, 14.65% that at least 35 will, and so on,
GIVEN THAT THIS TYPE OF FAILURE OCCURS IN THE "FIRST" PLATFORM.  Don't forget
that the probability that the first platform will fail is UNRELATED to the
probabilities in this table! 

As far as the relative probabilities of independent and coupled failures,
I haven't a clue.  The independent failures are the easiest to get a handle
on through reliability theory; the coupled failures may be the result of
unknown shortcomings in design, or due to unknown hostile actions.  (There
is an old saying that there are always more unknown errors than known errors,
because known errors are limited, but unknown errors are unbounded by 
definition!)
                                            Martin Moore
                                            mooremj@eglin-vax.arpa

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂12-Sep-85  1628	NEUMANN@SRI-CSLA.ARPA 	The foregoing mailings of RISKS-1.10 
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  16:28:06 PDT
Date: Thu 12 Sep 85 15:32:24-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: The foregoing mailings of RISKS-1.10
To: RISKS-LIST@SRI-CSLA.ARPA

I apologize for the almost null message.  I was debugging a macro to
automate the RISKS handling, and the ↑B to include a file did not get
properly escaped.  That character is also used by the macro processor
as an internal break character.  Yuk.  Sorry.  Thanks to those of you
commented on the RISKS of using message systems to run such a forum.

The real RISKS-1.10 was sent out a little later.  I have ascertained that
some of you got the second message, some didn't.  I try never to send
the Forum out during the day, to keep down net overhead, so this was
an exception.  If you haven't gotten it in a while longer, try to FTP
it from SRI-CSL:<RISKS>RISKS-1.10.  If you are gatewaylaid, then send
me a message.  I don't see why some of you should have gotten it and
others not -- unless it is just net delay.

There are several of you who have requested to be added to the list
whose addresses I cannot answer (for various different reasons).  I
generally try to find a net guru who can find a way through, so don't 
give up if you read this second-hand.

Let me take this less formal opportunity, since I hope this message and the
near-null one will subsequently be deleted from your mail boxes, to thank
you all for your messages and contributions.  Running this Forum has been
quite time consuming up until now (I am now largely automated), but very
rewarding.  I am enjoying it immensely.

Peter
-------

∂12-Sep-85  1750	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 12 Sep 85  17:50:33 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 12 Sep 85 17:48:51-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Thu 12 Sep 85 17:48:48-PDT
Received: by wisc-rsch.arpa; Thu, 12 Sep 85 19:19:00 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Thu, 12 Sep 85 17:16:09 cdt
Message-Id: <8509122216.AA14968@wisc-crys.arpa>
Received: from csnet-relay.arpa by wisc-crys.arpa; Thu, 12 Sep 85 17:16:02 cdt
Received: from btl by csnet-relay.csnet id ab08838; 12 Sep 85 18:07 EDT
To: theory%uwisc.csnet-relay.csnet@csnet-relay.arpa
From: dsj%slepian.btl.csnet@csnet-relay.arpa
Date: Thu, 12 Sep EDT 1985 14:10
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

There will be no NP-Completeness Column appearing in the
December issue of J.ALGORITHMS, as I am currently revising
the first 14 or so columns for book publication.

If you have any corrections or updates on these that I might
not know about, please send them to

	David S. Johnson
	Room 2C-355
	AT&T Bell Laboratories
	Murray Hill, NJ 07974

or use computer mail (dsj.btl@csnet-relay or research!dsj).

I also continue to welcome new material.  The column
will resume in the March 1986 issue.

∂13-Sep-85  0151	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.11  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  01:51:38 PDT
Date: Fri 13 Sep 85 00:47:58-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.11
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-FORUM Digest       Friday, 13 Sep 1985      Volume 1 : Issue 11

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  SDI and John McCarthy (Charlie Crummer)
  SDI and Safeguard (John Mashey)
  SDI and Robert Jastrow (Herb Lin)
  Some financial disaster cases from Software Engineering Notes
          (three contributions, totalling five reports)

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Thu, 12 Sep 85 19:00:29 PDT
From: Charlie Crummer <crummer@AEROSPACE.ARPA>
To:   risks@sri-csl
Subject: SDI and John McCarthy

>Date: 12 Sep 85  0057 PDT
>From: John McCarthy <JMC@SU-AI.ARPA
>Subject: SDI 
>To:   risks@SRI-CSL.ARPA    
>
>... there [is] no principle of computer
>science that says that programs of any particular task cannot be written and
>debugged.  

  Computer Science does not contain or deal with the principles operative 
  in the writing and debugging of large, nay, HUGE, eclectic, software 
  programs.  This is the realm of Software Engineering.  By a similar token
  there is no principle of the theory of random processes that says that
  the works of Shakespeare cannot be written by 1,000,000 monkeys pounding
  1,000,000 typewriters either, in fact in principle that would be one way
  of reproducing these works.  No serious student of Shakespeare who knew
  something about random processes would propose such an undertaking, of 
  course.  A mathematician who knew nothing about typewriters and little
  about Shakespeare, however, might if Ronald Reagan pursuaded him that
  the problem should be worked by assigning 1,000,000,000,000 monkeys to
  1,000,000,000,000 typewriters.  In software engineering as well as 
  mechanical engineering there is the concept of feasibility to be considered.


 
>	Now I shall say my opinion about SDI.
>
>If it can be done, it should.

  If you had a gun wouldn't you be more afraid to face a gunman with
  a bullet-proof vest than one without?  If he began deliberately to
  put this vest on as he stood before you with his gun leveled at you
  wouldn't you be inclined to fire before he got the vest on?

>If it affords complete protection that's great, 
>and if it affords partial protection, that's good.

  You speak in the present tense but "it" does not exist!  How can
  a non-existence afford anything?  At least one of the basic questions
  is whether it can be made at all.

>The balance of terror is a bad thing.  

  Yes, and SDI would only enhance the terror.  The "civilized" world has
  no defensive answer to the terrorists in such mundane places as on airliners,
  let alone in space, and such is not in the offing.

>Here are answers to some counter
>arguments to its desirability.  (a) Joe Weizenbaum says that it attempts a
>technological solution to a problem that should be solved morally.  

  MUST be solved between the terrorizer and terrorizee.  When someone's
  out to get you there's no place to hide.  (D. Corleone)

>Alas,
>moral progress has been so slow that almost the only moral problems to be
>even partially solved are those that can at least partially been turned into
>technological problems.  

  Not true, viz. cannibalism and slavery.

>For example, the technology of contraception has
>greatly reduced human unhappiness.  

  What evidence do you have of that?

>(b) It is argued that the Soviets would
>have to attack at the first sign of deployment.  Every past imminent advance
>by either side has in principle given the other side some temptation to
>strike before it can be deployed.  So far as we know, neither side has even
>come close to giving in to such temptation.  One reason is that the effect
>of any advance is always subject to a probabilistic estimate, so temporizing
>has always looked better than attacking.  Even if SDI works very well, it
>may be that no-one will be able to be sure that it is that good.

  You may be safe in saying that but I hope our leaders are not so cavalier.
  Most serious strategy is based on "worst case" scenarios.

>	However, most likely the main reason has been is that neither side
>ascribes the very worst intentions to the other with certainty.  Each side
>has always said, "Perhaps they don't actually mean to attack us.  Why have a
>nuclear war for sure instead of only a certain probability?"  Anyway the
>Soviets have experienced a period in which we had complete nuclear
>superiority and didn't attack them.
>
>2. My opinion is that if the physics of the problem permits a good
>anti-missile defense the programs can be written and verified.  However, it
>will be quite difficult and will require dedicated work.  It won't be done
>by people who are against the whole project.  Computer checked proofs of
>program correctness will probably play some role.  So will anticipating what
>kind of bugs would be most serious and putting the biggest effort into
>avoiding them.  Having many people go over and discuss all the critical
>parts of the program will also be important.
>

  Whether the physics of the problem admits a good anti-missile defense
  is a paramount question.  It will take much more than dedicated climbing
  of the automatic proof of correctness "tree" to get to the "moon" of
  an "astrodome" over the U.S. a la Reagan's definition of strategic
  defense.


  --Charlie

------------------------------

Date: Thu, 12 Sep 85 22:56:02 pdt
From: mips!mash@glacier (John Mashey)
To: Glacier!risks@sri-csl.ARPA
Subject: SDI and Safeguard

I used to work with many of the people at Bell Labs who worked on the
Safeguard ABM; they were competent people who knew how to build complex
systems.  Maybe there were some who believed that it was actually possible
to build a reliable, deployable, maintainable ABM that one could expect to
work in real use; if so, I never met any; most folks did not so believe,
and said so. [They did believe that you could shoot down missiles in
well-controlled tests, because they'd done it; they just didn't believe
it would work when it needed to.]

------------------------------

Date: Thu, 12 Sep 85 20:08:22 EDT
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  SDI and Robert Jastrow
To: JMC@SU-AI.ARPA
cc: LIN@MIT-MC.ARPA, risks@SRI-CSL.ARPA

    From: John McCarthy <JMC at SU-AI.ARPA>

    	At the suggestion of Robert Jastrow, who is one of the main
    scientific defenders of SDI, I made the same point in letters to three
    Congressmen, said to be influential in the matter of SDI appropriations.

Robert Jastrow is certainly a defender of the SDI, but he has admitted
publically in his own Congressional testimony that he does NOT carry
out scientific analyses of anything related to SDI.  He hardly counts
as a "scientific defender."

------------------------------

Date: Fri 13 Sep 85 00:22:19-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Some financial disaster cases from Software Engineering Notes
To: RISKS@SRI-CSLA.ARPA

I hope that the RISKS Forum will not degenerate into only an SDI Forum, so I
thought I would counterbalance this issue with a new topic.  I have
resurrected a contribution from the July 1985 SIGSOFT SEN, and also preview
some newer cases that will appear in the October 1985 SEN (which is just
about ready to go to press).  (The few of you who are ACM SIGSOFT members
please pardon me for the duplications.)

                         -------

     [FROM ACM Software Engineering Notes vol 10 no 3, July 1985]

Disasters Anonymous 1: A Rose is Arose is (Three) Z-Rose

Now and then I get a story that I cannot print.  (I do have a few, but don't
ask.  I have of course conveniently forgotten them all.)  Here, is one that can
be printed -- although its author must remain anonymous.  Note that the case of
the three extra zeroes resulting from two different assumptions about the human
interface bears an eerie resemblance in cause to the case of the shuttle laser
experiment, which follows after this one.  [PGN]

  A group within my company had a policy of dealing only in multiples
  of one thousand dollars, so they left off the last three digits in
  correspondence to the wire transfer area to make their job easier.
  Other groups, however, had to write out the full amount since they did
  not always deal with such nice round numbers.  One day, a transaction
  was processed that had a value of $500,000.  The person who entered the
  transaction thought that it was from the group who dealt in multiples
  of $1000 and entered it as $500,000,000.  Of course, this was not the case,
  so a $500,000 transaction became a $500,000,000 one.

  The only thing that prevented a disaster was that it was sent to a small
  company that called back to verify the amount, and the error was then
  caught.  However, this was a Federal Reserve transaction and the funds
  had been transferred, but the timing was good and the transaction was
  backed out before it became a disaster.  My opinion is that such critical
  software should have caught the error before the wire was sent to the
  Federal Reserve.

  Another error in a Federal Reserve transfer had to do with multiple
  transactions per communications transfer.  In this case, the Federal
  Reserve software put a pair of nulls in the data that should have been
  translated as blanks.  However, they were stripped out and a $200,000,000
  incoming wire lost.  To maintain the Fed balance, money was purchased
  to cover a deficit that didn't exist -- since the money was a credit.
  This was a substantial monetary loss because of inadequately tested
  software.

                         -------

     [FROM ACM Software Engineering Notes vol 10 no 5, October 1985]

Disasters Anonymous 2: Financial Losses

Our anonymous contributor from SEN 10 3 (July 1985) has come through again.

  Since I sent some disaster reports to you in May, another one has occurred.
  This one caused some financial loss and acute headaches among managers.

  Most large banks subscribe to the Federal Reserve's funds transfer system,
  frequently referred to as "Bankwire".  Our system that connects to Fedwire
  was being upgraded with a new DDA interface to the host to help protect
  against overdrafts.  During a review, it was determined that the software
  was not quite ready, but should be okay to put into production two days
  later.  I cautioned them against doing so since not all of the bugs had been
  resolved, and the software had not been "stress tested" (or whatever phrase
  you wish to use about testing that ensures that it will work in production).

  The first day of production went fine.  However, the main file in the new
  software was an ISAM file that had degraded significantly during the first
  day.  On the second day, that file continued to fragment and started to
  consume a large amount of the system resources.  This slowed response time
  so much that by the end of the banking day, we still had hundreds of wires
  to send to the Federal Reserve.  We had to request extensions every half
  hour for hours to try and squeeze the transactions through the system so
  that the money would get to our customers.

  In addition, the response-time problem and other bugs in the software
  prevented us from knowing our Federal Reserve balance.  Since we must
  maintain some 150 million dollars in our Fed "checking account", this lack
  of information could cause significant financial loss as 1.5 billion dolars
  were posted that day and we were off by hundreds of millions of dollars at
  first.

  Another part of this disaster is that the slow response time caused one
  program to assume that the host was down.  When a transaction finally went
  through, our system would transmit the DDA information, but the host did not
  acknowledge that they already had the wire.  Thus a large number of wires
  were being "double posted" (money sent twice).  At the end of the day, tens
  of millions had been double posted.

  As of this writing, the Fed balance had been straightened out, but not all
  of the double postings had been recovered.  Note that at current interest
  rates, a bank loses $350 per day per million dollars of unused money.

                         -------

     [FROM ACM Software Engineering Notes vol 10 no 5, October 1985]

Disasters Anonymous 3: Insurance, Reinsurance, and Rereinsurance

Perhaps anonymity is contagious.  Re: reinsurance, here is
another letter from a different contributor. 

  I'm newly receiving SEN and found the ``war stories'' quite interesting.
  Here are three more.  I would prefer anonymity should you choose to print 
  these.

  This first is hearsay (from a former co-worker).  Apparently he and his
  wife had a joint account with a $300 balance.  They needed $200 in cash, but
  due to miscommunication they both made $200 withdrawals - she at a teller's
  window (cage?) and he at an ATM (automatic teller machine) - within minutes
  of each other.  When the dust settled they found that their account had a
  zero balance:  the first $200 withdrawal left a $100 balance, the second
  should have left a negative balance of $100, but the computer generated a
  $100 credit to offset the shortfall.  The icing on the cake was my friend's
  inability to explain/convince the bank of this situation and have them accept
  restitution.

  I need to be circumspect about this second story -- it might well have
  involved fraud.  While a consultant, I was hired to review a reinsurance
  agreement.  The reinsurance industry is an old-boys, ``handshake is my bond''
  industry as insurors frequently offset their risk by selling it (reinsuring)
  to other insurors.  That is, I insure your building for $10,000,000 and
  re-sell all or part of that risk to another firm.  Apparently, late one
  Monday morning (nearly 11:00 a.m. EST), my client got notice across his
  computer network from another firm that it was reinsuring (i.e. off-loading
  risk) to my client to the tune of several million dollars.  The message was
  time-dated Friday evening (6:00 P.M., WST).  As ``luck'' would have it the
  property in question had suffered a catastrophic loss over the weekend.  The
  bottom line was that the message had been sent directly (not through any of
  the store-and-forward services) and the time-date was thus determined by the
  clock-calendar on the sender's computer.  Need I say more?

  Finally, a story told to me ``out of school'' by a friend at one of the
  nation's largest insurance companies.  They apparently are involved in so
  many reinsurance deals that it turned out that they were reinsuring
  themselves.  I.e., Jones reinsured with Smith who reinsured with Brown who
  reinsured with White who reinsured with Smith.  Smith, it turned out was
  paying both Brown and White commissions for accepting his own risk.  The
  computer system was not designed to look beyond the current customer,
  neglecting the loop.

------------------------------

End of RISKS-FORUM Digest
************************

-------

∂13-Sep-85  0307	NET-ORIGIN@MIT-MC.ARPA 	Enter on BBoard 
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  03:07:48 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 13 SEP 85  06:04:43 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 13 Sep 85 06:05-EDT
Received: from WISCVM.ARPA by MIT-MC.ARPA 13 Sep 85 06:02:50 EDT
Received: from (#D1Z)DDATHD21.BITNET by WISCVM.ARPA on 09/13/85 at
  05:04:04 CDT
Date: 13 Sep 1985, 12:00 CEST
To:  <#D1Z%DDATHD21.BITNET@WISCVM.ARPA> ,
    <PHILOSOPHY-OF-SCIENCE-REQUEST@MIT-MC> ,
    <PHILOSOPHY-OF-SCIENCE@MIT-MC> ,
    <JCMa@MIT-MC>
From:    <#D1Z%DDATHD21.BITNET@WISCVM.ARPA>
Subject: Enter on BBoard

UserID #D1Z     Node DDATHD21     Network EARN (= BITNET)
(Beware of the number sign (#).  It's part of my UserID!)


In the beginning of August, I requested to be entered into the mailing
list concerning philosophy of science.
Up to now I didn't receive any mail concerning this subject, nor did I
receive an acknowledge.

Could anybody tell me something about this bboard?


Mid freundlische Griess/With kind regards

               Wilhelm Mueller

               TH Darmstadt, HRZ, Beratung AB
               Petersenstr. 30
               D-6100 Darmstadt
               Federal Republic of Germany

               Tel +49/6151/16-3050)

∂13-Sep-85  0855	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  08:54:51 PDT
Received: by diablo with Sendmail; Tue, 3 Sep 85 13:03:43 pdt
Date: Tue, 3 Sep 85 13:03:43 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

tomorrow, 11AM.
Shukey will talk about recent work of his that may imply
something about optimal join ordering.

∂13-Sep-85  1117	BETSY@SU-CSLI.ARPA 	New Charge for Specially Processed Checks    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  11:17:38 PDT
Date: Fri 13 Sep 85 11:13:43-PDT
From: Betsy Macken <BETSY@SU-CSLI.ARPA>
Subject: New Charge for Specially Processed Checks
To: folks@SU-CSLI.ARPA



The Stanford Accounting Office has announced that as of October 1
they will charge $7.50 for each check that is "walked through".
In the past, CSLI has had to walk through checks quite frequently
to get them through Stanford's system with just a few days notice
-- usually in relation to operant funds.  Be aware that in the
future we will have to charge your project's operant funds for
checks we have to walk through.
Betsy
-------

∂13-Sep-85  1209	@MIT-MC.ARPA:Ghenis.pasa@XEROX.ARPA 	Is the Phi of Sci list alive?    
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  12:09:50 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 13 SEP 85  15:05:39 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 13 Sep 85 15:06-EDT
Received: from Xerox.ARPA by MIT-MC.ARPA 13 Sep 85 15:01:50 EDT
Received: from PinotNoir.ms by ArpaGateway.ms ; 13 SEP 85 12:03:32 PDT
Date: 13 Sep 85 11:18 PDT
From: Ghenis.pasa@Xerox.ARPA
Subject: Is the Phi of Sci list alive?
In-reply-to: <#D1Z@DDATHD21.BITNET>'s message of 13 Sep 1985, 12:00 CEST
To: #D1Z%DDATHD21.BITNET@WISCVM.ARPA
cc: PHILOSOPHY-OF-SCIENCE-REQUEST@MIT-MC.ARPA,
 PHILOSOPHY-OF-SCIENCE@MIT-MC.ARPA, JCMa@MIT-MC.ARPA
Message-ID: <850913-120332-2639@Xerox>

I have been on the Philosophy of Science list for several months now,
and as far as I can tell it is dead, dead, dead... Hope I'm wrong and
it's just a problem with my link to the list.

If anyone can report any signs of life or other relevant info, please
do.

∂13-Sep-85  1218	YAMARONE@SU-CSLI.ARPA 	TGIF TODAY @ 4:00
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  12:18:42 PDT
Date: Fri 13 Sep 85 12:13:37-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: TGIF TODAY @ 4:00
To: folks@SU-CSLI.ARPA



Folks, 

There will be a TGIF today beginning at 4:00 pm on the patio at CSLI.

All are invited to attend and BYOB-ing is optional. Chips, drinks and 
any other victual delights would be welcomed as the Bar-B-Que mistress
will be preparing CHICKEN for all.  And you all know how much I LOVE
CCCCCCCCCCC   H        H IIIIIII  CCCCCCCCC  K        K  EEEEEEEE  N        N
C             H        H    I     C          K       K   E         N N      N
C             H        H    I     C          K     K     E         N   N    N
C             H        H    I     C          K   K       E         N    N   N
C             HHHHHHHHHH    I     C          KKK         EEEEEEE   N     N  N
C             H        H    I     C          K   K       E         N      N N
C             H        H    I     C          K     K     E         N       NN
C             H        H    I     C          K       K   E         N        N
CCCCCCCCCCC   H        H IIIIIII  CCCCCCCCC  K        K  EEEEEEEE  N        N


So, plan on an afternoon of mingling and merriment.....put your
superstitions aside ......and we'll be seeing you at 4:00 for another grand




    TTTTTTTTTTTTTTTTT       GGGGGGGGGG      IIIII    FFFFFFFFFFFFFF
            T             G            G      I      F
            T            G              G     I      F
            T           G                     I      F 
            T           G                     I      F
            T           G                     I      FFFFFFFFF
            T           G          GGGGGG     I      F
            T            G              G     I      F
            T             G            G      I      F
            T              G          G       I      F
            T               GGGGGGGGGG      IIIII    F


Care of your friends and co-workers at

                                                                                        CCCCCCCC     SSSSSS       LLLLL               IIIII
      C         C   S      S        L                   I
     C             S                L                   I
    C              S                L                   I
    C               S               L                   I
    C                SSSSS          L                   I
    C                     S         L                   I
    C                      S        L                   I
     C                     S        L                   I
      C         C  S      S         L         L         I
        CCCCCCCC    SSSSSS        LLLLLLLLLLLLL       IIIII

Tell a friend and bring them along!!!!

Good food, Good drink, you can't go wrong!!!!

-------

∂13-Sep-85  1302	YAMARONE@SU-CSLI.ARPA 	ONE EXTRA TURKEY SANDWICH:2.50  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  13:02:51 PDT
Date: Fri 13 Sep 85 12:59:19-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: ONE EXTRA TURKEY SANDWICH:2.50
To: folks@SU-CSLI.ARPA



There happens to be one extra turkey sandwich on light rye

available at the front desk.

Don't let this go to waste....

Eat now!!!



The Ventura Sandwich Corp.

IGNORE THIS MESSAGE IF RECEIVED LATER THAN 1:30 PM. THANK YOU!!
-------

∂13-Sep-85  1505	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.12  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  15:05:19 PDT
Date: Fri 13 Sep 85 13:44:10-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.12
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA


RISKS-FORUM Digest       Friday the 13 Sep 1985      Volume 1 : Issue 12

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Wire-Transfer Risks; Risk of Non-application of Technology (Jerry Saltzer)
  Date-Time stamps (and errors therein) (Ted M P Lee)
  JMC's remarks (Joseph Weizenbaum)
  Subjective Factors in Risk Assessment (Lynne C. Moore)
  Moral vs. Technological Progress (Charlie Crummer)

[*** MODERATOR'S NOTE.  SOME OF THE SDI-RELATED DISCUSSION MAY BE
DEGENERATING INTO A NONCONVERGENT ITERATIVE LOOP.  LET'S TRY TO STICK A
LITTLE MORE TO COMPUTER-RELATED ISSUES, ALTHOUGH I RECOGNIZE THAT THE
TECHNICAL ISSUES MAY BE OVERWHELMED BY NONTECHNICAL ISSUES.  BUT PLEASE DO
NOT INTERPRET THIS AS AN ATTEMPT TO SQUELCH MEANINGFUL DISCUSSION.  PGN ***]

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date:  Fri, 13 Sep 85 10:51 EDT
From:  Saltzer@MIT-MULTICS.ARPA
Subject:  Wire-Transfer Risks; Risk of Non-application of Technology
To:  RISKS FORUM <RISKS@SRI-CSL.ARPA> (Peter G. Neumann, Coordinator)

Re: Wire-Transfer Risks

1) The current (September, 1985) issue of IEEE Communications magazine
on page 23 suggests that it may be typical in the wholesale financial
business to carry transactions in thousands of dollars rather than in
(ones of) dollars.  If so, you would think that the people responsible
for software in that business would check and recheck their specs and
the human engineering across those interfaces where dividing or
multiplying by 1000 is a possibility, wouldn't you?

2) The comment that current money prices lead to losses of about $350
per day for each mislaid million dollars seems to have been intended to
suggest that such mistakes are unacceptable.  The people in the
wholesale money movement business draw an opposite conclusion:  since
they can quantify their exposure so precisely, they can decide
rationally when the loss rate has become unacceptable and it is thus
worth paying someone to develop a more error-free system.  (For the
price of a contract to SRI to develop a verified 1000-line program one
could probably afford to mislay IBM's entire revenue stream for a week.)


Re:  Risk of Non-application of Technology

For another economically quantifiable example, the early reports on the
creation of the SABRE airline reservation system by American Airlines
explicitly mentioned a business decision, with two alternatives:  invest
in two more Boeing 707's, or in developing SABRE.  The first approach
provided more spare seat-mile capacity that could thus be managed with
less precision; the second offered the hope of better management of
available seat-mile capacity.  Two other considerations that were
explicitly mentioned were the cost of customer disatisfaction when
reservations were dishonored (accidental overbooking, as contrasted with
intentional overbooking, was a rampant problem at the time) and the cost
to the company in lost revenue if the prospective computer were to go
down for several hours or if the entire contents of the disks were lost.
The decision to develop SABRE thus represents an example of up-front
assessment of the risk of non-application of technology, compared with
the risk of applying it.

------------------------------

Date:  Fri, 13 Sep 85 12:15 EDT
From:  TMPLee@MIT-MULTICS.ARPA
Subject:  Date-Time stamps (and errors therein)
To:  Risks@SRI-CSL.ARPA

It was an interesting coincidence that the latest Risks←Forum had a
piece related to the correctness of the time-stamp on messages.  About
two days ago I logged on late (about half-past midnight, Central time)
and started going through my electronic in-basket.  One of the messages
struck me:  its header was time-stamped 03:56 EDT -- how could I
possibly be reading it two and a half-hours before it was sent?  (yes,
the dates were right -- it wasn't from the previous night/early-AM.)
Eventually got a copy of the original from its author.  The key to the
mystery is that Multics does a time-zone conversion on most (but not
all) time fields in incoming message headers.  The original message's
time-zone was clearly marked as PDT, so multics dutifully added three
hours and gave me the time in EDT.  When we (I and a multics guru) first
looked at just the multics version we speculated that perhaps multics
had taken the message's time-zone as GMT, which I think would have given
the same result.  I also thought perhaps since the original was before
midnight and the result after, that might have been the cause.  In the
process of writing this entry for the Risks forum I looked at the
original message one more time, and it struck me:  for some reason the
ISI clock had been set to run on Eastern Time (00:56) but the ISI mailer
software (or something else there) thought it was keeping Pacific, hence
the PDT tag.  What was further confusing was the fact that I looked at
several other messages from ISI from about the same period (two to four
days ago) and some came out right and some not.  Sounds like a good
ingredient for a mystery story, at least.

------------------------------

Date: Fri 13 Sep 85 12:57:15-EDT
From: Joseph Weizenbaum <JOSEPH@MIT-XX.ARPA>
Subject: JMC's remarks
To: risks@SRI-CSL.ARPA

Contrary to John McCarthy's inference that I hold to the "general
proposition, 'Don't do it if there's a way around it'", I think that
proposition to be (even purely logically) absurd.  The "way around it"
would be another "it" to which the rule would apply, and so on.

Another instance of John putting words in my mouth I didn't (and wouldn't)
utter is his "Joe Weizenbaum says that [SDI] attempts a technological
solution to a problem that should be solved morally".  He makes it easy for
himself to score a point by pointing  to the slowness of "moral progress"
and so on.  I believe I wrote that SDI is a technological fix for a problem
that is primarily political, cultural, economic, and so on, and that it has
to be attacked in these contexts, that we must actually confront the
problem of how peoples who organize their societies differently from one
another can peacefully share the same globe.  That is considerably
different from saying the problem should be "solved morally".  The trouble
with technological fixes is often, and I think in this case, that they give
the impression the problem has been dealt with and that no further efforts
to deal with it are necessary.  In the present instance the spread of such
an impression with respect to the peaceful coexistence of the Western and
the Eastern block nations could be fatal to the whole world.

------------------------------

Date: Fri, 13 Sep 85 13:48:04 CDT
From: moorel@EGLIN-VAX
Subject: Subjective Factors in Risk Assessment
To: RISKS@SRI-CSL.ARPA

	There is a very interesting article about various types of risks and 
the way that people perceive them in the October issue of ←Science←85←. In 
particular, it makes a couple of points that I feel are quite relevant to this
forum's discussions. First, that people respond to risks differently depending
on whether the risk is presented as a positive or a negative risk. "Because 
losses loom larger than gains, we are more willing to gamble to avoid them."
Second, it points out that most people are less concerned and aware of the risksof things over which they feel that they have some control. "If we can't be
certain about the risks we face, we at least want to have some control over the
technologies and activities that produce them."

	When we look for examples of the risks of using computer technology vs.
the risks of not using computer technology, we ought to keep these two ideas in
mind, and ask ourselves whether we are being truly objective about the risks
involved or are we letting other, subjective factors influence our judgement. I
recommend this article for your reading.

	Lynne C. Moore (MOOREL AT EGLIN-VAX.ARPA)

------------------------------

Date:           Fri, 13 Sep 85 10:56:21 PDT
From:           Charlie Crummer <crummer@AEROSPACE.ARPA>
To:             risks@sri-csl
Subject:        Moral vs. Technological Progress

> From: McNelly.OsbuSouth@Xerox.ARPA
> In-Reply-to: NEUMANN%SRI-CSLA:ARPA's message of 13 Sep 85 01:19:23 PDT
>  (Friday)

>>>Alas,
>>>moral progress has been so slow that almost the only moral problems to be
>>>even partially solved are those that can at least partially been turned into
>>>technological problems.  

>>Not true, viz. cannibalism and slavery.

> Actually, it's my understanding that the demise of slavery was due to
> technological advances which made slavery economically unfeasible.  The
> invention of the cotton gin, for example, made it only a matter of time
> here in the US before slavery died out.  As far as cannibalism goes, I'd
> say that was more caused by Western culture steam-rolling over the
> cannibals.

> -- John --

Actually McCarthy's original comment presupposes that moral and technological
progress are comparable.  It is that assumption that I disagree with.  Ethics
and the attendant morality provide the context within which all activity, and
in particular technological progress, exists.  Morality and technology are
not substitutes for one another and moral progress is not dependent on
technology nor vice versa.  There is always technological progress attendant
to moral progress just because there is always technological progress.
  

  --Charlie

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂13-Sep-85  1636	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.12  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 13 Sep 85  16:36:07 PDT
Date: Fri 13 Sep 85 14:28:03-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.12
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA


RISKS-FORUM Digest       Friday the 13 Sep 1985      Volume 1 : Issue 12

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Wire-Transfer Risks; Risk of Non-application of Technology (Jerry Saltzer)
  Date-Time stamps (and errors therein) (Ted M P Lee)
  JMC's remarks (Joseph Weizenbaum)
  Subjective Factors in Risk Assessment (Lynne C. Moore)
  Moral vs. Technological Progress (Charlie Crummer)

[*** MODERATOR'S NOTE.  SOME OF THE SDI-RELATED DISCUSSION MAY BE
DEGENERATING INTO A NONCONVERGENT ITERATIVE LOOP.  LET'S TRY TO STICK A
LITTLE MORE TO COMPUTER-RELATED ISSUES, ALTHOUGH I RECOGNIZE THAT THE
TECHNICAL ISSUES MAY BE OVERWHELMED BY NONTECHNICAL ISSUES.  BUT PLEASE DO
NOT INTERPRET THIS AS AN ATTEMPT TO SQUELCH MEANINGFUL DISCUSSION.  PGN ***]

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date:  Fri, 13 Sep 85 10:51 EDT
From:  Saltzer@MIT-MULTICS.ARPA
Subject:  Wire-Transfer Risks; Risk of Non-application of Technology
To:  RISKS FORUM <RISKS@SRI-CSL.ARPA> (Peter G. Neumann, Coordinator)

Re: Wire-Transfer Risks

1) The current (September, 1985) issue of IEEE Communications magazine
on page 23 suggests that it may be typical in the wholesale financial
business to carry transactions in thousands of dollars rather than in
(ones of) dollars.  If so, you would think that the people responsible
for software in that business would check and recheck their specs and
the human engineering across those interfaces where dividing or
multiplying by 1000 is a possibility, wouldn't you?

2) The comment that current money prices lead to losses of about $350
per day for each mislaid million dollars seems to have been intended to
suggest that such mistakes are unacceptable.  The people in the
wholesale money movement business draw an opposite conclusion:  since
they can quantify their exposure so precisely, they can decide
rationally when the loss rate has become unacceptable and it is thus
worth paying someone to develop a more error-free system.  (For the
price of a contract to SRI to develop a verified 1000-line program one
could probably afford to mislay IBM's entire revenue stream for a week.)


Re:  Risk of Non-application of Technology

For another economically quantifiable example, the early reports on the
creation of the SABRE airline reservation system by American Airlines
explicitly mentioned a business decision, with two alternatives:  invest
in two more Boeing 707's, or in developing SABRE.  The first approach
provided more spare seat-mile capacity that could thus be managed with
less precision; the second offered the hope of better management of
available seat-mile capacity.  Two other considerations that were
explicitly mentioned were the cost of customer disatisfaction when
reservations were dishonored (accidental overbooking, as contrasted with
intentional overbooking, was a rampant problem at the time) and the cost
to the company in lost revenue if the prospective computer were to go
down for several hours or if the entire contents of the disks were lost.
The decision to develop SABRE thus represents an example of up-front
assessment of the risk of non-application of technology, compared with
the risk of applying it.

------------------------------

Date:  Fri, 13 Sep 85 12:15 EDT
From:  TMPLee@MIT-MULTICS.ARPA
Subject:  Date-Time stamps (and errors therein)
To:  Risks@SRI-CSL.ARPA

It was an interesting coincidence that the latest Risks←Forum had a
piece related to the correctness of the time-stamp on messages.  About
two days ago I logged on late (about half-past midnight, Central time)
and started going through my electronic in-basket.  One of the messages
struck me:  its header was time-stamped 03:56 EDT -- how could I
possibly be reading it two and a half-hours before it was sent?  (yes,
the dates were right -- it wasn't from the previous night/early-AM.)
Eventually got a copy of the original from its author.  The key to the
mystery is that Multics does a time-zone conversion on most (but not
all) time fields in incoming message headers.  The original message's
time-zone was clearly marked as PDT, so multics dutifully added three
hours and gave me the time in EDT.  When we (I and a multics guru) first
looked at just the multics version we speculated that perhaps multics
had taken the message's time-zone as GMT, which I think would have given
the same result.  I also thought perhaps since the original was before
midnight and the result after, that might have been the cause.  In the
process of writing this entry for the Risks forum I looked at the
original message one more time, and it struck me:  for some reason the
ISI clock had been set to run on Eastern Time (00:56) but the ISI mailer
software (or something else there) thought it was keeping Pacific, hence
the PDT tag.  What was further confusing was the fact that I looked at
several other messages from ISI from about the same period (two to four
days ago) and some came out right and some not.  Sounds like a good
ingredient for a mystery story, at least.

------------------------------

Date: Fri 13 Sep 85 12:57:15-EDT
From: Joseph Weizenbaum <JOSEPH@MIT-XX.ARPA>
Subject: JMC's remarks
To: risks@SRI-CSL.ARPA

Contrary to John McCarthy's inference that I hold to the "general
proposition, 'Don't do it if there's a way around it'", I think that
proposition to be (even purely logically) absurd.  The "way around it"
would be another "it" to which the rule would apply, and so on.

Another instance of John putting words in my mouth I didn't (and wouldn't)
utter is his "Joe Weizenbaum says that [SDI] attempts a technological
solution to a problem that should be solved morally".  He makes it easy for
himself to score a point by pointing  to the slowness of "moral progress"
and so on.  I believe I wrote that SDI is a technological fix for a problem
that is primarily political, cultural, economic, and so on, and that it has
to be attacked in these contexts, that we must actually confront the
problem of how peoples who organize their societies differently from one
another can peacefully share the same globe.  That is considerably
different from saying the problem should be "solved morally".  The trouble
with technological fixes is often, and I think in this case, that they give
the impression the problem has been dealt with and that no further efforts
to deal with it are necessary.  In the present instance the spread of such
an impression with respect to the peaceful coexistence of the Western and
the Eastern block nations could be fatal to the whole world.

------------------------------

Date: Fri, 13 Sep 85 13:48:04 CDT
From: moorel@EGLIN-VAX
Subject: Subjective Factors in Risk Assessment
To: RISKS@SRI-CSL.ARPA

	There is a very interesting article about various types of risks and 
the way that people perceive them in the October issue of ←Science←85←. In 
particular, it makes a couple of points that I feel are quite relevant to this
forum's discussions. First, that people respond to risks differently depending
on whether the risk is presented as a positive or a negative risk. "Because 
losses loom larger than gains, we are more willing to gamble to avoid them."
Second, it points out that most people are less concerned and aware of the risksof things over which they feel that they have some control. "If we can't be
certain about the risks we face, we at least want to have some control over the
technologies and activities that produce them."

	When we look for examples of the risks of using computer technology vs.
the risks of not using computer technology, we ought to keep these two ideas in
mind, and ask ourselves whether we are being truly objective about the risks
involved or are we letting other, subjective factors influence our judgement. I
recommend this article for your reading.

	Lynne C. Moore (MOOREL AT EGLIN-VAX.ARPA)

------------------------------

Date:           Fri, 13 Sep 85 10:56:21 PDT
From:           Charlie Crummer <crummer@AEROSPACE.ARPA>
To:             risks@sri-csl
Subject:        Moral vs. Technological Progress

> From: McNelly.OsbuSouth@Xerox.ARPA
> In-Reply-to: NEUMANN%SRI-CSLA:ARPA's message of 13 Sep 85 01:19:23 PDT
>  (Friday)

>>>Alas,
>>>moral progress has been so slow that almost the only moral problems to be
>>>even partially solved are those that can at least partially been turned into
>>>technological problems.  

>>Not true, viz. cannibalism and slavery.

> Actually, it's my understanding that the demise of slavery was due to
> technological advances which made slavery economically unfeasible.  The
> invention of the cotton gin, for example, made it only a matter of time
> here in the US before slavery died out.  As far as cannibalism goes, I'd
> say that was more caused by Western culture steam-rolling over the
> cannibals.

> -- John --

Actually McCarthy's original comment presupposes that moral and technological
progress are comparable.  It is that assumption that I disagree with.  Ethics
and the attendant morality provide the context within which all activity, and
in particular technological progress, exists.  Morality and technology are
not substitutes for one another and moral progress is not dependent on
technology nor vice versa.  There is always technological progress attendant
to moral progress just because there is always technological progress.
  

  --Charlie

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂15-Sep-85  0339	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.13  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 15 Sep 85  03:39:28 PDT
Date: Sat 14 Sep 85 23:43:07-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.13
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-FORUM Digest     Saturday, 14 Sep 1985      Volume 1 : Issue 13

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Risks in RISKS (Peter G. Neumann)
  Preserving rights to Email messages (Larry Hunter)
  Risk Comparisons (T. Tussing)
  Risks history/philosophy (Nicholas Spies)       [long but interesting]

[NOTE: RISKS-1.12 went out on Friday the 13th.  Sorry for the double
mailing.  MIC is full of pitfalls.  I would like to believe that my macros
are now REALLY DEBUGGED, but I know better.  However, things will run much
more smoothly on my end from now on, and I hope from yours also.  Thanks for
your patience.  PGN]

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Sat 14 Sep 85 23:18:42-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
To: RISKS@SRI-CSLA.ARPA
Subject: Risks in RISKS 

The text of the message that follows this one is taken verbatim from the
HUMAN-NETS Digest (HUMAN-NETS@RUTGERS), 11 Sep 1985, Volume 8 : Issue 29, on
the topic of risks in EMAIL.  That topic is of vital significance to the
RISKS Forum, for at least two reasons:

  (1) You should recognize the risks that might be incurred by you in 
      submitting messages to this forum, and in sending on-line messages 
      in general.

  (2) The free propagation of messages not copyrighted can itself lead
      to significant risks to the contributor and to the RISKS Forum,
      if those messages were false or libelous, or if they are altered.

In general, you and I must assume that any message on this forum may be
forwarded indefinitely and read by anyone, and could appear in print in all
sorts of strange places.  (If something worthy of inclusion in the ACM
Software Engineering Notes is at all controversial, I ask for explicit
permission before publishing it.)

What is even RISKIER is that your message can be trivially altered along the
way as it traverses the network of networks, possibly drastically changing
your intended meaning.  Use of check sums and crypto seals can reduce the
risk of undetected alteration, but does not solve the problem entirely.

Peter G. Neumann 

(This message is not copyrighted.)

------------------------------

Subject: preserving rights to Email messages.
Date: Tue, 3 Sep 85 11:08:31 EDT
From: Larry Hunter <Hunter@YALE.ARPA>

Copyright-by: Larry Hunter, 1985

After consulting with several lawyer friends, the conclusion I
reach is that anything you send out over the nets is public
property -- ie, anyone can reproduce it verbatim, for profit
and the author has no right to control its use.  There is,
however, a very simple way to preserve the author's rights to
control the use his messages are put to.  The courts have
held that practically any clear attempt of an author to
preserve his/her rights to a written work are sufficient
to actually preserve them.  No need to add the 'circled c'
to ASCII, just add a 'Copyright-by:' line to the header of your
local mailer and voila! your rights are preserved.

                                               Larry

PS. I am not a lawyer and this is only my opinion - if you have
a vital interest in some related matter, talk to a real lawyer!

------------------------------

-------

Date: 13 Sep 85 13:27:12 PDT (Friday)
From: TTussing.es@Xerox.ARPA
Subject: Risk Comparisons
To: RISKS@SRI-CSL.ARPA

Someone sent me this and I thought the people on this mailing list might
be interested.

Excerpt from a pamphlet by Dow Chemical Corp, entitled Life is in the
Balance:

Dr. richard Wilson, who is professor of physics at Harvard, has devised
a mathematical formula that measures risks in terms of the minutes and
seconds of life lost.  Taking the average person of 30 who has a
life-span in the United States of approximately 73 years, Wilson says
that statistical person cuts time from his life in the following ways:

Smoking one cigarette - minus 12 minutes
Drinking a diet soft drink - minus about 9 seconds
Driving without a seat belt - 6 seconds for each trip
Being an unmarried male - minus 1800 days
Being male rather than female - minus 2700 days

We can also view risks by figuring which risks are qeual.  For example,
the following items all pose an equal risk of increasing the likelihood
of death by one chance in a million:

Drinking half a liter of wine
Spending three hours in a coal mine
Living two days in New York
Traveling six minutes by canoe
Riding 10 miles on a bicycle
Driving 300 miles in a car
Flying 1000 miles by jet

------------------------------

Date: 14 Sep 1985 01:06-EST
From: Nicholas.Spies@CMU-CS-H.ARPA
Subject: Risks history/philosophy
To: risks@sri-csl

This is definitely old news, but then again the facts behind the case have
only recently seen the light of day. In his recent biography "Alan Turing:
the enigma" (Simon & Schuster) Andrew Hodges reveals in some detail the
inner workings of the German Enigma encryption device (arguably a
"computer") which contributed (backhandedly) to the development of computers
as we know and love them today. (If you're interested in Turing, computers
or WWII history read it, if you haven't already.)

The portion of the book devoted to Turing's stint at Bletchley Park is
peppered with lost opportunities on the German side.  Apparently with little
additional effort the Germans could have rendered the "Bletchley bombes"
completely useless. In fact the only reason the Germans were not careful was
their unswerving faith in the Enigma device. Even when there was ample
evidence pointing to a message-security problem the integrity of Enigma was
never seriously questioned by the Germans. There were, of course, countless
other factors, but the German faith in their technological answer was in
large measure responsible for their losing the U-boat war and very likely
the war itself.

Another anecdote, not related to computers (recounted in either "The Ultra
Secret", Winterbotham or "Bodyguard of Lies", A. Cave Brown, two other
excellent books on the secret war) gave one reason for the German atomic
bomb project not really getting off the ground.  It seems that a German
professor of the old school was in charge of finding a material for
moderating a chain reaction (by absorbing neutrons). Graphite was tried but
failed which meant that deuterium (heavy water) was thought to be needed.
When it was suggested that the graphite might not be pure enough (which, as
it turned out, was the reason the test failed) the professor reacted with
rage that his authority was being questioned and he effectively derailed
German research in reactor design. (Later the plant for making heavy water
built in Norway was sabotaged by British agents which made a reactor
impossible which preventing the manufacture of fissile material.)

These examples suggest that excessive reliance on either technological
solutions or "authoritative opinion" may carry grave risks, albeit in these
cases for an evil regime. The question facing us is whether we (or the
Soviets, for that matter) have fallen into the same traps. I would say that
we (both) definitely have, for the means to power are now more than ever
technological (as opposed to political or diplomatic) and one or another
"expert" is routinely trotted out to "prove" the efficacy of this or that
technological scheme.

Indeed, how can it be otherwise? Hitler opened the Pandora's Box of applying
high-tech to warfare and it worked (at least until a higher-tech response
prevailed). After WWII a new era was born in which global political power no
longer rested on moral authority but on a command of the new applied
sciences and scientists. Engineers had provided political leaders with
instruments of war for centuries, but now scientists are looked upon as the
fountainhead of political power, by dictators, politicians and the people
alike. It may now be said truly that Knowledge is Power.

To some the risks of technology are the inevitable consequence of the
inexorable "progress" that technology itself represents. It seems to me that
this view places too great an emphasis on the growth of "technology" itself
at the expense of our ability to integrate it with human wants and needs.
It's almost to say that "technology" has a virtual life of its own that we
have no control over. This is manifestly untrue because "technology" is the
mere summation of the creative acts and compulsions of a great number of
people enamored of the idea of "technology". But if "technology" does have a
life of its own it must be based on a willing denial of responsibility on
the part of each member involved in furthering it, particularly when large
populations or the world are put at risk by introducing a new "technological
development". It seems, therefore, self-evident that morality and technology
are intimately interwoven.

In the largest sense, the risks of computers are the risks of having an
increasing number of computer experts that are in a position to tell people
what computers can be safely used for.  Their expert opinions may be well
thought out or erroneous, as the case may be, but they are in fact the only
opinions that the public, financial institutions, military establishments or
politicians can depend on. The fact that any or all may place a value on
this expert information and act on it puts a heavy moral burden on the
providers of this information, whether they like it or not.

The only population that I have had direct contact with who have faced this
primal issue of the risks of technology are the Amish-Mennonites of Ontario;
I made a film about their 150th Anniversary in Canada. (I have also edited
films about the Amish in Lancaster Co., PA.) The trigger for the Amish was
rubber-tired wheels on carriages around the 1870's because this allowed the
young of courting age to "go to town" more easily, with a perceived
disruption of Amish life not far behind. To this "improvement" they said
"No". Subsequently, the Amish have taken great care to keep the increasing
technological developments surrounding them at bay, but not by pure
rejection. In short, they have evaluated the risks of adopting technologies.

For instance, gasoline engines are permitted for stationary use (and also on
horse-drawn wagons) for harvesting, threshing, bailing and for powering milk
refrigerators.  There's no contradiction in using a valuable power source so
long as it isn't applied to providing the means for increased contact with
the outside world. Electricity is used if generated within the farm; and
public telephones may be used as well; as long as wires (i.e. connections)
to the outside world are avoided there is no reason to avoid using the
technology. The oddity of the Amish is based on common sense when their
objectives are known.

Although the Amish reaction to technology may strike many as merely "quaint"
they do show that it is possible to stop short the "inevitable" growth of
technology. (The Amish are renowned for their success in farming, which is
not the case for many others that have embraced modern technological ways.)

I am not advocating a general return to Amish ways (indeed this only makes
sense within the context of Amish values), but I will say that we all face a
similar confrontation with a technology that may do us great harm on many
fronts.  Unfortunately we are prone to treat our own creations (be they
buildings, cars, chemicals or computers) as if they are as benevolent as the
products of 5 billion years of co-adaptive evolution. As increasingly
complex and interdependent as our creations become, the more they will
reveal themselves as ill-suited to the tasks they were meant to perform; it
only stands to reason because of the lack of a truly global feedback in the
design process. And also, how are we to judge the efficacy of our machines
if we have lost sight of the reason we have created them?

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂15-Sep-85  1200	CS.AVI@R20.UTEXAS.EDU 	PODS call for paper   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 15 Sep 85  11:58:45 PDT
Received: from R20.UTEXAS.EDU by su-aimvax.arpa with Sendmail; Sun, 15 Sep 85 11:58:46 pdt
Date: Sun 15 Sep 85 13:37:13-CDT
From: CS.AVI@R20.UTEXAS.EDU
Subject: PODS call for paper
To: nail@SU-AIMVAX.ARPA

 
                      PRELIMINARY CALL FOR PAPERS
                         (dates are tentative) 
                
                 Fifth ACM SIGACT-SIGMOD Symposium on
                     PRINCIPLES OF DATABASE SYSTEMS

               Cambridge, Massachusetts, March 24-26, 1986
 
The  conference  will cover new developments in both the theoretical and
practical aspects of database  systems.    Papers  are  solicited  which
describe   original   and  novel  research  about  the  theory,  design,
specification,  or  implementation  of  database   systems   and   query
languages.
 
Some   suggested,  although  not  exclusive,  topics  of  interest  are:
artifical intelligence  for  databases,  concurrency  control,  database
design,  database  security, data models, data structures for databases,
dependency theory,  distributed databases,  file organization, logic for 
databases, performance evaluation of database systems,  query languages, 
and schema design.
 
You  are invited to submit nine (9) copies of a detailed abstract (not a
complete paper) to the program chairman:

                     Avi Silberschatz
                     Department of Computer Sciences
                     The University of Texas at Austin
                     Austin, TX  78712
                     (512) 471-4353
                     ARPANET: avi@utexas-20
 
Submissions will be evaluated on the basis of significance, originality,
and overall quality.  Each abstract should 1) contain enough information
to enable the program committee to identify the main contribution of the
work;  2)  explain  the  importance  of  the  work - its novelty and its
practical or  theoretical  relevance  to  database  management;  and  3)
include   comparisons   with  and  references  to  relevant  literature.
Abstracts should be no longer than ten double-spaced pages.
 
                     Program Committee:

      Francois Bancilhon           Christos Papadimitriou 
      Hector Garcia-Molina         Ed Sciore
      Jim Gray                     Avi Silberschatz  
      Alberto Mendelzon            Moshe Vardi
      Meral Ozsoyoglu                         
 
The deadline for submission of abstracts is October 11,  1985.    Authors
will  be  notified  of acceptance or rejection by December 6, 1985.  The
accepted papers, typed on special  forms,  will  be  due  at  the  above
address  by  January  10,  1986.  All authors of accepted papers will be
expected  to  sign  copyright  release  forms.    Proceedings  will   be
distributed  at  the  conference, and will be subsequently available for
purchase through ACM.
 
    General Chairman:                     Local Arrangements Chairman:
 
    Ashok K. Chandra                      Arvola Chan
    IBM T.J. Watson Research Center       Computer Corporation of America
    P.O. Box 218                          4 Cambridge Center
    Yorktown Heights, NY 10598            Cambridge, MA 02142.
    (914) 945-1752                        (617) 492-8860
 

-------

∂15-Sep-85  1251	@MIT-MC.ARPA:7thSon@SCRC-STONY-BROOK.ARPA 	philosophy of science 
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 15 Sep 85  12:51:21 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 15 SEP 85  15:49:52 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 15 Sep 85 15:45-EDT
Received: from SCRC-STONY-BROOK.ARPA by MIT-MC.ARPA 15 Sep 85 15:45:13 EDT
Received: from GOONEY.SCRC.Symbolics.COM by SCRC-STONY-BROOK.ARPA via CHAOS with CHAOS-MAIL id 313944; Sun 15-Sep-85 15:41:30-EDT
Date: Sun, 15 Sep 85 15:45 EDT
From: Christopher Garrigues <7thSon@SCRC-STONY-BROOK.ARPA>
Subject: philosophy of science
To: phil-sci@MIT-MC.ARPA
cc: 7thSon@SCRC-STONY-BROOK.ARPA
Message-ID: <850915154511.4.7THSON@GOONEY.SCRC.Symbolics.COM>

I'd like to be added to the mailing list, please.


∂15-Sep-85  1307	@MIT-MC.ARPA:7thSon@SCRC-STONY-BROOK.ARPA 	philosophy of science 
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 15 Sep 85  13:06:55 PDT
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 15 SEP 85  15:55:26 EDT
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 15 Sep 85 15:50-EDT
Received: from SCRC-STONY-BROOK.ARPA by MIT-MC.ARPA 15 Sep 85 15:49:10 EDT
Received: from GOONEY.SCRC.Symbolics.COM by SCRC-STONY-BROOK.ARPA via CHAOS with CHAOS-MAIL id 313945; Sun 15-Sep-85 15:42:45-EDT
Date: Sun, 15 Sep 85 15:46 EDT
From: Christopher Garrigues <7thSon@SCRC-STONY-BROOK.ARPA>
Subject: philosophy of science
To: phil-sci@MIT-MC.ARPA
cc: 7thSon@SCRC-STONY-BROOK.ARPA
Supersedes: <850915154511.4.7THSON@GOONEY.SCRC.Symbolics.COM>
Message-ID: <850915154622.5.7THSON@GOONEY.SCRC.Symbolics.COM>

[whoops, sent to the wrong place.  so sorry.]


∂16-Sep-85  0135	@SUMEX-AIM.ARPA:SAMUEL@SU-SUSHI.ARPA 	SIGBIG 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Sep 85  01:35:24 PDT
Received: from SU-SUSHI.ARPA by SUMEX-AIM.ARPA with TCP; Mon 16 Sep 85 01:37:06-PDT
Return-Path: <@SU-SCORE.ARPA:welch@ames-vmsb.ARPA>
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Sat 14 Sep 85 13:07:39-PDT
Received: from su-aimvax.arpa by SU-SCORE.ARPA with TCP; Sat 14 Sep 85 13:05:14-PDT
Received: from ames-vmsb.ARPA by su-aimvax.arpa with TCP; Sat, 14 Sep 85 13:05:46 pdt
Date: 14 Sep 85 12:53:00 PST
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: bayboards@diablo.ARPA
Reply-To: welch@ames-vmsb.ARPA
ReSent-Date: Mon 16 Sep 85 01:33:35-PDT
ReSent-From: Sam Hahn  <Samuel@SU-SUSHI.ARPA>
ReSent-To: aap@SUMEX-AIM.ARPA

Reply-To: "DYMOND, KEN" <dymond@nbs-vms>




	    NBS PARALLEL COMPUTER BENCHMARK COLLECTION



     The National Bureau  of Standards, since  its  founding, has been 
concerned with measurement, determining the precise values and metrics
for physical phenomena. The NBS has also made significant contri-
butions to metrology in numerous scientific and engineering
disciplines. In this tradition, the MPC (Measurement for Parallel
Computing) project at NBS is developing a set of metrics  and measure-
ment techniques to characterize the performance of parallel processing
systems.  As part of that effort, NBS is collecting benchmarks and code
kernels that represent a variety of applications which are candidates
for parallel processing.  NBS solicits benchmark codes and kernels from
researchers and scientists.  Programs which are computationally
intensive, I/O intensive, vectorizable or not and from non-numeric as
well as from numeric application areas are requested.  Especially
welcome are programs which have been used to produce timing or speedup
data on parallel computers, whose measurement results have been or may
be published in the technical literature, and which are in some fairly
widely used and higher-level programming  language such as FORTRAN,
"C", LISP, Ada, etc.

  
Contributions or inquiries should be directed to:

	Measurement for Parallel Computing
        Institute for Computer Sciences and Technology
	Materials Building MS B364
	National Bureau of Standards
	Gaithersburg, MD 20899 USA

	Telephone: 301-921-3274
	ARPANET: MEASURE@NBS-VMS

------
------

∂16-Sep-85  1004	EMMA@SU-CSLI.ARPA 	CSLI talk  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 16 Sep 85  10:04:44 PDT
Date: Mon 16 Sep 85 10:03:05-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: CSLI talk
To: friends@SU-CSLI.ARPA
Tel:  497-3479

                              CSLI TALK
                          ``Proto-Language''
           C. B. Martin, Dept. of Philosophy, U. of Calgary
           Friday, Sept. 20, 2:15, Ventura Conference Room

C. B. Martin has done a lot of important work in the philosophy of
religion, but also he wrote an extremely important paper on memory
with Max Deutscher, defending a causal theory of memory when this was
quite unfashionable.  I think this paper played an important role in
setting the stage for causal theories of reference and action.  Martin
is now doing work that sounds extremely interesting to me on the
semantics of non-verbal behavior.  This following paragraph from his
paper gives a good indication of what it is about.

	"The time is long overdue for the recognition of the semantic
	import of non-verbal behaviour.  Such behaviour is procedural
	and projective for an outcome (that may or may not have
	satisfaction).  Though "true" and "false" may be reserved for
	the verbal cases, there is a basic rightness and wrongness
	about the non-verbal behavioural, procedural, projective
	representations.  Such behaviour is formed in inter-related
	patterns strikingly and importantly analogous to that of
	verbal language.  I shall call such semantic non-verbal
	behaviour "proto-language"."


					     --John Perry


-------

∂16-Sep-85  2057	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp%ucbernie@Berkeley   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 16 Sep 85  20:57:23 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Mon 16 Sep 85 20:52:39-PDT
Received: from UCB-VAX.ARPA by SU-SCORE.ARPA with TCP; Mon 16 Sep 85 20:52:38-PDT
Received: from ucbernie.ARPA by UCB-VAX.ARPA (4.24/5.6)
	id AA06623; Mon, 16 Sep 85 14:36:05 pdt
Received: by ucbernie.ARPA (5.5/5.3)
	id AA10496; Mon, 16 Sep 85 12:18:41 PDT
Date: Mon, 16 Sep 85 12:18:41 PDT
From: karp%ucbernie@Berkeley (Richard Karp)
Message-Id: <8509161918.AA10496@ucbernie.ARPA>
To: aflb.all@su-score.ARPA

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley  642-0143
SEMINAR ANNOUNCEMENTS
Thursday, September 19  4:00  MSRI Lecture Hall
Neil Immerman
Expressibility and First-Order Reductions

Tuesday, October 1   2:00   MSRI Lecture Hall
Christos Papadimitriou
How Easy is Local Search?

∂16-Sep-85  2307	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.14  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 16 Sep 85  23:07:44 PDT
Date: Mon 16 Sep 85 21:54:25-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.14
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-FORUM Digest       Monday, 16 Sep 1985      Volume 1 : Issue 14

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Pitfalls of a Fail-Safe Mail Protocol? (Peter G. Neumann)
  Some Ruminations on an Ideal Defense System (Bob Estell)
  SDI, feasibility is irrelevant (Gopal)

Note on contributions:
  Let me remind you all that your contributions to RISKS should be objective,
  nonpolitical, nonpersonal, and on the subject of the Forum.  Some of you
  have been drifting away from the charter.  In fact, I rejected a message 
  from John McCarthy that seemed to violate the nonpolitical criterion,
  although in hindsight that was probably a subjective decision.   John 
  accepted that decision, but suggested that you could communicate 
  privately with him if you were curious.  PGN

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Mon 16 Sep 85 20:25:57-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Pitfalls of a Fail-Safe Mail Protocol?
To: RISKS@SRI-CSLA.ARPA

After reading the case of the double posting of hundreds of millions of
dollars in RISKS-1.11, some of you apparently experienced the multiple
posting of RISKS-1.13 -- all original mailings time-stamped Sat 14 Sep 85
23:43:07-PDT, all complete, and all identical.  Brint Cooper, for example,
received THREE identical copies at various time intervals.

  Received: from brl-aos.arpa by TGR.BRL.ARPA id aa20274; 15 Sep 85 3:23 EDT
  Received: from sri-csl.arpa by AOS.BRL.ARPA id a017885; 15 Sep 85 3:18 EDT

  Received: from brl-aos.arpa by TGR.BRL.ARPA id a021160; 15 Sep 85 4:47 EDT
  Received: from sri-csl.arpa by AOS.BRL.ARPA id a018065; 15 Sep 85 4:35 EDT

  Received: from brl-aos.arpa by TGR.BRL.ARPA id a022055; 15 Sep 85 6:31 EDT
  Received: from sri-csl.arpa by AOS.BRL.ARPA id a018257; 15 Sep 85 6:17 EDT

The new ARPANET message protocols are supposed to be fail-safe (never losing
a message, retrying sensibly after a failure, and eventually returning
undeliverable mail with a NAK), network-efficient (transmitting single
copies to each host and letting that host redistribute), and reliable (never
garbling a message) -- although they cannot guarantee message authenticity
in the presence of tampering.

After consulting with my Foonly-TOPS-20 gurus (Geoff Goodfellow, Mark
Lotter, and Dwight Hare), we discovered that just after I mailed RISKS-1.13
late Saturday night, Foonly's David Poole brought our system down and up
(several times?) after midnight PDT for installation of a 100% memory
increment.  Each time he brought it back up, the mailer seems to have
restarted sending some of the previously queued messages -- how many I do
not know.  It is NOT SUPPOSED TO DO THAT, but that is the most plausible
explanation at the moment.  If you again receive multiple copies, please
remail them back in their entirety to Geoff@SRI-CSL (and we'll hope that you
don't blow his mailbox).  Geoff insists the protocol is sound, so let's let
him in on the glitch-hunt!  Meanwhile, we'll follow a bunch of possibilities.

Analogous to the time-out problem described in RISKS-1.11 causing
double-posting of millions of dollars, a similar problem can arise in the
ARPANET if a receiving host does not acknowledge soon enough, causing a
time-out and retry -- even though the message was successfully received.
This has been known to happen, but probably not repeatedly, as in the
RISKS-1.13 duplication.

Sorry for the inconvenience.  We will take several measures to try to
prevent a recurrence, but if there is a real bug in the protocols or in
their implementation, then we will just have to slug it out until it is found.

Peter

------------------------------

Date: Mon, 16 Sep 85 12:21:52 PDT
From: estell@NWC-143B
Subject: Some Ruminations on an Ideal Defense System
To: risks@sri-csl.arpa

SOME RUMINATIONS on AN IDEAL DEFENSE SYSTEM	         16 SEP 85 [RGE]

WHAT ARE THE CHARACTERISTICS OF AN IDEAL DEFENSE SYSTEM?
(Never mind that it does NOT exist - and likely will not in my lifetime.)

1. Useful ONLY on the defensive; incapable of use as an offensive weapon.
2. Effectiveness: 100%.
3. Reliability: 100%.
4. Cost: Cheap.  So economical to manufacture and deploy that everyone who
   needs one (or two) could have it.
5. Simple to use; requires no training; requires no "technology base."
6. Easy to maintain; essentially never wears out, or breaks.
7. Side effects of use or possession: None.

EXAMPLES OF SYSTEMS THAT DO NOT MEET THE IDEAL REQUIREMENTS.

1. Planes, ships, tanks, guns, and bombs - of any size.
2. Chemicals, gas, germs, etc.

EXAMPLES OF SYSTEMS THAT MIGHT - or might not - SUFFICE.
(They clearly won't satisfy all the above requirements.)

1. SDI - whatever that turns out to be, a few billion $ from now.
2. MX - NOT the expensive "new" systems recently voted, but simple
   modifications of older, reliable, affordable technology.  The key
   is the nature of the modifications.  This isn't the place to say more.

WHY THIS PROPOSAL IS WORTH LOOKING AT.

1. The technology is mature, affordable, and operational.
2. The super-powers can USE this proposal, this year, or next.
   Not just the USA and the USSR; but a few other nations, too.
3. The "trouble makers" who haven't yet demonstrated the maturity to
   abstain from rash use of a "doomsday" device can't use this idea.
4. It saves money; and buys time - precious time, in which we can
   learn to trust each other, and to respect our differences -
   accepting the fact that "... east is east and west is west,
   and never the twain shall meet ..." [Kipling]

WHY THIS PROPOSAL WILL BE RESISTED - perhaps EVEN DISCARDED.

1. Not many vendors will win multi-year, multi-billion contracts to
   develop and manufacture these systems.  That has largely been done.
2. Not many career officers and bureaucrats will get multiple promotions
   for the development and deployment of these systems; again, that's done.

SOME CONFESSIONS ABOUT MY MOTIVES.

I believe that Americans and Russians, Christians, Moslems, Jews, Buddhists,
atheists, agnostics, capitalists, communists, socialists, pacifists, hawks,
doves, owls, and others CAN live together; not in "harmony" - but in some
civility, with respect.  My belief is based on two facts:

1. The lion and the lamb may never lie down together, but the lion and the
   antelope presently share the veldt; blood is shed, but genocide does not
   occur.  It's elsewhere called the "balance of nature."  There is even some
   evidence that the process improves the strain of antelope - and lion.

2. I cannot accept gross demeaning judgments made about a people who produce
   some of the world's really great classical music.  Tchaikovsky, Borodin,
   and others have composed the best - in my opinion.  I believe that the
   Russians love their country, their families, their music, and the Almighty
   - by whatever name He may be called - just as we do; and, like us, they fear
   (or at least don't enjoy the thought of) death, starvation, disease, hatred,
   suspicion, etc.  Why are they so "defensive" - an enigma wrapped in a
   mystery, surrounded by a riddle? (or whatever exact phrase Churchill
   turned in '48)  Their land was invaded twice in the lifetime of their most
   recent leaders - men born before WWI.  The losses of life were catastrophic;
   almost 25% of the adult male population died each time - though the WWI 
   figures get blurred with the totals from their own revolution.  And those 
   weren't the first times; Napoleon did it in the 19th century, and the 
   Mongols did it earlier.  That plus the oppressive cold of a lingering 
   winter give them a somewhat different world view.  
   But in a quarter century, they, like us, have NOT pushed the button.  That 
   speaks well for their responsibility.

Our REAL problem - in the USA and the USSR and the third world - indeed in the
entire world - is NOT communists, not Arabs, not any organized government, nor
religion; it is terrorists - criminals; men of desperation, who when armed will
extract their needs by violence, and damn the consequences.  We have more than
our share of such trouble-makers; they have attacked many of our prominent
citizens recently: John and Robert Kennedy, Martin Luther King, Jr., Ronald
Reagan, Patty Hearst, and Malcolm X are examples easily remembered.

If some of the money spent now on weapons can be re-invested in research in 
medicine, in electronics for space exploration, in agriculture, etc. there will
be no loss of income for the scientists and business men now supplying the 
Pentagon.  True, many of them will have to adapt; but then, they do that now.
The arms of WWII just don't sell today.  A Technology Base that supports 
sophisticated weapons to kill people can adapt to kill viruses; to guide 
rockets to the distant galaxies, instead of missiles to distant cities; to 
detect bombs of terrorists, instead of bombs of military commandos.  I will 
have to be one of those who adapt; my employer, the Naval Weapons Center, must 
adapt too.  We are intellectually capable of doing that.

   RGE≠

------------------------------

Date: 16 Sep 85 11:09:56 PDT (Monday)
Subject: Re: SDI, feasibility is irrelevant (Response to McCarthy)
To: RISKS@SRI-CSL.ARPA
From: Gopal <Gopal.es@Xerox.ARPA>

	Your analogy to SDI of a duelling gunman putting on a bullet-proof vest
seems slightly flawed, but your overall point is well taken:
 
	The vest is far from being bullet proof...and the vest dose not
exist yet. The gunman has just thought about making one, after many
fruitless years of holding guns at each other. It is likely that he may not
succeed in making a vest that stops any bullet. The other gunman knows this.
	
	 But IF he should succeed, THEN, the status quo is broken immediately:
the very fact of possession of a bullet proof vest indicates that he can
turn the outcome decisively in his favor. This clearly will be perceived
as an act of hostility by the worst-case-strategists at the Kremlin. If
I were them, I would not sit by idly, twiddling my thumb: I would launch.
	 
	Sure, the gunman trying to make the vest has promised to give one to
the other gunman also, so there would be no hostility. But, WHEN he has
one, he would not give it away. The strategists do not plan on the basis
of goodwill. They would wait until a 'crisis' develops that forces them
to make a hard choice: share SDI or die.
	
	Would they get rid of the guns, once each one has a vest to stop
bullets? Not likely. We all know that when he was the only gunslinger in
town, he opted to keep that gun in case the other guy gets hold of
one...but now he can't get rid of his gun BECAUSE the other guy has one
too! If there is enough goodwill to get rid of the guns with vests, it
should be even more feasible to get rid of them now, without having to
incur the cost of the vests. Such goodwill simply does not exist while
there is also fear.   
	   
	If SDI should succeed, even partially, (that is, if the monkey should
reach the moon, at least partially, by climbing the tree) then the only
option would be to share SDI with the other guy.  Even after sharing
SDI, we are still bound by fear, of a different kind: what if one
develops a way to take out the other guy's SDI somehow by first strike
and tip the scales in his favour? Another president will come along to fund
this project, because the other side "already started on it and we don't
want to compromise national security". The other side, of course, needs
no excuses like this. And, history will repeat...and repeat.
	
	SDI does not solve the basic nature of the conflict in purpose that
national leaders must exhibit: They must get rid of the guns to make
peace and rid their people of fear, and they must protect national
security. These are clearly not compatible goals in a world of hostile
Governments. The Governments of the world hold the people hostage to
fear and insecurity, much like a handful of gun-slinging gangsters
holding a vast majority hostage to fear of crime.
	
	I admit that it is much easier to shoot down an idea (like SDI) than to
offer an easy alternative in a complex problem like this. But let this
not be a reason to adopt a faulty approach like SDI. 
	
	Well, 'nuff  ramblin' on SDI. By the way, October issue of Science '85
magazine has a cover story on "RISKS: How it is perceived" and I thought
it was quite thought-provoking and may help us understand how
SDI-like-risk is perceived by the people. I highly recommend reading it.
	
Gopal

        [This message is marginally on the subject -- assuming the original
         message was.  I'm not sure whether it sharpens or blunts the
         would-be analogy, but let's blow the whistle on the bullet-proof 
         vest at this point.  Thanks.  PGN]

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂16-Sep-85  2355	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #15  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Sep 85  23:55:10 PDT
Date: 16 Sep 85 2342-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #15
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Tuesday, 17 Sep 1985      Volume 1 : Issue 15

Today's Topics:

                       Quickly computing quarks
              NBS Parallel Computer Benchmark Collection

----------------------------------------------------------------------

[From HUMAN-NETS Digest   V8 #27]

Date: 19-Aug-85 10:20 PDT
From: William Daul / McDonnell-Douglas / APD-ASD
From: <WBD.TYM@OFFICE-2.ARPA>
Subject: Quickly Computing Quarks (Science News, VOL. 128)
To: physics@sri-unix.arpa

...news of at least one IBM research effort in high-speed computing
surfaced at last month's National Computer Conference in Chicago.  A
team of physicists will soon take over a specially built computer
designed to solve a single physics problem.  According to an IBM
official, this computer is supposed to take less than a year to solve
a problem that would take a CRAY-1 supercomputer more than 300 years
to do.

The IBM machine, developed at the Thomas J. Watson Research Center in
Yorktown Heights, N.Y., consists of an array of 576 processors, each
one capable of 20 million "floating point" operations per second
(equivalent to multiplying two decimal numbers 20 million times).  In
contrast, a typical personal computer performs 1,000 or so such
operations per second.  When all the processors are working in
parallel, each one handling a small part of a computation, the IBM
computer can handle more than 10 billion floating point operations per
second.

The machine will be used to calculate the mass of a proton from "first
principles," applying quantum chromodynamics theory.  This year-long
exercise should give physicists some clues as to the validity of
their concepts about quarks and gluons.  Once this project is over,
the machine could be used for other purposes, says IBM's George Paul.
And the computer's design team is already thinking about how to extend
the ideas they developed for the original machine.


------------------------------

[Forwarded by Sam Hahn  <Samuel@SU-SUSHI.ARPA>]

Date: Saturday, 14 September 1985  13:53-PDT
From: welch at ames-vmsb.ARPA
Reply-To: welch at ames-vmsb.ARPA
To:   bayboards at diablo.ARPA
Re:   SIGBIG
Reply-To: "DYMOND, KEN" <dymond@nbs-vms>


	    NBS PARALLEL COMPUTER BENCHMARK COLLECTION

     The National Bureau of Standards, since its founding, has been
concerned with measurement, determining the precise values and metrics
for physical phenomena.  The NBS has also made significant
contributions to metrology in numerous scientific and engineering
disciplines. In this tradition, the MPC (Measurement for Parallel
Computing) project at NBS is developing a set of metrics and
measurement techniques to characterize the performance of parallel
processing systems.  As part of that effort, NBS is collecting
benchmarks and code kernels that represent a variety of applications
which are candidates for parallel processing.  NBS solicits benchmark
codes and kernels from researchers and scientists.  Programs which are
computationally intensive, I/O intensive, vectorizable or not and from
non-numeric as well as from numeric application areas are requested.
Especially welcome are programs which have been used to produce timing
or speedup data on parallel computers, whose measurement results have
been or may be published in the technical literature, and which are in
some fairly widely used and higher-level programming language such as
FORTRAN, "C", LISP, Ada, etc.


Contributions or inquiries should be directed to:

	Measurement for Parallel Computing
        Institute for Computer Sciences and Technology
	Materials Building MS B364
	National Bureau of Standards
	Gaithersburg, MD 20899 USA

	Telephone: 301-921-3274
	ARPANET: MEASURE@NBS-VMS

------------------------------

End of PARSYM Digest
********************

∂17-Sep-85  0027	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #16  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  00:27:53 PDT
Date: 16 Sep 85 2359-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #16
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Tuesday, 17 Sep 1985      Volume 1 : Issue 16

Today's Topics:

               PARSYM Special Issue: IJCAI-85 Abstracts

----------------------------------------------------------------------

[This special issue of PARSYM consists of abstracts of papers
presented at IJCAI-85, the Ninth International Joint Conference on
Artificial Intelligence, 18-23 August 1985.  The abstracts included
are from papers particularly relevant to parallel symbolic computing.
The papers are published in the Proceedings, available from Morgan
Kaufmann Publishers, 95 First Street, Los Altos, California 94022.
-- BD]


The Intelligent Channel: A Scheme for Result Sharing in Logic Programs

                     Simon Kasif and Jack Minker
                        University of Maryland

The separation of logic and control in logic programs allows the
programmer to write programs whose execution is determined by the
interpreter.  This characteristic of logic programs spurred research
towards diversifying the means for controlling the execution of logic
programs, and towards understanding and exploring the value of
parallelism in logic programming.  Much of these efforts belong to the
study of AND/OR-parallelism, i.e., parallel execution of
conjunctive/disjunctive goals respectively.

A brute force approach to AND/OR-parallelism, i.e., executing every
possible subgoal in a conjunctive goal, has two major drawbacks:
combinatorial explosion of processes and the need for many active
binding environments.  The latter arises due to the interaction of
shared variables in conjunctive goals.

We propose a scheme to alleviate the above difficulties.  For
simplicity, we demonstrate our approach with examples where atoms
share a single variable.  The approach is applicable to Horn-clause
logic programs.

------------------------------

    The Architecture of the FAIM-1 Symbolic Multiprocessing System

                             A. L. Davis
                            S. V. Robison

                   Schlumberger Palo Alto Research

The FAIM-1 is an ultra-concurrent symbolic multiprocessor which
attempts to significantly improve the performance of AI systems.  The
system includes a language in which concurrent AI applications programs
can be written, a machine which provides direct hardware support for
the language, and a resource allocation mechanism which maps programs
onto the machine in order to exploit the program's concurrency in an
efficient manner at run-time.  The paper provides a brief synopsis of
the nature of the language and resource allocation mechanism, but is
primarily concerned with the description of the physical architecture
of the machine.  The architecture is consistent with high-performance
VLSI implementation and packaging technology, and is easily extended
to include arbitrary numbers of processors.

------------------------------

         A Variable Supply Model for Distributing Deductions

                             Vineet Singh
                        Michael R. Genesereth

                         Stanford University

Multiple processors can be used to speed up a backward-chaining
deduction by distributing or-parallel deductions.  However, the actual
speedup obtained is strongly dependent on the amount of communication
required for the task allocation strategy.  A variable supply model
(VSM) is presented for multiple processors with replicated databases on
a broadcast network.  The term "model" refers to the set of procedures
and messages required to perform the computation.  VSM allows an
infinite class of strategies with varying amounts of communication.
The utility of VSM lies in the easy and powerful way it provides for
selecting a strategy that works satisfactorily given certain
communication constraints.  All strategies in VSM use a dynamic task
supply protocol (ESP) that works better than other supply protocols
described in the literature.

------------------------------

                      Parallelism in AI Programs

                           Dennis F. Kibler
                  University of California at Irvine

                             John Conery
                         University of Oregon

A folk theorem is developing which suggests that parallel solution of
AI programs will not afford a speedup of more than one order of
magnitude.  We critically review this folk theorem by analyzing some
of the problems used to "prove" it, and then cite work that provides
examples of better than one order of magnitude improvement for these
problems.  We examine two representative AI algorithms where
parallelism would achieve speedups of two orders of magnitude with a
reasonable number of processors.

------------------------------

          Recognition Algorithms for the Connection Machine

                            Anita M. Flynn
                            John G. Harris

                                 MIT

This paper describes an object recognition algorithm both on a
sequential machine and on a SIMD parallel processor such as the MIT
Connection Machine.  The parallel version is shown to run three to
four orders of magnitude faster than the sequential version.

------------------------------

            NON-VON's Applicability to Three AI Task Areas

                          David Elliot Shaw
                         Columbia University

NON-VON is a massively parallel machine constructed using custom VLSI
chips, each containing a number of simple processing elements.  A
preliminary prototype is now operational at Columbia University.  The
machine is intended to provide highly efficient support for a wide
range of artificial intelligence and other symbolic applications.
This paper briefly describes the current version of the NON-VON
machine and presents evidence for its applicability to the execution
of OPS5 production systems, a number of low- and intermediate-level
computer vision tasks, and certain "difficult" relational algebraic
operations relevant to knowledge base management.  Analytic and
simulation results are presented for a number of algorithms.  The data
suggest that NON-VON could provide a performance improvement of as
much as two to three orders of magnitude over a conventional
sequential machine for a wide range of AI tasks.

------------------------------


End of PARSYM Digest
********************

∂17-Sep-85  0911	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  09:11:30 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Sep 85 09:11:27 pdt
Date: Tue, 17 Sep 85 09:11:27 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

We'll meet at 11AM in 301 on Wednesday, as usual.
NExt Wednesday, the 25th, is Yom Kippur, and I will not
be here.  You folks can meet by yourselves if you wish,
but I suspect we should just cancel the meeting.

Anyway, the topic for this week will be further progress
on deciding whether a logic program is in NC or is P-complete.
Allen and I will hold forth.

∂17-Sep-85  0918	ullman@su-aimvax.arpa 	paper recieved   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  09:18:09 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Sep 85 09:17:58 pdt
Date: Tue, 17 Sep 85 09:17:58 pdt
From: Jeff Ullman <ullman@diablo>
Subject: paper recieved
To: nail@diablo, paco@diablo

Does anybody want to see: "Parallel Algorithms for Term MAtching""
by Dwork, Kanellakis, and Stockmeyer?
It improves the number of processors needed from n↑5 to n↑3,
while still running in polylog time.
				---Jeff Ullman

∂17-Sep-85  0922	RICHARDSON@SU-SCORE.ARPA 	Sr. Faculty Meeting September 24  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  09:22:20 PDT
Date: Tue 17 Sep 85 09:13:21-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Sr. Faculty Meeting September 24
To: tenured@SU-SCORE.ARPA
Message-ID: <12143997993.13.RICHARDSON@SU-SCORE.ARPA>

There will be a sr. faculty meeting on September 24, 1985 following the
general faculty meeting scheduled at 2:30 in Conference Room 146 of
Margaret Jacks Hall. Should you have any agenda items that you would like
to be included, please indicate to me.

Thanks,
Anne Richardson
-------

∂17-Sep-85  1140	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH REMINDER -- Wednesday, 11AM    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  11:38:42 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 17 Sep 85 11:29:52-PDT
Date: Tue 17 Sep 85 11:25:54-PDT
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH REMINDER -- Wednesday, 11AM
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanLehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA

			     SCRUFFY PLANNING

			       David Wilkins
                         SRI International AI Center

		        11:00 AM, WEDNESDAY, September 18
	         SRI International, Building E, Room EK242


Most of the talks in the planlunch series have been about "neat"
research.  While neat researchers know they have a more 
expressive formalism than do planners like SIPE, they have rarely
sat in front of a machine and watched their systems be overwhelmed
by the computational loads even simple representations can impose.

I'll defend scruffy planning (not to replace but to supplement neat planning),
show how SIPE represents a simple 6-room indoor world for Flakey, show 
its solutions and use of computation, describe the kind of hacks needed to
make things run fast, describe pitfalls, and other scruffy stuff like that.
Only a small amount of time will be spent explaining SIPE, depending on 
audience's desire.

-------
-------

∂17-Sep-85  1523	ullman@su-aimvax.arpa 	papers received  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  15:23:12 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Sep 85 15:22:02 pdt
Date: Tue, 17 Sep 85 15:22:02 pdt
From: Jeff Ullman <ullman@diablo>
Subject: papers received
To: nail@diablo

"Termination of Rewriting" by Nachum Dershowitz.
A survey about well-founded ordering proofs.

"Logic Programming cum Applicative Programming"
Dershowitz and Plaisted.

∂17-Sep-85  1622	MARJORIE@SU-CSLI.ARPA 	Books  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  16:22:34 PDT
Date: Tue 17 Sep 85 16:16:21-PDT
From: Marjorie Maxwell <MARJORIE@SU-CSLI.ARPA>
Subject: Books
To: folks@SU-CSLI.ARPA

We are missing some books from the consultants office - if anyone has
borrowed them and forgotten to return them please let us know.  They are
2 Scribe Users Manuals
1 Revised Maclisp Manual (Pitman, Kent)
1 A Practical Guide to Unix (Sobell, Mark)
1 Artificial Intelligence Programming (Charniak, Riesbeck)
Thanks,
Marj
-------

∂17-Sep-85  2008	PIERRE@SU-CSLI.ARPA 	Books    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Sep 85  20:08:47 PDT
Date: 17 Sep 1985  20:03 PDT (Tue)
Message-ID: <PIERRE.12144116343.BABYL@SU-CSLI.ARPA>
From: Peter King <PIERRE@SU-CSLI.ARPA>
To:   Marjorie Maxwell <MARJORIE@SU-CSLI.ARPA>
Cc:   folks@SU-CSLI.ARPA
Subject: Books
In-reply-to: Msg of 17 Sep 1985  16:16-PDT from Marjorie Maxwell <MARJORIE>


I have one of the Scribe Manuals.

Peter

∂18-Sep-85  0133	ACUFF@SUMEX-AIM.ARPA 	Explorer Mailing Lists 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 18 Sep 85  01:32:55 PDT
Date: Wed 18 Sep 85 01:34:22-PDT
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorer Mailing Lists
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12144176583.32.ACUFF@SUMEX-AIM.ARPA>

   In order to facilitate information exchange among projects using TI
Explorers, two mailing lists are being created.  INFO-TI-EXPLORER will
be used for general information distribution, such as operational
questions, or announcing new generally available packages or tools.
BUG-TI-EXPLORER will be used to report problems with Explorer
software, as well as fixes.  Requests to be added to or deleted from
these lists should be sent to INFO-TI-EXPLORER-REQUEST or
BUG-TI-EXPLORER-REQUEST, respectively.  All addresses are at
SUMEX-AIM.ARPA.  These lists signify no commitment from Texas
Instruments.  Indeed, there is no guarantee that TI representatives
will read the lists.  The idea of the lists is to provide
communication among the users of Explorers.  Archives will be kept on
SUMEX-AIM.ARPA in BBoard:*-TI-Explorer.txt, which can be with BBoard.

	-- Rich
-------

∂18-Sep-85  1536	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Sep 85  15:29:25 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Wed 18 Sep 85 15:21:22-PDT
Date: Wed 18 Sep 85 15:19:44-PDT
From: LANSKY@SRI-AI.ARPA
Subject: Next Monday's PLANLUNCH
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    vanlehn@XEROX.ARPA, araya@XEROX.ARPA, frayman@XEROX.ARPA,
    suchman@XEROX.ARPA, weld@XEROX.ARPA, mittal@XEROX.ARPA, dekleer@XEROX.ARPA,
    trigg@XEROX.ARPA


			THE SKY IS A BLUE COLOR

			   Marcel Schoppers
                      SRI International AI Center

		        11:00 AM, MONDAY, September 23
       SRI International, Building E, Room EJ228 (new conference room)


	I present a representation which will allow us to encode and
	access much of the information contained in simple descriptive
	statements. "The sky is a blue color" entails that blue is a
	color, that the sky has a color, that the sky is blue, that the
	sky is visible, and that the sky is located both spatially and
	temporally. These generalizations are so trivial that they border
	on presuppositions, and they have consequently been taken for
	granted in semantic nets and frames. Making such information
	explicit greatly increases the density and usefulness of stored
	knowledge. One interesting application is to disambiguate an
	adjective/predicate to suit a given noun/extension.

	My representation is parsimonious, having O(three) primitive con-
	structs (link types); is highly irredundant, since blue(sky) and
	color(blue) reference the same blue; and is static, being inten-
	ded to formalize and implement "massively parallel" deterministic
	connectionist question-answering systems. Predicates, relations,
	and simple forms of quantification all emerge as by-products of
	function applicability and set inclusion. Viewed as a logic the
	representation is potentially O(w), intensional and inconsistent.

	The talk will touch on issues in philosophy of logic and linguistics.
	I will especially appreciate constructive criticism in those areas,
	as I am a novice there.

-------

∂18-Sep-85  1615	SUSI@SU-CSLI.ARPA 	opportunity to share 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Sep 85  16:15:07 PDT
Date: Wed 18 Sep 85 16:10:00-PDT
From: Susi Parker <SUSI@SU-CSLI.ARPA>
Subject: opportunity to share
To: FOLKS@SU-CSLI.ARPA


     There will be a slide projector available in the Ventura Hall
     Conference Room if  you wish to use it (and the room, when available)
     do so~~~offer expires "this" Friday
-------

∂18-Sep-85  2111	davies%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar, Sept. 24, 1985   
Received: from UCB-VAX.ARPA by SU-AI.ARPA with TCP; 18 Sep 85  21:11:27 PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.6)
	id AA00948; Wed, 18 Sep 85 12:19:42 pdt
Received: by ucbcogsci.ARPA (5.5/5.7)
	id AA07592; Wed, 18 Sep 85 11:12:07 PDT
Date: Wed, 18 Sep 85 11:12:07 PDT
From: davies%ucbcogsci@Berkeley (Catherine Davies)
Message-Id: <8509181812.AA07592@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar, Sept. 24, 1985

                         BERKELEY COGNITIVE SCIENCE PROGRAM
                                     Fall 1985
                       Cognitive Science Seminar -- IDS 237A

           TIME:                Tuesday, September 24, 11:00 - 12:30
           PLACE:               240 Bechtel Engineering Center
              (followed by)
           DISCUSSION:          12:30 - 1:30 in 200 Building T-4

           SPEAKER:          Peter  Pirolli,  School  of  Education,   
                             UC Berkeley

           TITLE:          ``A Cognitive Model and Intelligent Computer
                             Tutor for Programming Recursion''


           Recursion is typically a novel concept  for  programming  stu-
           dents  that  causes  them  considerable  grief and difficulty.
           Thus, the study of how people learn to program recursive  pro-
           grams  provides a useful domain for addressing the psychologi-
           cal issue of how fundamentally new knowledge  is  acquired  as
           well  as  the  instructional issue of how to teach a difficult
           programming concept.  I will present a production system model
           that  addresses  expert  and  novice  problem-solving, problem-
           solving by  analogy,  and  skill  acquisition  in  programming
           recursive  functions.   This  research served as the basis for
           the development of recursion lessons in  an  intelligent  com-
           puter  tutor for programming LISP.  Specifically, a simulation
           model of ``ideal'' and ``buggy'' novice problem-solving was con-
           structed  for  coding  recursion.   Using this model, the LISP
           tutor provides instruction, hints, and feedback in the context
           of  programming.  The LISP tutor also maintains a model of the
           skill development of  individual  students.  Evaluations  show
           that the LISP tutor is more effective in teaching introductory
           LISP  programming  than   good   classroom   instruction   and
           approaches the effectiveness of human tutors.
           ---------------------------------------------------------------------
           UPCOMING TALKS

           Oct  1:      David Rumelhart, Cognitive Science, UCSD
           Oct  8:      Terry Winograd, Computer Science, Stanford
           Oct 15:     Ron Kaplan, Xerox PARC
           Oct 22:     Lotfi Zadeh, Computer Science, UCB
           Oct 29:     Mardi Horowitz, Psychiatry, UCSF
           Nov  5:     TBA
	   Nov 12:     TBA
           Nov 19:     Richard Alterman, Computer Science, UCB
           Nov 26:     Eve Clark, Linguistics, Stanford

∂18-Sep-85  2112	@SU-CSLI.ARPA:davies%ucbcogsci@Berkeley 	UCB Cognitive Science Seminar, Sept. 24, 1985    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Sep 85  21:12:43 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Wed 18 Sep 85 21:08:54-PDT
Received: from ucbcogsci.ARPA by UCB-VAX.ARPA (4.24/5.6)
	id AA00948; Wed, 18 Sep 85 12:19:42 pdt
Received: by ucbcogsci.ARPA (5.5/5.7)
	id AA07592; Wed, 18 Sep 85 11:12:07 PDT
Date: Wed, 18 Sep 85 11:12:07 PDT
From: davies%ucbcogsci@Berkeley (Catherine Davies)
Message-Id: <8509181812.AA07592@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley
Subject: UCB Cognitive Science Seminar, Sept. 24, 1985

                         BERKELEY COGNITIVE SCIENCE PROGRAM
                                     Fall 1985
                       Cognitive Science Seminar -- IDS 237A

           TIME:                Tuesday, September 24, 11:00 - 12:30
           PLACE:               240 Bechtel Engineering Center
              (followed by)
           DISCUSSION:          12:30 - 1:30 in 200 Building T-4

           SPEAKER:          Peter  Pirolli,  School  of  Education,   
                             UC Berkeley

           TITLE:          ``A Cognitive Model and Intelligent Computer
                             Tutor for Programming Recursion''


           Recursion is typically a novel concept  for  programming  stu-
           dents  that  causes  them  considerable  grief and difficulty.
           Thus, the study of how people learn to program recursive  pro-
           grams  provides a useful domain for addressing the psychologi-
           cal issue of how fundamentally new knowledge  is  acquired  as
           well  as  the  instructional issue of how to teach a difficult
           programming concept.  I will present a production system model
           that  addresses  expert  and  novice  problem-solving, problem-
           solving by  analogy,  and  skill  acquisition  in  programming
           recursive  functions.   This  research served as the basis for
           the development of recursion lessons in  an  intelligent  com-
           puter  tutor for programming LISP.  Specifically, a simulation
           model of ``ideal'' and ``buggy'' novice problem-solving was con-
           structed  for  coding  recursion.   Using this model, the LISP
           tutor provides instruction, hints, and feedback in the context
           of  programming.  The LISP tutor also maintains a model of the
           skill development of  individual  students.  Evaluations  show
           that the LISP tutor is more effective in teaching introductory
           LISP  programming  than   good   classroom   instruction   and
           approaches the effectiveness of human tutors.
           ---------------------------------------------------------------------
           UPCOMING TALKS

           Oct  1:      David Rumelhart, Cognitive Science, UCSD
           Oct  8:      Terry Winograd, Computer Science, Stanford
           Oct 15:     Ron Kaplan, Xerox PARC
           Oct 22:     Lotfi Zadeh, Computer Science, UCB
           Oct 29:     Mardi Horowitz, Psychiatry, UCSF
           Nov  5:     TBA
	   Nov 12:     TBA
           Nov 19:     Richard Alterman, Computer Science, UCB
           Nov 26:     Eve Clark, Linguistics, Stanford

∂19-Sep-85  0850	EMMA@SU-CSLI.ARPA 	Newsletter September 19, No. 46
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Sep 85  08:50:45 PDT
Date: Thu 19 Sep 85 08:14:26-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter September 19, No. 46
To: friends@SU-CSLI.ARPA
Tel:  497-3479

***** Sorry for the delay but SU-CSLI crashed just as I was about to send
the Newsletter yesterday.******



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
September 19, 1985              Stanford                       Vol. 2, No. 46
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *THIS* THURSDAY, September 19, 1985

   12 noon		TINLunch
     Ventura Hall       ``Some Remarks on the Relationship of Mind to 
     Conference Room    Meaning and Language''
			Discussion led by Daniel Isaacson, Oxford University

   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week
     Seminar Room	

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *NEXT* THURSDAY, September 26, 1985

   12 noon		TINLunch
     Ventura Hall       ``The Concept of Supervenience''
     Conference Room    Discussion led by Carol Cleland
			(Abstract on page 1)
			
   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week
     Seminar Room	

   3:30 p.m.		Tea
     Ventura Hall
                              ←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
                    ``The Concept of Supervenience''

      Traditionally the notion of supervenience has been associated with
   moral philosophy (particularly, value theory).  In recent years,
   however, there has been a growing interest among philosophers in
   developing a concept of supervenience that could be employed in the
   analysis of certain problematic relations, e.g., between the mental
   and the physical, between macrostates of the world and microstates of
   the world.
      The appeal of the concept of supervenience for philosophers
   involves several factors.  First, supervenience is a weaker relation
   that the relation of so-called ``reducibility.''  While reducibility
   is traditionally taken to involve the presence of bi-conditional
   correlations between every ``reduced'' property and every ``reducing''
   property, supervenience does not.  Yet, like reducibility,
   supervenience appears to be able to provide us with a robust notion of
   the determination of one family of properties by another.
      The question is: Can supervenience live up to its promise?
							--Carol Cleland
!
Page 2                     CSLI Newsletter                 September 19, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                  TALK
                           ``Proto-Language''
       Professor C. B. Martin, Dept. of Philosophy, U. of Calgary
           Friday, September 20, 2:15, Ventura Conference Room

      C. B. Martin has done a lot of important work in the philosophy of
   religion, but also he wrote an extremely important paper on memory
   with Max Deutscher, defending a causal theory of memory when this was
   quite unfashionable.  I think this paper played an important role in
   setting the stage for causal theories of reference and action.  Martin
   is now doing work that sounds extremely interesting to me on the
   semantics of non-verbal behavior.  This following paragraph from his
   paper gives a good indication of what it is about.
   	``The time is long overdue for the recognition of the semantic
   	import of non-verbal behaviour.  Such behaviour is procedural
   	and projective for an outcome (that may or may not have
   	satisfaction).  Though ``true'' and ``false'' may be reserved
        for the verbal cases, there is a basic rightness and wrongness
   	about the non-verbal behavioural, procedural, projective
   	representations.  Such behaviour is formed in inter-related
   	patterns strikingly and importantly analogous to that of
   	verbal language.  I shall call such semantic non-verbal
   	behaviour ``proto-language''.''			--John Perry
                              ←←←←←←←←←←←←
                                  TALK
    ``Crossing the Rubicon: From a Physics of Dead Coordinate Spaces
               to a Physics of Living Coordinate Spaces''
   Dr. Peter Kugler, The Crump Institute for Medical Engineering, UCLA
            Monday, September 23, 1985, 2:15pm, Ventura Hall

      This talk will be about self-organizing systems that involve
   low-energy (nonforce) coupling and the nature of the predicates that
   constitute the low-energy descriptors, and will be organized around
   issues pertaining to general problems of language and information.
   The emphasis will be on systems that generate (self-assemble) new
   levels of description.  These new levels constitute new languages
   parasitic on the lower level languages but not reducible to their
   predicates.  In the self-organizing systems of interest it is the
   ``coordinate spaces,'' which are themselves evolving, that become the
   important objects of study.  Instead of assuming a fixed coordinate
   space, when the interest focuses on trajectories, attention is devoted
   to the coordinate space itself, since this is what provides the semantics.
      This approach is very similar to developments in computer
   architecture that focus on parallel processing.  In these machines
   (connection machines, Boltzmann, etc.) the machine language
   self-organizes (e.g. programs itself through the emergence of new
   stable configurations), and the new predicate descriptions play the
   role of symbols in terms of their opacity with respect to the lower
   level language.  The machine language `gives birth' to the symbolic
   level of description.  This situation contrasts dramatically with that
   of von Neumann machines, for which the symbolic language is
   ontologically independent of the machine language.  A symbolic
   language can run on any of an infinite variety of mechanistic
   substrates, the primacy of the symbol prevailing over the substrate
   machine.  The approach advocated here, puts the focus on the machine
   level of interaction, thus preserving an ontological continuity and
   avoiding mind/body, syntactic/semantic, etc. problems.
!
Page 3                     CSLI Newsletter                  September 19, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
      SUMMARY OF A TALK TO THE DISCOURSE INTENTION AND ACTION GROUP
           ``Reference and Denotation: The Descriptive Model''
                      Ami Kronfeld, AI Center, SRI
                         Tuesday, September 10.

      The descriptive approach to the problem of reference has
   recently been challenged. One of the most devastating weapons
   against it has been the Referential/Attributive distinction.  I
   argue that this distinction is defined by two criteria which are
   independent of each other.  The first is the ability to refer using
   the ``wrong'' description; the second is based on the notion of
   ``having a particular object in mind.''  The first criterion is
   explained in terms of a distinction between a functionally relevant
   description (where the description is used only to identify), and a
   conversationally relevant description (where the description takes
   part in a Gricean implicature).  The second criterion is explained
   in terms of the de-re/de-dicto distinction.  I examine the claim
   that an individual concept is neither necessary nor sufficient for
   a de-re belief, and I argue that a Russellian notion of
   acquaintance and a theory of the pragmatics of reporting beliefs
   can provide a descriptive account of de-re thought.  The discussion
   that followed the talk focused on (a) the ability of the
   descriptive model to handle reference to objects that were
   perceived in the past, (b) the role of the self in the
   individuation of beliefs, and (c) whether the concept of ``simple''
   reference, where the description is only functionally relevant, is
   really necessary.				--Ami Kronfeld

-------

∂20-Sep-85  0040	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	separating DL and NL using an oracle ?  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Sep 85  00:40:18 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 20 Sep 85 00:33:36-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Fri 20 Sep 85 00:33:43-PDT
Received: by wisc-rsch.arpa; Fri, 20 Sep 85 01:40:21 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Fri, 20 Sep 85 01:32:49 cdt
Received: from sally.UTEXAS.EDU by wisc-crys.arpa; Fri, 20 Sep 85 01:31:34 cdt
Date: Thu, 19 Sep 85 13:11:26 cdt
From: krishna@sally.UTEXAS.EDU
Posted-Date: Thu, 19 Sep 85 13:11:26 cdt
Message-Id: <8509191811.AA21205@sally.UTEXAS.EDU>
Received: by sally.UTEXAS.EDU (4.22/4.22)
	id AA21205; Thu, 19 Sep 85 13:11:26 cdt
To: theory@wisc-crys.arpa
Subject: separating DL and NL using an oracle ?
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

     I have read somewhere that separating deterministic logspace
from nondeterministic logspace using an oracle would imply
separating the classes in the unrelativised case.
     Can somebody give pointers about this result. I am specially
interested in the case where the oracle is NOT a random oracle.

     Specific references or a proof would be helpful!!

     Thanks in advance.

∂20-Sep-85  1019	VSINGH@SUMEX-AIM.ARPA 	Comments    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 20 Sep 85  10:19:29 PDT
Date: Fri 20 Sep 85 10:14:55-PDT
From: Vineet Singh <vsingh@SUMEX-AIM.ARPA>
Subject: Comments
To: aap@SUMEX-AIM.ARPA
Message-ID: <12144795632.39.VSINGH@SUMEX-AIM.ARPA>

This is an advertisement for my paper "PM: A Parallel Execution Model
for Backward-Chaining Deductions".  It is available as KSL-85-18 in
bldg. C.  Any comments will be appreciated.  I am including the
abstract below.

				Abstract

This paper describes @i[PM], an execution model for automating
backward-chaining deductions on multiple processors.  The term
@i[execution model] refers to the state, messages and procedures
required to perform the computation @i[correctly].  The target
multiprocessor is characterized by (1) a large number of small
processors, (2) inter-processor communication via messages, and (3) a
distributed database.  The key distinguishing feature of @i[PM] is
simultaneous exploitation of @i[and-parallelism], @i[or-parallelism]
and @i[pipelining] in this scenario.

Happy reading!

Vineet
-------

∂20-Sep-85  1211	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.15  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 20 Sep 85  12:11:19 PDT
Date: Fri 20 Sep 85 10:34:12-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.15
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-FORUM Digest      Friday, 20 Sep 1985      Volume 1 : Issue 15

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  SDI Panel at 8th ICSE in London (David Weiss)
  Risks to the Moderator (PGN)
  Mailer Protocol Woes (Marty Moore)
  Another Horror Story -- Sidereal Time Rollover (Marty Moore)
  Article: Health Hazards of Computers (Ted Shapin)
  Two More SDI Related Queries (douglas schuler)
  CAL ID -- computerized fingerprint system (douglas schuler)

Rejections:
  Either your contributions are drifting off the mark, or I must be jacking 
  up my standards (or both).  I rejected several items this time.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Wed 18 Sep 1985 16:51:10 EST
From: David Weiss <wanginst!weiss@nrl-css.arpa>
Subject: SDI Panel at 8th ICSE in London
To: RISKS@SRI-CSLA

    Panel Discussion on the Strategic Defense Initiative at the 8th
         International Conference on Software Engineering

One of the more interesting sessions at the 8th International
Conference on Software Engineering was a discussion of software for
the Strategic Defense Initiative (SDI).  The moderator was Manny
Lehman and the panelists were Fred Brooks, David Parnas, and Alan
Perlis.  The discussion was originally intended to be a debate, but
Fred Brooks was not willing to participate in a debate because he had
not yet reached a resolution of the issues.  (I understand that
volunteers for the position of opposing Parnas in a debate on SDI
were hard to find.  Dr. Brooks deserves credit for being willing to
engage in a public exploration of touchy issues about which he feels
unsettled in concert with a strongly opinionated colleague in an
effort for both audience and panel to learn more.)  

The discussion started with a presentation by Parnas of the technical
reasons why reliable SDI software cannot be built.  (Readers of this
newsletter will be familiar with many of the arguments put forth by
Parnas.  A complete discussion in hard-copy is available from him at
the University of Victoria, Department of Computer Science, P.O.  Box
1700, Victoria, B.C., Canada.)  

Brooks responded with reasons why he thought we could build such a
system.  His major point was that we have built similar systems in
the past.  He identified the Apollo missions software as an example,
suggesting that we start with such a system and incrementally build
from it towards an SDI system, using what's learned along the way.

Perlis then added a few comments, explaining why SDI software would
be more complex than existing software and why it is of the hardest
type of software to build.  His argument was that the SDI system
represents a moving target in terms of requirements and design.

Following some further discussion among the panelists the floor was
opened to technical questions from the audience.  

The major place in which Parnas and Brooks seemed to disagree was
whether or not similar systems have been built.  Brooks tried to use
the Apollo and Space Shuttle as examples.  Parnas's point was that in
those systems everything can be predicted in advance.  In an
anti-missile system, the number, shape, and trajectories of launched
missiles can't be predicted.  In addition, the system must
distinguish decoys from real warheads.  Finally, the defense system
itself will be under attack.  As a result, realistic tests and
simulations of operating conditions for such a system could not be
conducted.  

All the discussants seemed to agree that an SDI system could not be
built error-free, and that it would not be completely reliable.
Nonetheless, there were advocates of building it on such grounds as
that it would only be needed for a short time, and could be turned
off the rest of the time, or that we now place our trust in systems
that are also untested and probably unreliable.  

In summary, there were no good responses to any of the questions that
Dave Parnas raised.  Nonetheless, there were arguments put forth for
the construction of an SDI system on the grounds that it need not be
completely reliable.  

David Weiss
Wang Institute of Graduate Studies

------------------------------

Date: Thu 19 Sep 85 17:40:08-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Risks to the Moderator!
To: RISKS@SRI-CSLA.ARPA

RISKS-1.15 was ready to go out Wednesday night.  Murphy hit in spades.  The
SRI MICOM switch for dial-up access was unavailable for two days, and is
still down.  An alternative route might have been available through the only
system that receives dial-ups directly from a split-speed modem, but that
system went down for five hours.  Several other more circuitous alternative
routes all ran into broken gateway, which resulted from a power failure
Tuesday night.  It would not have helped to drive back to SRI, because the
SRI-CSLA (which kept running through all this) was out of net contact as a
result of a gateway problem.  All of the gurus were invisible.  If you ever
get this issue, you will know that things have improved.  Peter

------------------------------

Date: Tue, 17 Sep 85 07:40:58 CDT
From: mooremj@EGLIN-VAX
Subject: Mailer Protocol Woes
To: risks@sri-csl.arpa

Receiving 2 or 3 duplicate messages is a minor annoyance.  Receiving over
a hundred is a *major* annoyance.  A few weeks ago, a message from SECURITY
(a non-digest, re-mail distribution list) contained a record that was too
long for our mailer to handle.  The result was that the message would be
transmitted from Rutgers, and our mailer would have a problem with the long
record.  The message would not be successfully acked, but the mailer would
send me the partial message up to the problem.  Since the message was not
acked, Rutgers kept remailing it to me.  Before the Rutgers wizards flushed
the message, I had received 173 (I think) copies of the partial message;
at times I was receiving one every 15 minutes!  This happened just before I
left on vacation, and I was seriously concerned about returning to work and
finding my disk quota used up by N*1000 copies of the busted message...

                                     Marty Moore (mooremj@eglin-vax.arpa)

    [I'm glad we're not the only ones!  I think the protocols really
     need further ruggedization.  Thanks.  PGN]

------------------------------

Date: Tue, 17 Sep 85 12:41:04 CDT
From: mooremj@EGLIN-VAX
Subject: Another Horror Story -- Sidereal Time Rollover
To: risks@sri-csl.arpa

How many of you real-time programmers have been bitten by time rollover at 
midnight?  How about *sidereal* time rollover?  It happened like this:

In the late 70's I worked on the USNS Redstone, which is the primary tracking 
and support ship for at-sea test launches of the Trident Submarine Launched
Ballistic Missile.  I wrote a section of program which took telemetry data
from the Trident's Inertial Guidance Unit and reduced it to provide track 
data.  Now, Inertial Guidance is like the little girl in the famous rhyme:
when it's good, it's very very good, but when it's bad, it's very very bad.
As such, we had some fairly extensive reasonableness checks on the data.
One in particular took the data's time tag (in sidereal hour angle format),
differenced it with a reference hour angle computed at program initialization, 
converted the answer to seconds, and compared this to the program's running
time.  If the two times were dissimilar, the IG data was rejected.  This
check worked beautifully on numerous tests, with both simulated and actual 
input data.

Unfortunately, the programmer (blush, cringe, hang head in shame) completely 
overlooked the possibility that the sidereal hour angle could reach 2*pi
radians and roll over during the mission.  This eventually happened on a "2+2" 
test launch.  In a "2+2" launch, two missiles are launched close together,
then two more are launched close together after a lengthy delay.  The sidereal 
hour angle rolled over about five minutes before the first missile was 
launched.  The program decided that the IG data had a bad time tag and promptly
rejected it.  Fortunately, other devices were tracking the missiles; mission 
rules stated that if no track data was received for a certain period, missiles
in flight must be destroyed.

During the delay between the first and second missile pairs, I carefully -- 
very, very carefully -- patched the running program to disable the time check.
On the second pair of missiles, the IG data was great, which was a good 
thing, because for about 40 seconds, no other device tracked them; if the IG
had also failed, the missiles would have been destroyed.  If the sidereal 
rollover had occurred *between* the two pairs of launches...(gulp)

The moral: the check worked great on numerous tests, until a peculiar set of
conditions occurred.  When the bug bit, we were able to save the test; but
with just a small change in conditions, we could have destroyed two Trident 
missiles unnecessarily.  I don't know what they cost, but I'm sure it's at 
least $10,000,000 each.

                                   Marty Moore (mooremj@eglin-vax.arpa)

------------------------------

Date: Wed 18 Sep 85 15:59:24-PDT
From: Ted Shapin <BEC.SHAPIN@USC-ECL.ARPA>
Subject: Article: Health Hazards of Computers
To: RISKS@SRI-CSL.ARPA
Phone: (714)961-3393; Mail:Beckman Instruments, Inc.
Mail-addr: 2500 Harbor Blvd., X-11, Fullerton CA 92634

"The Health Hazards of Computers" edited by Art Kleiner. Pages 80-93,
Whole Earth Review, No. 48, Fall 1985, $4.50. PO Box 15187, Santa Ana, CA.
92705, or your news-stand.

   "This amalgamation of information, conjecture, experiment, and reporting is
    the end of a 12-month odyssey.  It started last June, when we were planning
    the 'Computers as Poison' issue (Fall '84 WER)."

   "We really should have something on the health hazards of video-display 
    terminals (VDTs)." I said to Kevin and Stewart. "After all, it's a major 
    uncertainty. You sit with you nose squeezed up against the beast for hours 
    every day; you hear vague reports of cataracts and birth defects; you hear,
    on the other hand, industry groups saying there's nothing wrong with the 
    machines .... Whom should you believe?"

  A tip from Mike Castleman of  Medical Self-Care Magazine  led me to the
  Center for Investigative Reporting in San Francisco.  A reporter there named
  Diana Hembree had already been investigating VDT radiation health hazards
  for several months, with a particular interest in its effects on women
  workers -- most VDT terminal grunt workers, such as airline reservation
  clerks and data- entry operators are women. At my request, she assembled a
  group of investigators to look into potential radiation hazards from
  personal computers.

  Their original article arrived in time for the Computers as poison issue,
  but because it reported on a situation that was simultaneously
  controversial, extremely technical, and inconclusive, we didn't feel
  comfortable printing the article without scientific review.

  Thus we held it and sent it to two dozen physicists, radiologists,
  biophysicists, and doctors -- all people with a preestablished interest in
  this topic. Diana's original theme wasn't particularly incendiary; it
  basically said, "There seems to be a cause for concern, but nothing
  conclusive; more research is needed."  We got back a dozen replies, some
  complimentary and other criticizing us for everything from hysterical
  sensationalism to underplaying the danger. Some of those replies led to
  further interviews that supplemented Diana's already exhaustive research.
  Meanwhile, discussion of the EIES computer network began turning up comment
  from other people who had investigated the issue.

  Ultimately, I edited Diana's article, plus some of the replies and other 
  comments into these 14 pages.

The article ends with a bibliography and notes.  Ted.

------------------------------


Date: Thu, 19 Sep 85 12:41:36 pdt
From: bcsaic!douglas@uw-june (douglas schuler)
To: RISKS@SRI-CSL
Subject: Two More SDI Related Queries

I have two queries   r e l a t e d   to SDI but (I hope) of general
risks interest.

1.  Does anybody have rough heuristics for comparing the complexity of
    large projects?  I'd like to see a matrix where several very large 
    projects were compared feature by feature (e.g. person-years, LOC, 
    cost, function, etc)

2.  I would be very curious to see the results of using Boehm's estimating 
    techniques (←Software Engineering Economics←) on the SDI software.  The 
    techniques were developed at TRW and, hence, may be applicable to SDI.

	Doug Schuler     (206) 763-5295
	{allegra,ihnp4,decvax}uw-beaver!uw-june!bcsaic!douglas
	uw-june!bcsaic!douglas@washington.arpa

------------------------------

Date: Mon, 16 Sep 85 07:55:43 pdt
From: bcsaic!douglas@uw-june (douglas schuler)
To: Neumann@SRI-CSLA.arpa
Subject: CAL ID -- computerized fingerprint system

This isn't really a submission, just a noteworthy subject that I heard on
NPR this morning.  The "CAL ID" computer system is a $40 million system
(from NEC) for storing and retrieving finger prints.  The system has not
been officially accepted as of yet as only 2 million of the 2.5 million
fingerprinted California citizens are stored.  It is still being tested.
The system was used successfully to identify the "nightstalker" from
fingerprints.  Only males born since 1960 had been included.  Ramirez was
born in February, 1960.  It was estimated that the new system will result in
20,000 additional arrests per year in California.

        [I thought this was worth including.  There are all sorts of 
         associated risks.  PGN]

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂20-Sep-85  1248	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 20 Sep 85  12:48:03 PDT
Received: from ames-vmsb.ARPA by SU-SCORE.ARPA with TCP; Fri 20 Sep 85 12:14:09-PDT
Date: 18 Sep 85 18:34:00 PST
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA

Reply-To: "DYMOND, KEN" <dymond@nbs-vms>




	    NBS PARALLEL COMPUTER BENCHMARK COLLECTION



     The National Bureau  of Standards, since  its  founding, has been 
concerned with measurement, determining the precise values and metrics
for physical phenomena. The NBS has also made significant contri-
butions to metrology in numerous scientific and engineering
disciplines. In this tradition, the MPC (Measurement for Parallel
Computing) project at NBS is developing a set of metrics  and measure-
ment techniques to characterize the performance of parallel processing
systems.  As part of that effort, NBS is collecting benchmarks and code
kernels that represent a variety of applications which are candidates
for parallel processing.  NBS solicits benchmark codes and kernels from
researchers and scientists.  Programs which are computationally
intensive, I/O intensive, vectorizable or not and from non-numeric as
well as from numeric application areas are requested.  Especially
welcome are programs which have been used to produce timing or speedup
data on parallel computers, whose measurement results have been or may
be published in the technical literature, and which are in some fairly
widely used and higher-level programming  language such as FORTRAN,
"C", LISP, Ada, etc.

  
Contributions or inquiries should be directed to:

	Measurement for Parallel Computing
        Institute for Computer Sciences and Technology
	Materials Building MS B364
	National Bureau of Standards
	Gaithersburg, MD 20899 USA

	Telephone: 301-921-3274
	ARPANET: MEASURE@NBS-VMS

------
------

∂20-Sep-85  1751	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Final FOCS Forewarning   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Sep 85  17:51:31 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 20 Sep 85 17:45:40-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Fri 20 Sep 85 17:43:54-PDT
Received: by wisc-rsch.arpa; Fri, 20 Sep 85 19:23:56 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Fri, 20 Sep 85 19:20:41 cdt
Message-Id: <8509210020.AA14822@wisc-crys.arpa>
Received: from CSNET-RELAY.ARPA by wisc-crys.arpa; Fri, 20 Sep 85 19:20:36 cdt
Received: from uoregon by csnet-relay.csnet id aa08603; 20 Sep 85 20:10 EDT
Received: by uo-vax3.UONET (4.12/1.0.uoregon)
	id AA13115; Fri, 20 Sep 85 15:21:34 pdt
Date: Fri, 20 Sep 85 15:21:34 pdt
From: Fall Conference Mail <focs%uoregon.csnet@CSNET-RELAY.ARPA>
Posted-Date: Fri, 20 Sep 85 15:21:34 pdt
To: theory@wisconsin.ARPA
Subject: Final FOCS Forewarning
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;


A final reminder on advance registration and room reservations for FOCS -


The deadline for Advance Registration is September 27th.  Non-student
registration fees go up by $35 after that date.  This policy is firm.

The quoted room rates at the Marriott presume reservations 21 days in
advance of arrival.  Rates are likely to be considerably higher in
the absence of such reservations.


Also, if you are counting on the US Mail, it travels here via covered
wagon over the Oregon Trail.  There are already rumors that Grant's
Pass is snowed in.


∂20-Sep-85  2106	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	parallel machine bibliography 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Sep 85  21:06:36 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 20 Sep 85 20:59:26-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Fri 20 Sep 85 20:59:34-PDT
Received: by wisc-rsch.arpa; Fri, 20 Sep 85 22:44:29 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Fri, 20 Sep 85 20:23:44 cdt
Received: from CSNET-RELAY.ARPA by wisc-crys.arpa; Fri, 20 Sep 85 20:23:38 cdt
Received: from lsu by csnet-relay.csnet id a008909; 20 Sep 85 21:14 EDT
Received: by lsu.CSNET ; Fri, 20 Sep 85 13:15:30 cdt
Date: 20 Sep 1985 13:13-EST
From: Steve & <cater%lsu.csnet@CSNET-RELAY.ARPA>
Subject: parallel machine bibliography
To: theory@uwisc.csnet
Message-Id: <496088003/cater@lsu>
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;

I am constructing a bibliography for the general topic of parallel
automata. Subjects include
    
    any parallel machine (TM, stack machine, etc.)
    recursive machines
    cellular automata

as well as anything remotely related to any of the above. If you have any
references that might apply, please send them to me at

     cater@lsu.CSNET

If there is demand for the results, I am willing to either mail
individual copies of the bibliography (small demand), or post the
results on the net (large demand).

Thanks.

Steven Cater

←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←


∂21-Sep-85  1550	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #17  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 21 Sep 85  15:50:14 PDT
Date: 21 Sep 85 1549-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #17
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Saturday, 21 Sep 1985      Volume 1 : Issue 17

Today's Topics:

                     IJCAI-85 abstracts: omission
            Connectionist network teaching / learning tool
                    Parallel machine bibliography
               Parallel rule execution: examples needed


----------------------------------------------------------------------

Date: Saturday, 21 September 1985 15:34-PDT
From: Byron Davies <Davies@SUMEX-AIM.ARPA>
Re:   IJCAI-85 abstracts: omission

I inadvertently omitted the following abstract from the collection of
IJCAI-85 articles in the last digest.  No offense was intended; my
apologies to the authors.


               Yes, An SIMD Machine Can Be Used For AI

                             Ruven Brooks
                             Rosalyn Lum

                                 ITT

Nearly all of the proposed parallel architectures for artificial
intelligence applications use a multiple instruction, multiple data
stream (MIMD) approach.  While this approach offers the greatest
opportunity for the ultimate exploitation of parallelism, it has been
difficult to actually achieve a high level of parallelism in practice
because most of the existing algorithms require a higher
interprocessor communications bandwidth than the hardware can achieve.

An alternate approach is the use of single instruction, multiple data
stream parallelism (SIMD) machines.  While SIMD machines do not offer
the same ultimate exploitation of parallelism as MIMD architectures,
they may, in fact, procide more useable parallelism.  This is because
they can be treated as serial machies with very long data words, so
that existing algorithms may be more readily adapted in ways which
better use available parallelism.  We illustrate this concept with the
Cellular Array Processor (CAP) being developed at the ITT Advanced
Technology Center.  This SIMD architecture is based on a rectangular
processing array which accesses a single memory and which has very
high speed data paths among the processing elements.  We discuss the
implementation and manipulation of data for two applications on the
CAP: the OPS5 production interpreter and a representation of an
associative network.  The expected performance of the CAP in these
applications will also be presented.

------------------------------

Date: Friday, 13 September 1985  13:22-PDT
From: Hon Wai Chun <hon%brandeis.csnet at CSNET-RELAY.ARPA>
Re:   Connectionist network teaching / learning tool

A connectionist network teaching / learning tool called AINET-1 is available
for distribution (educational and research purposes only) from the Computer
Science Department, Brandeis University.

AINET-1 is a graphic-oriented software package which can be used to
interactively create, manipulate, and experiment with connectionist
networks.  Most commands are conveniently driven by a mouse.  Nodes in
AINET-1 are shaded to reflect their activation levels.  Once a network is
created, the user can run an animated simulation (network relaxation).  In
the simulation, AINET-1 will change the various node shadings as the
activation levels change during each run cycle.  After a simulation, the
user can plot the results or display tables of previous activation levels.
Networks can be stored into binary files and reloaded later for further
editing.

AINET-1 is intended to be used mainly as a learning tool to give the user a
flavor of how connectionist networks behave.  A more sophisticated version,
called AINET-2 (under development), may be useful for development work.

AINET-1 is written in Symbolics Common Lisp and presently runs on Symbolics
Lisp machines (Release 6.0).  The system is offered on a non-commercial,
non-disclosure, and as-is basis for a nominal fee.

The fee is $150 for universities, and $250.00 for laboratories.  Interested
parties should send requests to (or call).


Hon Wai Chun       hon@brandeis.csnet
Computer Science
Brandeis University
Ford 232A
Waltham, MA 02254

617-647-2650 or
617-647-2119 (main-office)

------------------------------

Date: 20 Sep 1985 13:13-EST
From: Steve & <cater%lsu.csnet at CSNET-RELAY.ARPA>
To:   theory at uwisc.csnet
Re:   parallel machine bibliography

[Forwarded to PARSYM by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

I am constructing a bibliography for the general topic of parallel
automata.  Subjects include:

    any parallel machine (TM, stack machine, etc.)
    recursive machines
    cellular automata

as well as anything remotely related to any of the above. If you have any
references that might apply, please send them to me at:

     cater@lsu.CSNET

If there is demand for the results, I am willing to either mail
individual copies of the bibliography (small demand), or post the
results on the net (large demand).

Thanks.

Steven Cater

------------------------------

Date: Tue, 17 Sep 85 09:04:36 pdt
From: Thomas L. Zimmerman <zimmer%marlin at nosc.ARPA>
To:   AIList
Re:   Parallel Rule Execution

[Forwarded to PARSYM by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

The Naval Ocean Systems Center (NOSC) and Goodyear Aerospace are
working on a series of experiments to demonstrate the feasibility of
running expert systems on the Goodyear ASPRO parallel processor.  A
unique representation for rules and data has been developed which on
paper allows the ASPRO to test 2,000,000 rules per second for
satisfaction.  This representation scheme does put some limitations on
the system involved - it appears to require a very pure production
system with no embedded control or functions in the rules.  So far a
small (500 rule) system was written with this application in mind and
run on both a Symbolics and the ASPRO sucessfully.  We would now like
to convert an existing sequential expert system for parallel execution
in order to determine the degree of speedup actually available and to
discover the limitations of converting a system not originally
designed for this application.  Unfortunatly I am having trouble
finding a system to convert for this demonstration - which is why I am
appealing to all of you.  We need a reasonably sized (200-700 rule)
system, preferably military in character, that meets the above
limitations that we can attempt to run on our parallel inference
engine.  This would be a no-cost way for someone to have their system
speeded up by an estimated three orders of magnitude.  Any takers?  If
you're interested please contact me:

        Lee Zimmerman
        Naval Ocean Systems Center
        Code 421
        San Diego  CA  92152
        (619) 225-6571  or zimmer@nosc

------------------------------

End of PARSYM Digest
********************

∂22-Sep-85  2315	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH reminder   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Sep 85  23:13:09 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Sun 22 Sep 85 23:07:58-PDT
Date: Sun 22 Sep 85 23:00:12-PDT
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH reminder
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, rsimmons@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    araya@XEROX.ARPA, frayman@XEROX.ARPA, suchman@XEROX.ARPA, weld@XEROX.ARPA,
    mittal@XEROX.ARPA, dekleer@XEROX.ARPA, trigg@XEROX.ARPA


			THE SKY IS A BLUE COLOR

			   Marcel Schoppers
                      SRI International AI Center

		        11:00 AM, MONDAY, September 23
       SRI International, Building E, Room EJ228 (new conference room)


	I present a representation which will allow us to encode and
	access much of the information contained in simple descriptive
	statements. "The sky is a blue color" entails that blue is a
	color, that the sky has a color, that the sky is blue, that the
	sky is visible, and that the sky is located both spatially and
	temporally. These generalizations are so trivial that they border
	on presuppositions, and they have consequently been taken for
	granted in semantic nets and frames. Making such information
	explicit greatly increases the density and usefulness of stored
	knowledge. One interesting application is to disambiguate an
	adjective/predicate to suit a given noun/extension.

	My representation is parsimonious, having O(three) primitive con-
	structs (link types); is highly irredundant, since blue(sky) and
	color(blue) reference the same blue; and is static, being inten-
	ded to formalize and implement "massively parallel" deterministic
	connectionist question-answering systems. Predicates, relations,
	and simple forms of quantification all emerge as by-products of
	function applicability and set inclusion. Viewed as a logic the
	representation is potentially O(w), intensional and inconsistent.

	The talk will touch on issues in philosophy of logic and linguistics.
	I will especially appreciate constructive criticism in those areas,
	as I am a novice there.

-------

∂23-Sep-85  0342	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #18  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 23 Sep 85  03:41:56 PDT
Date: 23 Sep 85 0327-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #18
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 23 Sep 1985       Volume 1 : Issue 18

Today's Topics:
                    Another overlooked IJCAI abstract
                  Parallel symbolic processing in Japan


----------------------------------------------------------------------

Date: Sun 22 Sep 85 20:08:03-EDT
From: Dave.Touretzky@C.CS.CMU.EDU
Subject: another overlooked IJCAI abstract

		 Symbols Among the Neurons:
	Details of a Connectionist Inference Architecture

	    David S. Touretzky and Geoffrey E. Hinton
		 Computer Science Department
		  Carnegie Mellon University

ABSTRACT:

Pattern matching and variable binding are easily implemented in
conventional computer architectures, but not necessarily in all
architectures.  In a distributed neural network architecture each symbol
is represented by activity in many units and each unit contributes to
the representation of many symbols.  Manipulating symbols using this
type of distributed representation is not as easy as with a local
representation where each unit denotes one symbol, but there is evidence
that the distributed approach is the one chosen by nature.  We describe
a working implementation of a production system interpreter in a neural
network using distributed representations for both symbols and rules.
The research provides a detailed account of two important symbolic
reasoning operations, pattern matching and variable binding, as emergent
properties of collections of neuron-like elements.  The success of our
production system implementation goes some way towards answering a
common criticism of connectionist theories: that they aren't powerful
enough to do symbolic reasoning.

------------------------------

Date: Mon, 23 Sep 85 14:36:49 EST
From: munnari!aegir.oz!crw@seismo.CSS.GOV (charles watson)
Subject: PARALLEL SYMBOLIC PROCESSING IN JAPAN

PLEASE FORWARD TO INTERESTED PARTIES

                 JAPAN TRIP REPORT - 13 Aug 85 to 3 Sep 85

                            Charles Watson
                   Division of Information Technology
                  CSIRO, PO Box 4, Woodville, SA 5011


        This report  describes  research  at  12  Japanese  laboratories
visited  by  the  author. The main goal of the visit was to see Japanese
research in the areas of parallel symbolic computing hardware, and  VLSI
design tools and methodology.


THE INSTITUTE FOR NEW GENERATION COMPUTER TECHNOLOGY (ICOT)

        ICOT was started in 1982 as the flag-ship for the Japanese Fifth
Generation Computer Systems (FGCS) project. Researchers (with an average
age  under  30)  are  drawn  from  Japanese  companies  for  short terms
(typically 2 or 3 years). The  high  turnover  of  researchers  provides
rapid dissemination of expertise to industry.

        Shinichi  Uchida  introduced  the  work at ICOT and demonstrated
some prototype hardware. ICOT has about 10 Personal Sequential Inference
(PSI)    machines  for  FGCS  research  ranging  from  natural  language
translation to operating system development. Each PSI has 80  Mbytes  of
main memory using 256K DRAMs on 16 printed circuit boards (PCBs), and 11
PCBs of MSI TTL for the CPU. They are currently  implementing  a  single
board CPU with gate arrays.

        They  have  a  large  relational  database  machine  (DELTA) for
research  in  knowledge  acquisition  and  retrieval,    and    parallel
processing.  ICOT  is  responsible for co-ordinating many other research
projects  in  the  major  Japanese  electronics  companies,   and    the
Universities.  These  include  several parallel inference machines (PIM)
using either dataflow or reduction as the computational  mechanism,  and
an  implementation  (called  CHI)  of Warren's recent Prolog instruction
set.


OKI ELECTRIC INDUSTRY CO. LTD. (SYSTEMS LABORATORY)

        Hiroshi Yasuhara is manager of the AI Section. He has  published
work  on  speech  recognition  and  an  abstract Prolog machine based on
bundled OR-parallelism. His group is now working on the operating system
(SIMPOS)  for the PSI machine, and on parallel inference. Their parallel
inference machine (PIM-D) is data-flow based, and  uses  a  hierarchical
bus structure for passing instruction packets and data. It utilizes both
AND and OR parallelism. Their prototype has two clusters connected by  a
token  bus.  Each  cluster  contains  4  processing elements (PEs) and 4
structure memories (SMs) and a local bus. Each PE uses  8  PCBs  of  TTL
(eg.  AMD  2901  bit-slice)  and  each SM uses 3 PCBs. Simulation of the
PIM-D shows reduced utilization for more than 16 PEs.  This  is  due  to
limited bus bandwidth, and limited parallelism in the test programs. The
simulation of 64 PEs took 4 hrs CPU on a VAX 11/780.


ELECTROTECHNICAL LABORATORY (ETL)

        ETL is the largest national research organization in Japan, with
560  researchers  in  the  areas of energy, measurement, electronics and
information processing. It has been  located  at  Tsukuba  Science  City
since  1979.  Motoi  Suwa  is  chief of the Man-Machine Systems Section.
Their work includes image and speech  processing,  and  knowledge  based
systems.

        Toshitugu  Yuba  is  chief of the Computer Architecture Section.
They have a data-flow computer for scientific computation (SIGMA-1). The
single  PE  prototype  has  6  PCBs  of TTL and is being redesigned as a
single board of gate array chips. Their target for 1987 is a 100  Mflops
MIMD  machine  with about 100 PEs. They also have an 8 PE prototype of a
data-flow Lisp machine (EM-3) based on MC68000 processors. This is being
used as test-bed for new parallel control mechanisms. For example, the 4
queens problem took 10  seconds,  and  10,000  tokens  to  produce  both
solutions,    giving  a  utilization  of  70%  for  8  PEs.  The  target
implementation for 1987 is a 500 KLips, 100 processor Lisp  machine.  It
will  utilize  parallel  evaluation of conditionals (COND statement) and
concurrent procedure calls.


NTT ELECTRICAL COMMUNICATION LABORATORY (MUSASHINO)

        NTT  provides  most  of  Japan's  domestic   telecommmunications
services.  Ryuzo Hasegawa has a small research group working on parallel
execution of Logic Programs. They have a  single  PE  TTL  prototype  of
their data flow machine (DFM), and plan a gate-array implementation. The
DFM applies lenient CONS mechanism for eager and lazy evaluation of list
structures.  Dynamic  load  balancing  is  applied  at all levels of the
network hierarchy. They have simulator for  DFM  that  gives  a  dynamic
graphic representation of token flow.

        Other  work  at  Musashino  includes  robotics,  recognition  of
hand-drawn  Chinese  and  Japanese    characters,    natural    language
translation, and speech processing.


DEPT. OF ELECTRICAL ENGINEERING, UNIVERSITY OF TOKYO

        They have a prototype relational data-base machine called GRACE,
and a parallel inference engine (PIE). The PIE prototype  has  a  single
PE. They plan to build a 256 PE PIE machine when funds become available.
Software simulation predicts a utilization of 60% to  70%  It  will  use
OR-parallelism.


DEPT. OF COMPUTER SCIENCE, TOKYO INSTITUTE OF TECHNOLOGY

        Naoyoshi  Tamura  demonstrated  a  bottom-up  parser for English
sentences. It uses attribute grammars, to distinguish context  dependent
semantics.  All possible parsings for ambiguous sentences are generated.

        Tomohiro Yoneda has developed a fault  tolerant  multiprocessor.
His prototype uses 4 busses and 4 PEs. Computations are triplicated, and
the 4th PE is a hot spare to be switched in when the  three  active  PEs
have  differing  results.  The  faulty  PE  becomes  the  new spare when
repaired. Data transfer is also fault  tolerant,  using  triple  modular
redundancy (TMR) and a spare bus.


DEPT. OF APPLIED PHYSICS, OSAKA UNIVERSITY

        Hiroshi  Yasui  has  developed  a  multiprocessor  LISP  machine
(EVALIS). The prototype has 2 PEs and took about a man  year  to  build.
Utilization  is  over  90%  for  2  PEs.  The 8 queens problem took 6.31
seconds on a single PE and 3.31 seconds on 2 PEs.  Lists  are  evaluated
concurrently by spawning NCONC evaluations. Microcode can also be loaded
for diagnostics and Prolog interpretation.  An  interesting  feature  is
special CAR and CDR registers for rapid access to LISP structures.


DEPT. OF INFORMATION SCIENCE, KYOTO UNIVERSITY

        Shinji  Tomita has developed a microprogrammable multi-processor
(QA-2). The prototype has 4 arithmetic logic units (ALUs), and is  built
from  17,000  ECL  and  Schottky  TTL  ICs. It is used as a test-bed for
several projects, including a parallel 3-D graphics processor, pipelined
arithmetic  calculation,  and  a  parallel  reduction machine for Prolog
(KPR). KPR applies OR-parallelism and pipelined AND-parallelism, and  is
demand  driven.  A  prototype  is under construction. Several sequential
Prolog machines have also  been  emulated  on  the  QA-2,  including  an
implementation of Tick & Warren's pipelined processor.

        Other  work in the department includes arithmetic hardware using
redundant binary representation, circuit simulators,  and  computational
complexity.  The  data  processing centre has a 250 Mflops VP-100, and a
FACOM M-382 main-frame.


NTT ELECTRICAL COMMUNICATION LABORATORY (ATSUGI)

        NTT have done  some  excellent  work  in  VLSI  Design,  and  IC
fabrication technology. Technological innovations include a self aligned
process for 1 micron bipolar  gates  with  30  picosecond  gate  delays;
implanted  oxide  isolation; wafer planarization for multi-layer wiring;
and trench capacitors, used now in most commercial  Japanese  1  megabit
DRAMs.  INTEL,  the  leading US commodity chip manufacturer is trying to
compete in the  1  megabit  DRAM  market  with  low  density  horizontal
capacitors. No wonder the Americans are loosing the memory market to the
Japanese. Their clean rooms have overhead tracks for wafer transport  to
minimize human contact.

        In  the  area of VLSI Design they have a floorplanner (CHAMP), a
logic    synthesis    system    (ANGEL),    integrated    into     their
standard-cell/gate-array    design   environment.  They  have  an  array
processor (AAP) with 65,536 PEs, for auto  place  and  route  and  logic
simulation.

        They   also  have  a  prototype  Prolog  machine  using  content
addressable memory (CAM)  chips  for  binding  variables.  The  expected
performance is 100 kLips.


NEC VLSI DEVELOPMENT DIVISION, KAWASAKI

        NEC  has  a  2-D  gate-array  compaction  system  (COCKTAIL),  a
symbolic  layout  system  (ADULTS)  for  automatic  cut  selection   and
non-orthogonal  grid-free  compaction,  and  a  PC-based schematic entry
system (VISTAS).


TOSHIBA R&D CENTER, KAWASAKI

        Tadashi Shibata showed me the clean rooms used to develop  their
1  Megabit DRAMs. They have excellent quality control. When the 1 M DRAM
was transferred from a development  fabrication  line  to  a  brand  new
production  line  the initial yield was 10% This compares with new lines
in the past taking as much as a year to come up to a workable  yield.  I
saw  plasma etched line widths of 0.1 microns, and E-beam testing of ICs
with 0.1  micron  positioning  resolution  and  a  20  mV  signal  level
resolution.  This  enables  analog transients at the edge of a submicron
gate to be measured. They are also  developing  robot  controlled  wafer
transport and processing.

        In  the  area of VLSI design, they have a logic synthesis system
based on a hierarchical hardware description language, and  a  procedure
for  placing  processors  and  memories on a single chip with gate array
glue logic.


FUJITSU LABORATORIES, INFORMATION PROCESSING DIVISION, KAWASAKI

        Fujitsu's FACOM alpha is a commercial LISP/Prolog machine with a
hardware stack, costing about $100,000, and with 3 times the performance
of the Symbolics 3600. They also have a  prototype  OR-parallel  machine
for  executing  pure  Prolog.  They  have  an  array  processor for high
performance  parallel  graphics.  Other    work    includes    character
recognition, VLSI Design, expert systems, and robotics.


CONCLUSION

        I  found  the Japanese researchers very courteous and open about
discussing research ideas, but sometimes the  language  barrier  limited
communication.  For  many  researchers, translating work into English is
their hardest task. Speaking only English, I probably saw just  the  tip
of  the  iceburg.  I was pleasantly surprised by the high quality of the
computing research I did see. Considering the Japanese ability for rapid
commercialization  of  technical innovation, we can expect some exciting
developments in the next few years. It is paradoxical  that  the  highly
competitive  Japanese  electronics  companies seem to co-operate so well
with government agencies such as MITI and ICOT. There is a  lot  we  can
learn from the Japanese.
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
Charles Watson, CSIRO, PO Box 4, Woodville, S.A. 5011, Australia.

			UUCP:  {decvax,pesnta,vax135}!mulga!aegir.dmt.oz!crw
PHONE: +61 8 268 0137	ARPA:  crw%aegir.dmt.oz!crw@seismo.arpa
			CSNET: crw@aegir.dmt.oz

------------------------------

End of PARSYM Digest
********************

∂23-Sep-85  0651	PATASHNIK@SU-SUSHI.ARPA 	AFLB resumes   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 23 Sep 85  06:51:14 PDT
Date: Mon 23 Sep 85 06:47:54-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: AFLB resumes
To: aflb.all@SU-SUSHI.ARPA

     AFLB starts the year with Umesh Vazirani as our speaker on
October 3 and Eli Shamir on October 10.  I'll send abstracts early
next week, but in the meantime please let me know if you'd like to
speak later this quarter.  The open dates are October 17, October 31,
November 7, November 14, November 21, December 5, and December 12.
Also let me know if you'd like your address in the mailing list
changed or if you know anyone who'd like to be added to the list.
Send requests to me at aflb-request@su-sushi.arpa.
	--Oren
-------

∂23-Sep-85  0847	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@wisc-rsch.arpa 	Reminder, COming up this Friday    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 23 Sep 85  08:47:10 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Mon 23 Sep 85 08:42:08-PDT
Received: from wisc-rsch.arpa by SU-SCORE.ARPA with TCP; Mon 23 Sep 85 08:42:23-PDT
Received: by wisc-rsch.arpa; Mon, 23 Sep 85 10:16:03 cdt
Received: from wisc-crys.arpa by wisc-rsch.arpa; Mon, 23 Sep 85 09:13:23 cdt
Message-Id: <8509231413.AA08979@wisc-crys.arpa>
Received: from cs.columbia.edu by wisc-crys.arpa; Mon, 23 Sep 85 09:13:13 cdt
Date: Mon 23 Sep 85 10:09:48-EDT
From: Zvi Galil <GALIL@COLUMBIA-20.ARPA>
Subject: Reminder, COming up this Friday
To: theory@WISC-CRYS.ARPA
Sender: udi@wisc-rsch.arpa
Resent-From: udi@wisc-rsch.arpa (Udi Manber)
Resent-To: theory:;



THE
SEVENTH THEORY DAY
at Columbia University

Sponsored by the Department of Computer Science

FRIDAY, SEPTEMBER 27, 1985


10:00   PROFESSOR JOHN E. HOPCROFT
        Cornell University

	"MATHEMATICAL PROBLEMS IN OBJECT
	 REPRESENTATION SYSTEMS"


11:00	PROFESSOR LASZLO LOVASZ
	Eotvos Lorand University, Budapest, Hungary

	"MATCHINGS, POLYHEDRA,
	 LATTICES, AND ALGORITHMS"


 2:00	DR. FAN K.R. CHUNG
	Bell Communications Research

	"DYNAMIC SEARCH ON GRAPHS"


 3:00	PROFESSOR TOM LEIGHTON
	Massachusetts Institute of Technology

	"WAFER-SCALE INTEGRATION AND
	 THE GRID RECONSTRUCTION PROBLEM"


Coffee will be available at 9:30 a.m.
All lectures will be in the Kellogg Conference Center on the
fifteenth floor of the International Affairs Building, 118th
Street and Amsterdam Avenue.
The lectures are free and open to the public.
Call (212) 280-2736 for more information.
-------
-------

∂23-Sep-85  1005	RICHARDSON@SU-SCORE.ARPA 	Sr. Faculty Meeting
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 23 Sep 85  10:05:02 PDT
Date: Mon 23 Sep 85 09:57:47-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Sr. Faculty Meeting
To: tenured@SU-SCORE.ARPA
Message-ID: <12145578945.18.RICHARDSON@SU-SCORE.ARPA>

There will be a sr. faculty meeting following the general faculty meeting
at 2:30 on September 24 in Margaret Jacks Hall Conference Room 146.

Anne
-------

∂23-Sep-85  2016	PATASHNIK@SU-SUSHI.ARPA 	[Andrew Yao <YAO@SU-SCORE.ARPA>: Lower Bound Course]   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 23 Sep 85  20:16:31 PDT
Date: Mon 23 Sep 85 20:13:11-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: [Andrew Yao <YAO@SU-SCORE.ARPA>: Lower Bound Course]
To: aflb.su@SU-SUSHI.ARPA

Date: Mon 23 Sep 85 12:14:35-PDT
From: Andrew Yao <YAO@SU-SCORE.ARPA>
Subject: Lower Bound Course
To: aflb.su@SU-SUSHI.ARPA
cc: yao@SU-SCORE.ARPA
Message-ID: <12145603850.28.YAO@SU-SCORE.ARPA>


In this year's CS 366, the emphasis will be on Boolean circuit size complexity.
We will discuss the past and recent results on this subject, and its relations
with Turing computational complexity.

The first meeting will be on Thursday (Sept. 26) as scheduled in the class
schedule booklet.  It will be just an organizational meeting, mainly to
fix the meeting time for the rest of the quarter.  Tentatively, I would
like to have the subsequent meetings on MW 1:00-2:15 in Margaret Jacks
Hall 301.  The reason to make the change of meeting time is, as pointed
out by several people, that many of you may wish to attend the various
talks at MSRI on Tuesdays and Thursdays.

Thanks.

--Andy
-------
-------

∂24-Sep-85  1119	HAUNGA@SUMEX-AIM.ARPA 	Title and abstract for the SIGLUNCH, September 27th.
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  11:19:44 PDT
Date: Tue 24 Sep 85 11:10:07-PDT
From: Ana Haunga <HAUNGA@SUMEX-AIM.ARPA>
Subject: Title and abstract for the SIGLUNCH, September 27th.
To: SIGLUNCH@SUMEX-AIM.ARPA
Message-ID: <12145854258.59.HAUNGA@SUMEX-AIM.ARPA>




                     KSL/Symbolic Systems Resources Group

                        Tom Rindfleisch and Bill Yeager
                              Stanford University

This is the first of several SIGLunches this fall that will summarize work in
each of the five sublabs of the Stanford Knowledge Systems Laboratory (KSL),
including the Heuristic Programming Project, HELIX Group, Medical Computer
Science Group, Logic Group, and Symbolic Systems Resources Group (SSRG).  This
week's talk will consist of a brief overview of the KSL as an AI laboratory and
a survey of SSRG research and development activities.

Since 1980, the computing environment for KSL research has been moving slowly
away from central time-shared mainframes (like the SUMEX 2060 and VAX) toward
networked Lisp workstations.  Improvements in workstation performance, falling
prices, better packaging, and a wider vendor selection are now accelerating
this transition.  Over the next five years, we are proposing to phase out the
SUMEX research mainframes so that all KSL computing will be workstation-based
-- not only research program development but common tasks like text processing,
mail, file management, and budgeting.  This raises several important issues
that will require a community system software effort comparable to that in the
1970's that led to the current TOPS-20 and UNIX environments.

How can the user computing environment be improved using workstation bitmapped
graphics and AI methods for more intelligent systems/applications programs?

How can user displays connect flexibly to workstations -- from home, over
remote networks like ARPANET, and locally over Ethernet?

How can the considerable computing power distributed among many workstations be
combined to support individual user tasks?

What are the impacts on network protocols and services (file servers, gateways,
printing, etc.) of large numbers of workstations?
-------
-------

∂24-Sep-85  1144	REULING@SU-SCORE.ARPA 	LOTS Registration
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  11:44:07 PDT
Date: Tue 24 Sep 85 11:09:48-PDT
From: John Reuling <Reuling@SU-SCORE.ARPA>
Subject: LOTS Registration
To: tas@SU-SCORE.ARPA
cc: instructors@SU-SCORE.ARPA, fl%LOTS-A@SU-SCORE.ARPA
Office: 206 Margaret Jacks, 415/497-3549
Message-ID: <12145854201.24.REULING@SU-SCORE.ARPA>

TAs-

If your course will be using the LOTS Computer Facility this quarter
and you have not yet registered with LOTS, please stop by my office
TODAY and fill out a "Class Registration" form.  I will deliver the
completed forms to LOTS this afternoon.

Miscellaneous administrative requests (e.g. setting up class directories,
adding TA's to user groups, etc.) should be directed to FL@LOTSA (Faculty
Liaisons) AFTER you have turned in your registration form.

I am in Margaret Jacks 206.

-John Reuling
-------

∂24-Sep-85  1144	@SU-CSLI.ARPA:KAELBLING@SRI-AI.ARPA 	Course Announcement    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  11:44:32 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 24 Sep 85 11:39:24-PDT
Date: Tue 24 Sep 85 11:38:08-PDT
From: Leslie Kaelbling <KAELBLING@SRI-AI.ARPA>
Subject: Course Announcement
To: friends@SU-CSLI.ARPA


                      THE LOGIC OF ROBOT DESIGN


This course will explore theoretical issues in the design of software
for intelligent agents.  Its aim is to provide conceptual tools for
coping with complexity in robot design, covering processes from the
sensorimotor level through reasoning, planning, and linguistic
communication, emphasizing the role of formal methods in analysis and
synthesis of robot software.

The following topics will be covered:
  - applications of epistemic and temporal logic to robotics
  - automata-theoretic models of knowledge
  - inference and planning
  - logic-based tools for programming intelligent robots.

Some familiarity with basic logic and computer programming will be
assumed.  Coursework will consist of problem sets and one programming
assignment.

Instructor    : Stan Rosenschein   (stan@sri-ai; 859-4167)
Time          : TTh 11-12:15
Place         : e208
Course number : CS428
Units         : 3
TA            : Leslie Kaelbling   (kaelbling@sri-ai, pack@su-sushi,
				    k.kaelbling@su-lots-b; 859-2578)

-------

∂24-Sep-85  1209	@SU-SCORE.ARPA:TW@SU-AI.ARPA 	Program for this year's computer forum  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  12:09:06 PDT
Received: from SU-AI.ARPA by SU-SCORE.ARPA with TCP; Tue 24 Sep 85 11:52:21-PDT
Date: 24 Sep 85  1153 PDT
From: Terry Winograd <TW@SU-AI.ARPA>
Subject: Program for this year's computer forum  
To:   faculty@SU-SCORE.ARPA
CC:   forum-committee@SU-SCORE.ARPA    

After a good deal of discussion, we have decided to hold this year's
computer forum on campus, even though the available rooms are not adequate
for its steadily growing size.  As a result, we will not stick to our old
single-track format, but will make creative use of parallel sessions,
special-interest sessions (e.g., perhaps at CIS or with KSL), etc.  In
order to plan the schedule we need to get started early finding out what
presentations there might be.  Of course, there will still be room for
some changes, but anything that doesn't get into the original plan has no
guarantee of finding space and/or time.  The forum will be held on
February 5 and 6.

My request to each of you is that by October 11 you let me know the
following:

    What students (with corresponding topics) would you like to have speak?
    Preference goes to PhD students who will be finishing before the following
    forum meeting, and who have not spoken at a forum before.

    Which faculty members would like to make individual presentations?  In the
    past this has been limited to new faculty, for whom it is an introduction to
    the forum members, or for faculty speaking on behalf of newly created
    groups whose existence is of interest to the members.

    What special sessions/events/programs might you be interested in hosting
    (i.e, organizing)?  What demands would they have for space and time?

Our lack of space may turn out to be a benefit, rather than a problem, if
it helps us get out of the rut and think of new ways to make the forum
better for both the industrial members and our own students and faculty.
I'm counting on your help to do that.

--t

∂24-Sep-85  1317	MARJORIE@SU-CSLI.ARPA 	CSLI Computing pamphlet    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  13:17:24 PDT
Date: Tue 24 Sep 85 13:12:52-PDT
From: Marjorie Maxwell <MARJORIE@SU-CSLI.ARPA>
Subject: CSLI Computing pamphlet
To: folks@SU-CSLI.ARPA

There is a new CSLI Source pamphlet (explaining the DEC system, Turing,
getting help, MM, & editors) prepared by Susana Wessling, our consultant,
and edited by the rest of the computer staff.  It is helpful for new users
and is really just an introduction. I am currently giving it out to the
new accounts but if anyone else is interested in getting a copy, see me
at G-4 in the trailers.
Marjorie
-------

∂24-Sep-85  1346	LIBRARY@SU-SCORE.ARPA 	Socrates:  New Forms, What To Do If You Misplaced Your Old Account 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  13:45:12 PDT
Date: Tue 24 Sep 85 13:41:36-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates:  New Forms, What To Do If You Misplaced Your Old Account
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12145881834.9.LIBRARY@SU-SCORE.ARPA>


I have received the new forms for free Socrates accounts.  It you already have
an ITS account, you can search Socrates at no cost on your ITS account.  If
you don't have an ITS account you will want to come to the library, fill out
a form, and you will then be given an account that will already be activated.
Since it is the beginning of the year, it may be best if you send me an 
electronic note that you will be coming to the library to get a form.  That
was I can attempt to have enough forms on hand at all times.  Just send a
message with the Subject: Socrates Account and include your name etc. 

If you had an account last year and have misplaced your information.  You
may call 497-0020 and Richard Clark will verify your name and ID number 
and give you your account information.  

The enhanced version of Socrates will not be up this week.  It may be up
next week.

Harry Llull
-------

∂24-Sep-85  1400	BSCOTT@SU-SCORE.ARPA 	Salaries
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  14:00:15 PDT
Date: Tue 24 Sep 85 13:55:41-PDT
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: Salaries
To: Faculty@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12145884397.45.BSCOTT@SU-SCORE.ARPA>


This is just for CS faculty who are being paid from CS funds:  would you all
please check your 9/22 salary payments.  We changed account numbers and 
salaries and it will be a big help to me if you will make sure you received
the amount of salary you were expecting.  There is no need to respond if your
salary was o.k., but I will much appreciate hearing from anyone whose salary
is not as expected.

Thank you.

Betty
-------

∂24-Sep-85  1412	DAVIES@SUMEX-AIM.ARPA 	No architecture meeting this week    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  14:12:24 PDT
Date: Tue, 24 Sep 1985  14:07 PDT
Message-ID: <DAVIES.12145886464.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No architecture meeting this week
cc:   Davies@SUMEX-AIM.ARPA

There will be no meeting this week.

	-- Byron

∂24-Sep-85  1544	LIBRARY@SU-SCORE.ARPA 	Math/CS Library:  New Books
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  15:44:31 PDT
Date: Tue 24 Sep 85 15:40:59-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library:  New Books
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12145903567.9.LIBRARY@SU-SCORE.ARPA>

A Geometric Investigation of Reach. ACM Distinguished Dissertation. 1984.
MIT Press. by James Korein.

Numerical Grid Generation: Foundations and Applications. by Thompson,
Warsi, Mastin.  Norht-Holland. QA377.T524 1985.

Design and Analysis of Distributed Real-Time Systems. by Paul Fortier.
Intertext Publ. McGraw-Hill. (8512417)

Operating Systems: Structures and Mechanisms. by Philippe Janson.
Academic Press. QA76.6.J364 1985.

Formal Theories Of The Commonsense World. by Jerry Hobbs and Robert
Moore editors SRI International.  Ablex Series In Artificial Intelligence.
Q360.F66 1985 c.2

SPSS Graphics. SPSS Inc. HQ32.S62 1985

H. Llull
-------

∂24-Sep-85  1722	@SU-CSLI.ARPA:KAELBLING@SRI-AI.ARPA 	Room Change  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  17:22:16 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 24 Sep 85 17:17:14-PDT
Date: Tue 24 Sep 85 17:18:22-PDT
From: Leslie Kaelbling <KAELBLING@SRI-AI.ARPA>
Subject: Room Change 
To: friends@SU-CSLI.ARPA


The initial meeting of The Logic of Robot Design will be in MJH252
(Building 460), not e308 as previously announced.
-------

∂24-Sep-85  1941	@SU-SCORE.ARPA:fournier@su-navajo.arpa 	Re:  Socrates:  New Forms, What To Do If You Misplaced Your Old Account    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  19:41:09 PDT
Received: from su-navajo.arpa by SU-SCORE.ARPA with TCP; Tue 24 Sep 85 19:36:10-PDT
Received: by su-navajo.arpa with Sendmail; Tue, 24 Sep 85 19:36:12 pdt
Date: Tue, 24 Sep 85 19:36:12 pdt
From: Alain.Fournier@su-navajo.arpa, 226C@su-navajo.arpa, 0474@su-navajo.arpa,
        2472276 <fournier@su-navajo.arpa>
Subject: Re:  Socrates:  New Forms, What To Do If You Misplaced Your Old Account
To: LIBRARY@SU-Score, su-bboards@SU-Score
Cc: faculty@SU-Score

I would like a Socrates account. I am a visiting prof. Does that make a difference?
I'll go by the library probably tomorrow. Thanks.

∂24-Sep-85  1951	@SU-SCORE.ARPA:fournier@su-navajo.arpa 	Re:  Socrates:  New Forms, What To Do If You Misplaced Your Old Account    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Sep 85  19:51:34 PDT
Received: from su-navajo.arpa by SU-SCORE.ARPA with TCP; Tue 24 Sep 85 19:48:13-PDT
Received: by su-navajo.arpa with Sendmail; Tue, 24 Sep 85 19:48:14 pdt
Date: Tue, 24 Sep 85 19:48:14 pdt
From: Alain.Fournier@su-navajo.arpa, 226C@su-navajo.arpa, 0474@su-navajo.arpa,
        2472276 <fournier@su-navajo.arpa>
Subject: Re:  Socrates:  New Forms, What To Do If You Misplaced Your Old Account
To: LIBRARY@SU-Score, su-bboards@SU-Score
Cc: faculty@SU-Score

Sorry about unnecessary message. I forgot to shift.

∂25-Sep-85  0104	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #37
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  01:04:27 PDT
Date: Saturday, August 24, 1985 1:32PM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #37
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Monday, 26 Aug 1985       Volume 3 : Issue 37

Today's Topics:
                         Query - Simulation,
                Programming - Building Large Systems,
                  LP Philosophy - Hewitts Challenge,
                    Implementation - Performance,
                   LP Library - Fix & Public Domain
----------------------------------------------------------------------

Date: Mon 12 Aug 85 20:11:56-PDT
From: Larry Widman <WIDMAN@SUMEX-AIM.ARPA>
Subject: Simulation

I am interested in semi-quantitative simulation as an
approach to generating scenarios of causal networks
which include feed-back loops.  The scenarios could
then undergo feature extraction to permit temporal
reasoning and learning-from-experience by pattern
matching of  specific scenarios with a library of
ideal scenarios ("compiled knowledge") and of
previously seen cases ("experience").

I have implemented such a program in MACLISP for use in
the medical domain.  I presented it at the latest meeting
of the Artificial Intelligence in Medicine meeting in
Bethesda, in July.  I should think that this approach
would be generally useful, and may well be under independent
investigation by others.

Therefore, I would like to read the logic programming
literature (the Digest being among the best, according
to my colleagues).  If you think it is appropriate, you
could insert the above paragraphs into the Digest together
with a request for information from others with similar
interests.

I would of course be happy to summarize the responses for
the Digest.

Thank you for your help.  -- Larry Widman

------------------------------

Date: Sun, 18 Aug 85 22:43:03 +0200
From: Anjo@seismo.CSS.GOV
Subject: Building Large Systems

   We have had a similar problem with developing a window system
with a large number of utilities on top of Prolog.  As the window
system should have a fast user interface (responding to mouse
clicks immediately), Prolog itself is not the most appropriate
language to use.  Our solution is implement the user interface
and all the windowing stuff in C, along with text-editors, graphics
editors, the mouse-manager etc. (at the moment this is about 300+
pages of C source code).  This package is called PCE.  Obviously
the Prolog programmer wants full access to this package, he wants
to create windows, graphical representations of knowledge inside
his Prolog application and get mouse clicks if he needs them.
This calls for bi-directional interface between PCE and Prolog.
Our solution for this problem is to add a small number of
predicates to Prolog that implement object orientedness, and a
few functions that can asserted information available in PCE in
the Prolog data-base (total source code of these functions is
about 15 pages C).

   The advantages of this approach (in our case) are that the
performance of the user interface doesn't depend on the performance
of the Prolog interpreter, so all I/O functions are as fast as C can
do it.  The small interface assures that PCE can be modified without
also needing to change part in the interface.  For example if we
decide to create another lay-out for a window only PCE needs to be
changed, while both the interface between Prolog and PCE, and the
Prolog applications remain unchanged.

   Whether an approach like this is also feasible for your problem
depends, I think, on how well you succeed in keeping the interface
between Prolog and the "Data Base System (DBS)" simple.  From a
software engineering point of view the saying "small is beautiful"
certainly applies here.  I think that it is possible to build an
object oriented shell around a DBS, in that case the approach we
have used becomes possible.  A dangerous approach is to add
predicates to Prolog for each operation you need, before long the
interface will be large and porting it to another implementation
of Prolog will for sure give problems.  If we need to port PCE
Prolog to another implementation only 15 pages of C-code needs to
be rewritten, which is acceptable.

   The Prolog implementation we use is C-Prolog 1.5 with some local
additions.  If you would like to have more information or technical
details (the predicates used for instance) let me now it.

Greetings,

-- Anjo Anjewierden

------------------------------

Date: Mon, 19 Aug 1985  12:47 EDT
From: Hewitt@MIT-MC.ARPA
Subject: Prolog will fail as the foundation for AI

Date: Saturday, 17 August 1985  13:36-EDT
From: PEREIRA at SRI-AI.ARPA
Re:   Prolog will fail as the foundation for AI

Carl Hewitt's message is based on several misconceptions:

1. (the least interesting one) All the so-called commercially
viable Prolog systems in Lisp are not really Prolog systems
written IN Lisp, but rather Prolog systems written FOR Lisp
machines. Or is it that a microcode sublanguage or Lisp machine
pointer-smashing operations are part of List as we know it?

Yes.  They are DEFINITELY part of Common Lisp as we know it
being implementations of reading and writing operations on
record structures.  Such implementation methods are NOT part
of Logic as a Programming language.

Without those machine-level operations, those Prolog systems
would run too slow and use too much memory to be useful for
serious Prolog programming. From the Prolog implementation
point of view, what is important about the Lisp machines is
not that they run Lisp, but that they can be microcoded and
have good support for tagged data types and stack operations.

It is important to many users that they can make use of ALL
the software available to the community and not just be limited
to the tiny amountin Prolog.  Furthermore in the future good
software will be ported from stand alone Prolog systems to Prolog
implemented on Lisp.  However to good Lisp software will not be
able to be ported to the stand alone Prolog systems.

2. If the decisions (actions) of a system are not determined
by its inputs, the system is nondeterministic. Nondeterminism
in a system can be either an artifact of our incomplete knowledge
(or lack of interest) of the detailed operation of the system;
or it can be ``real physical'' nondeterminism. It would take us
to far to discuss whether the second kind of nondeterminism is
``real'' or also an artifact. In any case, most uses of
nondeterminism, say in models of concurrency, are of the first
kind, and can be expressed appropriately in various temporal
dynamic logics. Admittedly, these are not Prolog, but then
Common Lisp is not Lisp 1.5! (Prolog is 13 years old, Lisp 25).

Yes indeed there is a large problem here that poses fundamental
problems for using Logic as a Programming language to make
decisions in Open Systems.

3. The first logic course dictum ``from a contradiction one
can conclude anything'' is getting in the way. Notice that
the dictum says ``can'', not ``must''. There is an enormous
difference between things that are in principle true and
things that an agent knows to be true in a way that can affect
its action. An agent might know ``p'' and ``not p'', but it
might well never come to infer the dreaded totally unrelated
``q'' which IN PRINCIPLE follows from the contradiction. This
inference might not happen either because of inference control
mechanisms of the agent (eg. it uses the set-of-support strategy)
or because the agent's logic is just TOO WEAK to conclude
anything  from a contradiction (vide Hector Levesque's work in
the proceedings  of the last AAAI). In any case, Horn clauses
(the basis of Prolog)  are too weak to represent contradictions
... :-)

I claim that in practice the contradictions lie close to the
surface and occur in any nontrivial application.  Thus the
contradictions pose fundamental problems for using Logic as
a Programming Language.

4. The question of whether ``taking action'' fits in a logic
paradigm tends to be answered negatively after an hour's worth
of consideration.  If you persist for several years, though,
this question becomes a source of insight on the relations
between knowledge, state and action that is not available to
those who started by dismissing the question after that initial
hour. There is just too much work on logics of knowledge and
action in AI and computer science for me to try to discuss it
here. Some of this work has been applied to logic programming,
either in the form of new logic programming languages based on
temporal or dynamic logics or in representations of temporal
reasoning and decision in, say, Prolog.

I have been thinking about the problem for many years having
designed Micro-Planner, the first "procedural embedding of
logic" programming language in 1967.  I claim that the problem
of taking action poses fundamental problems for using Logic as
a Programming language.

5. It is curious to see someone by implication defend Lisp as a
language for expressing the taking of action!

I claim that current Lisp systems are better than current Prolog
systems for taking action because the only ways to take action in
current Prolog systems are kludges.


We know by now that the most difficult issue in ``reactive
systems'' is not EXPRESSING action, but rather keeping it under
control to prevent unwanted interactions. In this area, less
is REALLY more, and highly complex languages like Lisp are
just not suitable for the formal reasoning about programs
that is needed to help us believe our programs do what we
intend. To those who say ``...but we can keep to a carefully
constrained subset of Lisp, use only message passing for
interactions,...'' I will answer that the history of all large
Lisp programs I know (and I think that is a representative sample)
is littered with the brutalized corpses of constrained
programming styles. Anyone who has looked at the current
flavor mechanism in Zetalisp and its use in the window system
will know what I mean...

5. Underlying Carl Hewitt's misconceptions is the old chestnut
``logic is static, systems are dynamic''.

Note that the above quotation is NOT anything that I said.

Any language, be it first-order logic or Lisp, is static;
it is its USE which is dynamic (changes the state of
communicating agents). A good analogy here is the use of
differential equations to model dynamic systems in classical
mechanics. The differential equations themselves do not say
directly what happens when (they are ``static'' in Hewitt's
jargon).

I do not deny that dynamic systems can be DESCRIBED using
logic only that they can be CONTROLLED.

It is their SOLUTIONS that tell us the sequence of events.
Even the solutions are given as static objects (functions
from an open interval of the reals to some space). Does
anyone worry that such equations do not ``really'' describe
the dynamic behavior of a system? Is it not possible to
combine such ``static'' entities with an incremental
solution procedure to build systems that actually control
their (classical mechanical) environment?

I do not believe that the control system can be implemented
using Logic as a Programming language

------------------------------

Date: Tue, 20 Aug 85 11:20:47 PDT
From: Rick McGeer mcgeer%ucbkim@Berkeley
Subject: Prolog In Lisp...

[Hewitt]:
Prolog (like APL before it) will fail as the foundation for
Artificial Intelligence because of competition with Lisp.
There are commercially viable Prolog implementations written
in Lisp but not conversely.

Wrong.  I spent the better part of last year attempting to hack
up a decent Prolog in Lisp: the general idea was to use the
Prolog implementation for the top levels of the program, and a
Flavors-like "Object Engine" for low-level data structure hacking:
there were hooks in the logic engine to manipulate data structures
by calling the object-oriented stuff, and the trail was modified
to permit destructive assignment.

To make a long story short, the project failed because I couldn't
get decent performance out of the logic engine.  So I threw out all
the data structures stuff and the Object Engine, and concentrated
on squeezing Prolog performance any way I could.

The result, after months of effort, traces, special purpose hacks
for fast unification of special cases, and so forth?  100 LIPS on
the concat loop on a 4MB Vax 11/780 running 4.2BSD.  The
implementation was in Franz, Opus 38.91.

Why?  Franz is simply too slow.  I tried the concat loop, in
interpreted Franz:

(defun concat(l1 l2)
   (cond ((null l1) l2)
         (t (cons (car l1)
                  (concat (cdr l1)
                          l2)))))

 and got these results:

                Iterations      Seconds         Lips Equivalent
                ------------------------------------------------
                100             .166            600
                250             .3              833.333...
                350             .417            840
                400             Died            N/A

I then tried running the same lists through Franz' native, tail
recursion optimized append function, and got:

                Iterations      Seconds         Lips Equivalent
                -----------------------------------------------
                100             .03333          3000
                250             .06666          3750
                350             .1              3500
                500             .13333          3750
                1000            .26666          3750

Or, in short, compiled Franz runs at about 2.5 times the speed
of interpreted CProlog, and interpreted Franz runs at about half
the speed of interpreted CProlog.

Anyway, these figures were enough to convince me that Prolog in
Lisp wasn't going to be competitive with native-coded versions
(Cprolog and the compilers) for a long time, and I gave up.  And
before you mention Lambda's Prolog to me, microcode support is
hardly a Prolog coded entirely in Lisp.  Given microcode support
on the PLM, I'm sure that a viable Lisp-in-Prolog could be built.

-- Rick

------------------------------

Date: Tue, 20 Aug 1985  19:04 EDT
From: HEWITT@MIT-XX.ARPA
Subject: Prolog on Lisp vs. Lisp on Prolog

In the exhibitors hall at the IJCAI in Los Angeles, you
can find several commercially viable Prologs on top of
Lisp and Lisp Machines but no commercially viable Common
Lisp on top of Prolog.

Does anyone believe that it will EVER be possible to build
a commercially viable Common Lisp on Prolog?

-- Carl

------------------------------

Date: Tue, 20 Aug 85 16:58:10 PDT
From: Rick McGeer mcgeer%ucbkim@Berkeley
Subject: Prolog on Lisp vs. Lisp on Prolog

I'd be interested to see some figures on the performances
of these Prologs.  Last year, various people told me that
any Prolog that didn't perform to the level of Warren's
compiled DEC-10 Prolog (which is to say, 20-30KLIPS)
wouldn't be commercially viable: on a VAX, (other things
being equal) such a Prolog should obtain about 5 KLips.

Now, it's clear that a Prolog In Lisp can't possibly do
better than the Lisp interpreter.  Since even compiled Franz
can't hit 5 KLips, it is clear that no Prolog-In-Franz will
be commercially viable.  Now, it's possible that dedicated
Lisp machines such as the 3600 or the Lambda can run a variant
of Lisp fast enough that Prolog can run in the commercially
viable range.  However, if you're going to make that argument,
then the question you should be posing is whether dedicated
Prolog machines can run Lisp fast.  And the answer is, sure
it can: a 200+ KLIP processor can run almost anything fast.

-- Rick.

------------------------------

Date: Wed, 21 Aug 1985  17:53 EDT
From: HEWITT@MIT-XX.ARPA
Subject: Prolog on Lisp vs. Lisp on Prolog

Why should a Prolog on Lisp be limited by the speed of
interpreted Lisp? I would imagine that the limitation
on speed would be that of highly optimized compiled Lisp.

-- Carl

P.S.  What is your definition of a KLIP?  For what purposes
do you think that KLIPs is a useful measure of performance?

------------------------------

Date: Wed, 21 Aug 85 15:19:58 PDT
From: Rick McGeer mcgeer%ucbkim@Berkeley
Subject: Prolog on Lisp vs. Lisp on Prolog

[Hewitt]:
Why should a Prolog on Lisp be limited by the speed of
interpreted Lisp? I would imagine that the limitation
on speed would be that of highly optimized compiled Lisp.

Well, I'd like to be wrong.  If I could think of a way to
compile Prologcode into Lisp, then I'd agree.  Unhappily,
I can't.  The real trouble is maintaining the state of the
stack.

[Hewitt]:
P.S.  What is your definition of a KLIP?  For what purposes
do you think that KLIPs is a useful measure of performance?

A KLIP is 1000 Logical Inferences Per Second: I suppose that
would best be defined as succesful unifications, or function
calls.  If you are going to argue that this is at best a
flawed measure of performance (since not all clauses are created
equal), then of course I'd agree.  Performance studies here at
Berkeley have shown that it's easy to write the same program
two different ways, one of which runs faster and the other at a
higher LIPS rate.  Still, a LIP is a handy catchall, much as a
MIP or a FLOP is...

-- Rick

[ Yawn.  This definition is old, more or less resolved for
  flap, peruse Volume 1 of the Archive -ed ]

------------------------------

Date: Wed, 21 Aug 1985  18:42 EDT
From: Hewitt@MIT-MC.ARPA
Subject: Prolog on Lisp vs. Lisp on Prolog

How do you measure the speed in KLIPS in practice.  Do
you just measure the speed of APPEND in Prolog?

--Carl

------------------------------

Date: Wed, 21 Aug 85 18:27:08 PDT
From: Rick McGeer  mcgeer%ucbkim@Berkeley
Subject: Prolog on Lisp vs. Lisp on Prolog

That is the way that I do it.  Basically, I just
calculate the count of the call tree (not counting
failed unifications).  Others may have other ideas.

Of course, you use other benchmarks.  concat (or
append, if you prefer) tends to give you a pretty
optimistic answer.

BTW, a word about Common Prolog.  It is fairly clear
that Prolog will need destructive assignment before
it becomes commercially viable: it turns out that the
major problem with destructive assignment (undoing the
assignments upon backtrack) is not all that hard to
solve.  Once that happens, I imagine that you'll see a
large number of powerful, commercial Prolog systems.

-- Rick

------------------------------

Date: Thu, 22 Aug 85 12:55:46 pdt
From: Allen VanGelder <avg@diablo>
Subject: Re: minor fix for library

To users of my ranpkg in the Prolog library at
SCORE, if any: I replaced the expression

        \(1 << (Wsize - 1))

by the more precise version,

        \(\(0) << (Wsize - 1))

This has no effect using the current DEC-20 Prolog,
but makes the program less implementation-dependent.

Unfortunately my Cprolog converts large integers to
floating point without being asked, so extensive changes
were necessary to get around that "feature" and these
changes are not in the library.

------------------------------

Date: Mon, 5 Aug 85 10:09:19 pdt
From: Peter Ludemann <ludemann%ubc.csnet@csnet-relay>
Subject: Public domain Prolog, Hope

The August 1985 issue of BYTE magazine has a number of
articles on Declarative Langauges (Prolog, Hope, FP).

Public domain versions of Prolog (from Automata Design
Associates) and Hope (from Imperial college) for MS-DOS
machines are apparently available from BYTEnet at (617)
861-9774.  Would some kind soul who lives in that area
please post these to this Digest? Phone lines here are
bad just going across the city; going across the
continent would be a nightmare.

-- Peter Ludemann

------------------------------

End of PROLOG Digest
********************

∂25-Sep-85  1133	CONTRERAS@SU-SCORE.ARPA 	Parking Stickers    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  11:32:30 PDT
Date: Wed 25 Sep 85 11:28:53-PDT
From: Tina Contreras <CONTRERAS@SU-SCORE.ARPA>
Subject: Parking Stickers
To: Faculty@SU-SCORE.ARPA
cc: Staff@SU-SCORE.ARPA
Message-ID: <12146119817.9.CONTRERAS@SU-SCORE.ARPA>


You'll never believe it but the Police Station finally  called and the stickers
are ready,  so you'll beable to pick up your sticker from me at 1:00.

Tina
-------

∂25-Sep-85  1202	@SU-SCORE.ARPA:MDIXON@SU-SUSHI.ARPA 	maps to new student lunch   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  12:02:19 PDT
Received: from SU-SUSHI.ARPA by SU-SCORE.ARPA with TCP; Wed 25 Sep 85 11:59:00-PDT
Date: Wed 25 Sep 85 11:57:56-PDT
From: Mike Dixon <MDIXON@SU-SUSHI.ARPA>
Subject: maps to new student lunch
To: new-phd@SU-SUSHI.ARPA, new-msai@SU-SUSHI.ARPA, new-csms@SU-SUSHI.ARPA,
    faculty@SU-SUSHI.ARPA, student-advisors@SU-AIMVAX.ARPA

i've left a bunch of the maps showing how to find the pratts with tina
(the receptionist, second floor).                           .mike.

p.s.  if you haven't rsvp'd yet, what are you waiting for?
pps.  if you replied to my earlier message about carpooling and haven't
      received a reply, don't worry -- i'm collecting all the responses
      and will get back to you soon.
-------

∂25-Sep-85  1353	CHEADLE@SU-SCORE.ARPA 	Reminder--CSD Reception tomorrow
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  13:53:28 PDT
Date: Wed 25 Sep 85 13:45:52-PDT
From: Victoria Cheadle <CHEADLE@SU-SCORE.ARPA>
Subject: Reminder--CSD Reception tomorrow
To: csd@SU-SCORE.ARPA
Office: Margaret Jacks 258, 497-1519
Message-ID: <12146144756.31.CHEADLE@SU-SCORE.ARPA>


Remember that tomorrow is the annual CSD reception for new
graduate students.  It will be held at 4:00 on the patio
between the Psychology building and CS building.  Please join
us in welcoming our new students.

Hope to see you all there!

Victoria
-------

∂25-Sep-85  1408	REGES@SU-SCORE.ARPA 	Department Lecture Series    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  14:08:44 PDT
Date: Wed 25 Sep 85 14:02:26-PDT
From: Stuart Reges <REGES@SU-SCORE.ARPA>
Subject: Department Lecture Series
To: faculty@SU-SCORE.ARPA
Office: Margaret Jacks 210, 497-9798
Message-ID: <12146147772.29.REGES@SU-SCORE.ARPA>

As a result of the discussion last week, I have taken this course off of the TV
network and have moved it to a room in the history corner.  If you are
interested in giving a presentation, please let me know now so that I can work
out the schedule for the quarter.  The class meets on Thursdays from 2:45 to 4
starting tomorrow.  I'm planning on speaking tomorrow unless I get a last minute
volunteer.  I plan to give a general overview of the department and its research
and curriculum.  Please let me know if you want to speak on one of the other
Thursdays.
-------

∂25-Sep-85  1412	@SU-CSLI.ARPA:chertok%ucbcogsci@Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 1  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  14:10:57 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Wed 25 Sep 85 14:07:32-PDT
Received: by UCB-VAX.ARPA (5.26/5.9)
	id AA04219; Wed, 25 Sep 85 14:08:06 PDT
Received: by ucbcogsci.ARPA (5.5/5.7)
	id AA28422; Wed, 25 Sep 85 14:09:11 PDT
Date: Wed, 25 Sep 85 14:09:11 PDT
From: chertok%ucbcogsci@Berkeley.EDU (Paula Chertok)
Message-Id: <8509252109.AA28422@ucbcogsci.ARPA>
To: cogsci-friends%ucbcogsci@Berkeley.EDU
Subject: UCB Cognitive Science Seminar--Oct. 1

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                    Cognitive Science Seminar -- IDS 237A

        TIME:                Tuesday, October 1, 11:00 - 12:30
        PLACE:               240 Bechtel Engineering Center
        (followed by)
        DISCUSSION:          12:30 - 1:30 in 200 Building T-4

        SPEAKER:          David  Rumelhart,  Institute  for  Cognitive
                          Science, UCSD

        TITLE:            ``Parallel Distributed Processing:  Explora-
                          tions in the Microstructure of Cognition''

        Parallel Distributed Processing (PDP) is the name which I  and
        my  colleagues  at  San  Diego  have  given  to  the  class of
        neurally-inspired models of cognition we have  been  studying.
        We  have  applied  this  class  of "connectionist" models to a
        variety of  domains  including  perception,  memory,  language
        acquisition  and motor control.  I will briefly present a gen-
        eral framework for the class of PDP  models,  show  how  these
        models  can  be applied in the case of acquisiton of verb mor-
        phology, and show how such  macrostructural  concepts  as  the
        schema  can be seen as emerging from the microstructure of PDP
        models.  Implications of the PDP perspective  for  our  under-
        standing of cognitive processes will be discussed.
        ----------------------------------------------------------------
        UPCOMING TALKS
        Oct 8:      Terry Winograd, Computer Science, Stanford
        Oct 15:     Ron Kaplan, Xerox PARC
        Oct 22:     Lotfi Zadeh, Computer Science, UCB
        Oct 29:     Mardi Horowitz, Psychiatry, UCSF
        Nov 5:      Edward Zalta, CSLI, Stanford
        Nov 12:     Robert Wilensky, Computer Science, UCB
        Nov 19:     Richard Alterman, Computer Science, UCB
        Nov 26:     Eve Clark, Linguistics, Stanford
        Dec 3:      Bernard Baars, Langley Porter, UCSF
                                  * * * * *
        ELSEWHERE ON CAMPUS
        Steven Christman will be speaking on ``Visual Persistence'' on
        Friday,  October 4, 1985, at 4:00 p.m. in the Beach Room, 3105
        Tolman Hall, UCB.

∂25-Sep-85  1738	EMMA@SU-CSLI.ARPA 	Newsletter September 26, No. 47
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  17:38:40 PDT
Date: Wed 25 Sep 85 16:54:08-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter September 26, No. 47
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
September 26, 1985              Stanford                       Vol. 2, No. 47
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *THIS* THURSDAY, September 26, 1985

   12 noon		TINLunch
     Ventura Hall       ``The Concept of Supervenience''
     Conference Room    Discussion led by Carol Cleland

   2:15 p.m.		CSLI Talk
     Ventura Hall	No talk this week

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←
          CSLI ACTIVITIES FOR *NEXT* THURSDAY, October 3, 1985

   12 noon		TINLunch
     Ventura Hall       ``Idealized Cognitive Models'' and ``Metonymic Models''
     Conference Room    Sections 4, 5 of ``Women, Fire, and Dangerous Things''
			by George Lakoff
			Discussion led by Douglas Edwards
			(Abstract on page 2)

   2:15 p.m.		CSLI Seminar
     Ventura Hall	``Notes from the STASS Underground''
     Seminar Room	David Israel, CSLI and SRI
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←
                     THIS YEAR'S THURSDAY ACTIVITIES

      CSLI's year will be starting next Thursday, October 3, and several
   changes have been made. 

      TINLunches will be organized by Chris Menzel and Mats Rooth, two
      CSLI postdoctoral fellows.  They will continue to meet at noon in
      the Ventura Conference room.

      Thursday Seminars will have a different format this year and will
      consist of either individual presentations from the postdocs or a
      presentation by one of the new projects of its goals and progress.
      
      Thursday Colloquia will be rarer and of more general interest.
      Each project will be responsible for one colloquium, and we hope to
      have three colloquia a quarter.  Time and location of the colloquia
      may vary.

   Next week's newsletter will contain a list of the new projects and a
   tentative calendar for the Fall quarter.
!
Page 2                     CSLI Newsletter                 September 26, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
         ``Idealized Cognitive Models'' and ``Metonymic Models''
         Sections 4, 5 of ``Women, Fire, and Dangerous Things''

      According to Lakoff, many words are understood by reference to
   ``Idealized Cognitive Models'' (ICMs) which describe the ideal
   circumstances in which the phenomena these words refer to are
   conceived to exist.  Some uses of a word can only be understood by
   treating the word's ICM as true even when it is known to be false in
   general.  Other uses modify the word's meaning by more or less
   explicitly calling the ICM in question or by focusing on cases to
   which the ICM clearly fails to apply.
      Thus linguistic puzzles can arise.  For instance ``bachelor'' is
   often defined as ``unmarried man,'' and ``to lie'' as ``to make a
   false statement,'' even though it is well known that these terms are
   not coextensive with their definitions.  When a word is defined, its
   ICM is taken for granted, but when a purported example is judged,
   failure of applicability of the ICM can make the purported example
   illegitimate or at least atypical.  The ICMs for ``bachelor'' and
   ``lie'' fail partly or totally for priests, children, polygamists,
   misleading true statements, polite nothings, and accidental errors.
      Syncategorematic noun modifiers often affect the ICM.  Thus we get
   ``social lie,'' ``white lie,'' ``eligible bachelor'' (this one
   reinforces the ICM), ``foster mother,'' ``surrogate mother,'' and so
   on.
      ICMs are interesting in that they seem to be used in reasoning
   generally, not just in lexical semantics.  They are akin to, but not
   identical with, various constructs developed for artificial
   intelligence, such as frames, scripts, contexts, data pools, etc.
						--Douglas Edwards
                              ←←←←←←←←←←←←
                  ABSTRACT OF NEXT WEEK'S CSLI SEMINAR
                  ``Notes from the STASS Underground''

      I will try to explain the meaning and import of one of the hottest
   acronyms at CSLI -- ``STASS.''  In particular, I will try to explain
   why there should be a Situation Theory as well as a Situation
   Semantics.					--David Israel
                              ←←←←←←←←←←←←
                                CSLI TALK
                           ``Verbs and Time''
                    Dorit Abusch, Tel-Aviv University
            Tuesday, October 1, 1 pm, Ventura Conference Room

      In ``Word Meaning and Montague Grammar,'' David Dowty analyzed
   aspectual clauses in terms of an ``aspectual calculus'' consisting of
   stative predicates and operators such as BECOME and CAUSE.  For
   instance, achievements, including many morphological inchoatives, are
   analyzed as having the form lambda x[Become(P(x))].  Accomplishments,
   including many morphological causatives, are analyzed in terms of
   CAUSE.  Dowty and Lauri Carlson noted that some inchoatives, such as
   (the verb) ``cool,'' meet the test for process verbs, I discuss these
   inchoatives, and similar causatives.  The relation between the
   operators and the verb classification is complex.  I argue that the
   classification breaks down for certain causatives, such as the
   transitive versions of ``gallop'' and ``darken.''
!
Page 3                     CSLI Newsletter                  September 26, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                               AFA SEMINAR

      This quarter there will be a small informal seminar going through
   Peter Aczel's work on the anti-foundation axiom (AFA) in set theory,
   together with some of the applications found by people here at CSLI.
   We will start at the beginning, but assume familiarity with the
   cumulative hierarchy and ZFC.  The seminar will be Thursdays at 4:15
   when there is no CSLI colloquium, in the Ventura Conference room.  Jon
   Barwise will give a brief introduction on September 26, and then we
   will organize the rest of the quarter.  If you would like to be added
   to the AFA mailing list, contact Westerstahl@csli.
                              ←←←←←←←←←←←←
                   NEW PROJECT MEETING ON ENVIRONMENTS
              Mondays 1-2 in the trailer classroom, Ventura

      Beginning Monday, September 30 there will be a weekly meeting on
   environments for working with symbolic structures (this includes
   programming environments, specification environments, document
   preparation environments, ``linguistic workstations,'' and
   grammar-development environments).  As a part of doing our research,
   many of us at CSLI have developed such environments, sometimes as a
   matter of careful design, and sometimes by the seat of the pants.  In
   this meeting we will present to each other what we have done, and also
   look at work done elsewhere (both through guest speakers and reading
   discussions).
      The goal is to look at the design issues that come up in building
   environments and to see how they have been approached in a variety of
   cases.  We are not concerned with the particular details (``pop-up
   menus are/aren't better than pull-down menus'') but with more
   fundamental problems.  For example:

      What is the nature of the underlying structure the environment
      supports: chunks of text? a data-base of relations? a tree or graph
      structure?  How is this reflected in the basic mode of operation
      for the user?

      How does the user understand the relation between objects (and
      operations on them) that appear on the visible representation
      (screen and/or hardcopy) and the corresponding objects (and
      operations) on some kind of underlying structure?  How is this
      maintained in a situation of multiple presentations (different
      views and/or multiple windows)?  How is it maintained in the face
      of breakdown (system failure or catastrophic user error in the
      middle of an edit, transfer, etc.)?

      Does the environment deal with a distributed network of storage and
      processing devices?  If so, does it try to present some kind of
      seamless ``information space'' or does it provide a model of
      objects and operations that deals with moving things (files,
      functions, etc.)  from one ``place'' to another, where different
      places have relevant different properties (speed of access,
      security, shareability, etc.)?
!
Page 4                     CSLI Newsletter                  September 26, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

      How is consistency maintained between separate objects that are
      conceptually linked (source code and object code, formatter source
      and printer-ready files, grammars and parse-structures generated
      from them, etc.)?  To what extent is this simply left to user
      convention, supported by bookkeeping tools, or automated?

      What is the model for change of objects over time?  This includes
      versions, releases, time-stamps, reference dates, change logs,
      etc., How is information about temporal and derivational
      relationships supported within the system?

      What is the structure for coordination of work?  How is access to
      the structures regulated to prevent ``stepping on each other's
      toes,'' to facilitate joint development, to keep track of who needs
      to do what when?

      Lurking under these are the BIG issues of ontology, epistemology,
   representation, and so forth.  Hopefully our discussions on a more
   down-to-earth level will be guided by a consideration of the larger
   picture and will contribute to our understanding of it.
      The meeting is open to anyone who wishes to attend.  Topics will be
   announced in advance in the newsletter.  The first meeting will be
   devoted to a general discussion of what should be addressed and to
   identifying the relevant systems (and corresponding people) within
   CSLI, and within the larger (Stanford, Xerox, SRI) communities in
   which it exists.				--Terry Winograd
                              ←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
   ``Cree Verb Inflection: Linking Features to Grammatical Functions''
                 Summary of the meeting on September 12

      Cree (Algonquian) is a non-configurational language in which
   grammatical functions are encoded by means of a complicated system of
   verbal inflection.  The verb has ten inflectional affix positions; no
   single position is dedicated to a particular grammatical function.
   The shape of the person and number affixes is the same for both
   subject and object.  The task of linking person and number feature
   values with the appropriate grammatical function falls to a set of
   morphemes traditionally called ``theme signs.''
      The talk focussed on the role of the theme signs.  Some recent
   theoretical accounts have analyzed the theme signs as marking a voice
   opposition; on these accounts, the theme signs would be derivational,
   rather than inflectional. A subset of the theme signs would mark the
   application of a rule like passive, or a rule of ergative relinking,
   in which the theme argument is linked to subject, and the agent
   argument is linked to object.  However, syntactic tests (copying to
   object, quantifier float, complement control) show that the passive
   and the ergative relinking hypotheses must both be rejected.
      In Dahlstrom's analysis, the theme signs are inflectional, acting
   as a filter on possible linkings of person and number features to
   grammatical functions.  The other inflectional affixes carry specific
   feature values for person and number, but are unspecified for
   grammatical function.  Ungrammatical linkings of feature values to
   grammatical functions are ruled out by general conditions of
   completeness, coherence, and consistency.		--Amy Dahlstrom
!
Page 5                     CSLI Newsletter                 September 26, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                             NEW CSLI REPORTS

      Report No. CSLI-85-31, ``A Formal Theory of Knowledge and Action''
   by Robert C. Moore, and Report No. CSLI-85-32, ``Finite State
   Morphology: A Review of Koskenniemi'' by Gerald Gazdar, have just been
   published.  These reports may be obtained by writing to David Brown,
   CSLI, Ventura Hall, Stanford, CA 94305 or Brown@SU-CSLI.
-------

∂25-Sep-85  2035	JF@SU-SUSHI.ARPA 	BATS at Berkeley 10/11/85  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 25 Sep 85  20:35:49 PDT
Date: Wed 25 Sep 85 20:32:26-PDT
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: BATS at Berkeley 10/11/85
To: aflb.su@SU-SUSHI.ARPA

There will be a BATS (for newcomers, that's Bay Area Theory Seminar) at
Berkeley on Friday, October 11.  Please let me (jf@sushi) know:

1) Would you like to speak?

2) Do you need a ride?

3) Can you offer a ride?

4) Would you like to replace me as Stanford BATS coordinator?

Thanks,
Joan
-------

∂26-Sep-85  0059	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #38
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 26 Sep 85  00:59:17 PDT
Date: Tuesday, September 24, 1985 2:24PM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #38
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest           Wednesday, 25 Sep 1985     Volume 3 : Issue 38

Today's Topics:
                   Query - Books-in-print database,
                    Challenge Problem - Ken Laws,
                 LP Philosophy - Hewitt's Challenge,
                Programming - Building Large Systems,
         Implementations - Cut Test & Destructive Assignment,
                         LP Library - Update
----------------------------------------------------------------------

Date: Mon, 9 Sep 85 10:42 PDT
From: RGARRETT%LAJ.SAINET.MFENET@LLL-MFE.ARPA
Subject: Books-in-print database

I assume you are familiar with "Books in Print" , the multi-volume
listing of most of the books currently in print in the USA.  I
would  like something similar but on magnetic media;i.e., mag tape
and hopefully cheap.  The purpose  of this is to serve as the
foundation for an expert system retrieval system to be written in
(what else) PROLOG.  Since this is both an experimental system (how
many other kinds of expert systems are there?) and done as a
graduate student, I am only interested in public domain or very low
cost sources.  I am already familiar with commercial systems such
as DIALOG and with IRList.  I hope this answers your questions, but
if something is still unclear, just drop me a note.

-- Randy Garrett

------------------------------

Date: Thu, 29 Aug 85 10:39:26 pdt
From: Allen VanGelder <avg@diablo>
Subject: Ken Laws' challenge problem

The problem statement assumes some familiarity with image
processing jargon.  I am guessing that he is talking about
a 2-dimensional array and that two elements are "connected"
if they are within 1 of each other on both indexes, i.e.,
diagonal adjacency as well as horizontal and vertical
adjacency is connected.

The advantage of a logic-based language, if any, is not to
remove ALL procedural responsibility from the programmer, but
to provide a way to write procedures that have a close enough
connection to the specifications that they can be informally
verified.  For example, I can write

        neighbor((i,j), (k,l)) :-
                not ((i,j) = (k,l)),
                adjacent(i, k),
                adjacent(j, l).

        adjacent(i, j) :-
                (i = j; i is j+1; j is i+1).

Note that semi-colon is read "or" and ":-" is read "if" and
comma is read "and".  This "procedure" can simply be read in
English as a specification of "neighbor", yet it is also the
Cprolog source code.

Arrays are certainly a problem in logic programming, and Ken's
problem is a good challenge.  But he should not expect to see a
naive problem specification that runs as a Prolog program.  He
IS entitled to ask for a logic program that is both efficient
and informally verifiable. I don't have that for him.

------------------------------

Date: 05 Sep 85 18:29:55 PDT (Thu)
From: Sanjai Narain <Narain@rand-unix.ARPA>
Subject: Response to Hewitt

>Prolog (like APL before it) will  fail  as  the  foundation  for
>Artificial   Intelligence  because  of  competition  with  Lisp.
>There are commercially  viable  Prolog  implementations  written
>in Lisp but not conversely.

For the same reason, Lisp should have failed as a foundation for
computing because  of  competition  with  assembly language.
There  are commercially viable implementations of Lisp in assembly
language  but not conversely.

>LOGIC as a PROGRAMMING Language will fail as the foundation
>for AI because:

>1.  Logical inference cannot be used to infer the decisions
>    that need to be taken in open systems because the decisions
>    are not determined by system inputs.

>>2.  Logic does not cope well with the contradictory knowledge
>>    bases inherent in open systems.  It leaves out
>>    counterarguments and debate.

>>3.  Taking action does not fit within the logic paradigm.

1.  Hewitt clearly states in his  recent  BYTE  article  that
traditionalnotions  of  computation  as  defined,  for example,
by Turing machines or recursive functions cannot model the behavior
of open systems.  Hence even Lisp  is  inadequate for such modeling
(by his reasoning).

2.  The notion of contradiction (i.e. inconsistency) is well
understood in logic.

3.  The statement is too vague for debate.  What do the words
"action" and"fit"  mean?   Certainly,  if  action  can  be  modeled
by  an  effective procedure, it can be modeled by logic, cf. 1.

-- Sanjai Narain

------------------------------

Date: Thu, 5-Sep-85 13:40:43 PDT
From: (Tom Khabaza) mcvax!ukc!warwick!cvaxa!tomk@Seismo
Subject: On Hewitt's "Prolog and logic programming will fail"

I have read with interest the discussion following Carl Hewitt's
"Prolog will fail as the foundation for AI and so will Logic
Programming". I particularly enjoyed Vijay Saraswat's reply, most
of which I agree with. However, I would like to add a few comments:

In some ways I was surprised by the original message; I should
have thoughtthat if AI has taught us anything, it is that to solve
a given problem, we need a good representation language.  Why anyone
might think that logic is the BEST representation language for every
problem is beyond me.  (No Kowalskiist flames please, I know the
arguments, and I don't regard the case as proven.)

On the other hand, we don't yet know what the limits of logic
programming are; researchers in the field are constantly coming up
with new techniques.  There is convincing evidence that logic
programming is better than conventional programming for some kinds
of task, at least with regard to ease and clarity (though probably
not yet efficiency).

But I think the basis of the original comment goes deeper than the
virtues and vices of logic programming.  As I understand it (and I
wasn't around at the time) some earlier AI programming languages,
such as perhaps micro-Planner and its successors, WERE expected to
become a "foundation" for AI.  Perhaps this was because people still
had hopes for the notion of some "ultimate" representation language,
or family of languages.

AI is older and perhaps more cynical now; I don't think we expect
some single foundation to the field in the form of a representation
language.  Logic programming may be very useful for some parts of
AI; for example some kinds of rule based systems, but I don't expect
it to be the best tool for all kinds of AI programming.  In fact my
personal opinion is that logic programming will find its forte in
more conventional Computer Science, where formal specification is a
more practical proposition than in the relatively exploratory
activity of AI programming.

But I will say this in its favour: logic programming is IMPORTANT.

Logic programming is as different from conventional programming as
programming is from not programming at all.  I have met people who
have given up on Prolog because it was difficult for them and they
(rightfully) considered themselves competent programmers - and so
thought it must be Prolog's fault!  (I don't mean to imply that
anyone who has posted in this discussion is such a person.) But
logic programming is different in fundamental ways; it's worth
presevering to get to the bottom of it, and as logic programming
languages improve, it will become even more so.

So for all you computer people out there, USE Prolog, and study
how other people have used it.  It really is worth it.

------------------------------

Date: Tue, 20 Aug 85 21:07:41 EDT
From: Stephen Wolff <steve@BRL.ARPA>
Subject: Another precinct heard from

"Cut test" results for University of York Portable
Prolog, V2.1:

Test

 1  2  3  4  5  6
 Y  Y  Y  Y  Y  N

------------------------------

Date: Wed, 28 Aug 85 09:06:26 -0200
From: David McKelvie <mcvax!unido!ecrcvax!david@seismo.CSS.GOV>
Subject: User Interfaces - Building Large Systems

We, the ECRC Man-machine interaction group, have had similar
problems with the slowness of Prolog interpreters for handling
interactive graphics, and have come up with a similar solution.
We have split off the user interface part into a seperate
process (we are using UNIX) and communicate to Prolog via a
pipe, rather than a direct connection to the Prolog database
as you seem to have.  However we are having problems in defining
what sort of interface there should be between the UI and Prolog.

Therefore, we would be rather interested in the technical details
of your interface, in particular the Prolog predicates you have
defined, and what you mean by 'object orientedness' ( a horrible
term for a good idea). Does this allow a more declarative than
imperative way of defining pictures from Prolog?

Thank you

-- David Mckelvie

(ECRC stands for European Computer-Industry Research Centre, a
 centre funded by European computer companies, mainly doing work
 on Logic programming and Knowledge bases. A sort of miniature
 version of ICOT or MCC )

------------------------------

Date: Wed, 28 Aug 85 22:32:37 PDT
From: (Rick McGeer) mcgeer%ucbkim@Berkeley
Subject: Destructive assignment

Dedicated hardware isn't magic, you know.  You still fill up
space with garbage, and there is no particular reason to believe
that you're going to be able to predict or control where you leave
it.  Also keep in mind that if you have a paged virtual memory
system, you experience an enormous degree of internal
fragmentation if your wastage is large and is spread evenly over
more or less half your pages.  That's a recipe for thrashing....

The other, major problem with no destructive assignment is time.
If you want to change a small field buried deep inside a large
record, you have to copy the entire record.  On interpreted systems
this is not so bad since structure is shared.  On compiled systems,
such as Warren & Tick and Berkeley PLM, it's an absolute disaster
because such machines copy rather than share structures.  In that
sense compiled code on special-purpose h'ware performs more
poorly than interpreted code on a gp machine.

Some computations are history-sensitive, and that's all there is
to it.  Either your programming language recognizes these realities
or it doesn't.  If it does, it's useable; if it doesn't, it becomes
a delight for mathematicians and unused by everyone else.  Lisp
implementers swallowed hard, gave up purity, and put in setq and
rplacd.  Lisp is used and useable.  FP and Lucid implementers,
among others, stuck by no-assignment semantics and washed up on
the ash-heap of programming languages.  People who implement
Prolog are faced with the same choice, and will face the same
results.

-- Rick

------------------------------

Date: Fri, 30 Aug 85 13:24:56 PDT
From: (Rick McGeer) mcgeer%ucbkim@Berkeley
Subject: Destructive assignment

Here's a good example of when you really need destructive
assignment in Prolog.  I'm currently writing an IC layout
expert in Prolog.  At one point in the program, nets are
assigned to rows.  When this is done, the row's freelist
and wirelist is updated, and so, of course, a new row is
generated.

Now, the difficulty is that the net structure includes the
row that the net runs on.  Unfortunately, this row may be
regenerated later with a different freelist.  When this happens,
the row pointed to by the net structure disappears from the
list of rows in the grid, so that later processing reports
that the row doesn't exist.

The solution is to keep an (atomic) ID field in each row and
index through that.  This both complicates the code and
increases the running time, and obscures what's really going
on.

I might point out that other researchers in this area started
out with Prolog and then went to other languages (eg, C), largely
for these sorts of reasons.  Dwight Hill of BTL, who gave a paper
at ICCD '84 on "The Silicon Converter: A Case Study in Uses For
Prolog" has now gone over to C for IC placement and routing.

-- Rick

------------------------------

Date: Fri, 13 Sep 85 01:09:57 edt
From: Tom Blenko  <blenko@rochester.arpa>
Subject: Destructive assignment

I agree with your points about the time/memory costs
and difficulties associated with either copying or sharing
data structures.

First point: if you add destructive assignment to PROLOG,
it isn't (even nearly) PROLOG anymore.

Second point: the issues you raise are more clearly critical
for functional languages.  I don't agree that FP and LUCID
are washed up on the ash-heap of programming languages: FP
was basically a model for a functional language, not a proposal
for an implementable, production-oriented language.  There are
people working quite seriously  on the implementation of functional
languages (I am less familiar with serious work on implementation
of declarative languages).  I agree that there are as yet no
visible successes, but it is certainly too soon for the jury to
come back in.

We can look at how the LISP machine people addressed the problem
in their design: large VM, large physical memory, cacheing of
everything they could think of, and more recently (not yet released)
garbage collection out of the cache. I don't know the details of
the scheme, but I think it may have been published, and they are
claiming improvement by a factor of two in performance as a result.

I know also of a major computer manufacturer which is implementing
a chip based on SASL. They claim on-the-fly garbage collection,
although emergency (out-of-space) collection is provided.

One performance benefit of functional (and perhaps logic
programming) languages may come in their parallel execution --
specifically that this may be easier to accomplish than for
languages  with destructive assignment.  I have looked at this
for Concurrent Prolog, and the
write-once property of variables plays a very important role in
reducing the complexity of inter-process communication.

In sum, I agree that there are problems, and I agree that the
solutions are not in hand, but I think it is hasty to conclude
that solutions are not possible, especially since we are really
in the infancy of architectures designed with such languages in
mind.

-- Tom

------------------------------

Date: Sun, 11-Aug-85 19:32:53 PDT
From: ihnp4!houxm!mhuxt!mhuxr!ulysses!burl!clyde!watmath!utzoo
Subject: Supplement to C-Prolog manual, Grammars, Exanmples, etc.

The User's Manual for DEC10 Prolog (circa 1978) had several useful 
sections on definite-clause grammars, grammar rules for "Edinburgh" 
syntax, example programs, and (early) Prolog references.  None of this
information was in the C-Prolog manual distributed with C-Prolog 1.5,
so before our beloved (:-)) TOPS-10 machine vanished into obliv
-ion, I converted these sections into "troff -ms" format to serve as 
an appendix to the C-Prolog manual which we distribute to students in 
a LP course here. There is no copyright on this material, as far as I
can tell, so I hope F. Pereira & Co. won't mind.  I'm postingthe 
excerpts to the net, at the suggestion of a colleague.  I wouldn't 
mind hearing from anyone who found it useful (or who objects to long 
articles containing material of ancient vintage); it's hard for me to 
gauge whether or not purely pedagogical material such as this should 
be circulated on the net.

[ This is available from the SCORE: <Prolog> libray as:

       CProlog←Supplementary←Manual.Txt and can be netted over,
  
  if you can't access SCORE: send a request to Prolog-Request -ed ]

------------------------------

End of PROLOG Digest
********************












∂26-Sep-85  0751	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #39
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 26 Sep 85  07:51:04 PDT
Date: Wednesday, September 25, 1985 4:20AM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #39
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest           Thursday, 26 Sep 1985      Volume 3 : Issue 39

Today's Topics:
                 LP Philosophy - Hewitt's Challenge,
                 Programming - DCGs + Left Recursion,
               Implementation - Destructive Assignment,
----------------------------------------------------------------------

Date: Fri, 13 Sep 85 10:41:27 bst
From: William Clocksin <wfc%computer-lab.cambridge.ac.uk@ucl-cs>
Subject: Lisp/Prolog

I received issues 36 and 37 late owing to netproblem somewhere
between Stanford and Cambridge.  I am puzzled by this Lisp/Prolog
debate started by Carl Hewitt.  I use both Prolog and Lisp, and
have never felt the need to use one exclusively.  I suppose they
are like screwdrivers and chisels;  both are roughly the same,
but for slightly different purposes; to a person unfamiliar with
one of them, the other one might seem redundant.  I am also
puzzled about the question of a "foundation" for AI.  How can a
language be a "foundation" for anything?  Was Latin a "foundation"
of Western civilisation?  Seen any fundamental native speakers
lately? Besides, does AI deserve to have foundations attributed
to it anyway?

Another problem is this question about logic.  Prolog is a
programming language.  It was inspired by logic, but it is not
programming in logic.  Proponents of using logic do have a problem
matching impedance with the real world.  But Prolog is to logic
as Lisp is to lambda calculus.  Those who advocate programming in
lambda calculus have the same problem as those who advocate
programming in pure logic.  If Prolog can be said to have any
connection with logic, it is as the FORTRAN of logic programming.
Prolog is useful because you can grow data structures that have
actual variables in them, and because it is easy to define
nondeterministic methods.  I know how Prolog searches for a solution
just as I know how flow of control happens in Lisp, say.  I am not
disappointed with Prolog's strict strategy just as I am not
disappointed with Lisp's inability to run programs backwards, say.
I take it as it comes, and it is useful for some things.  Talking
hypothetically about the "ideal" language is another topic entirely,
and it only muddies the water to bring Prolog and Lisp into it.

------------------------------

Date: Fri, 13 Sep 85 10:07:09 bst
From: William Clocksin <wfc%computer-lab.cambridge.ac.uk@ucl-cs>
Subject: Prolog on/in Lisp

Compiling Prolog into Lisp is one way to provide a fast
Prolog/Lisp.  A student did this as a project last year;
no doubt it has been done elsewhere.  The Warren New Engine was
used as a model.  The compiler, took Prolog procedures and
generated Lisp code in a mixture of procedure calls and macros.
The clever part was getting backtracking right.  This was then
compiled into IBM machine code, and run on a 3081.  It was not
fast, owing probably to the large size of code body.  I forget
how slow, say 10KLIPS or so.  The final speed was not important
for the goals of the project.  More important was finding out
what special provision could be made in the future (in the form
of Prolog-specific support) to provide a more efficient system.
Also, it was the student's very first Lisp program, so hhe can be
forgiven for doing a few extremely inefficient things.

------------------------------

Date: Thu, 19 Sep 85 16:15:07 PDT
From: Adolfo Di-Mare <dimare@LOCUS.UCLA>
Subject: DCGs + Left Recursion = (:-[) !

I'm writing a little parser for a Pascal like language.
I was going to copy Wirth's grammar into a DCG Prolog
program, but I noticed that left recursion can't be used.
If you have a rule like:

a --> a, [+], b.      a --> b.      b --> [b].

it translates into:

a(A, B) :- a(A, C), 'C'(C, +, D), b(D, B).      'C'([H|T],H,T).

Any goal of the form a(Sentence,[]) would make the above
rule recurse until end←of←stack.
(Yes, a([b,+,b],[]) doesn't work either: try it!).

My problem is that the grammars I know for parsing expressions
are left recursive. The example in most Prolog manuals uses left
recursion. I ran it, and I got the following result (I include
the code at the end of the message):

| ?- expr(Z,"2-3-4",[]).

Z = 3 ;

no
| ?- Z is 2-3-4.

Z = -5 ;

no
| ?- halt.

Obviously, the 'standard' example is wrong. My question is: Does
any body know a way to do expressions using DCGs that doesn't make
the grammar look like a plate of spaghetti? I know I could translate
my grammar into Griebach Normal Form (or $%#!"%#$" Normal Form, for
that mater), but I could also as well write the whole parser in
8080 assembly language. Here's the grammar:

Script started on Thu Sep 19 15:16:07 1985
h[2:101] c dcg0
% File   : /usr/lib/prolog/teach/dcg0
% Author : lost in the mists of time
% Updated: 23 Nov 1983
% Purpose: the standard example of using Prolog grammar rules

% This is a simple grammar which parses an arithmetic expression (of
% digits and operators) and computes its value.

expr(Z) --> term(X), "+", expr(Y), {Z is X+Y}.
expr(Z) --> term(X), "-", expr(Y), {Z is X-Y}.
expr(Z) --> term(Z).

term(Z) --> number(X), "*", term(Y), {Z is X*Y}.
term(Z) --> number(X), "/", term(Y), {Z is X/Y}.
term(Z) --> number(Z).

number(C) --> "+", number(C).
number(C) --> "-", number(X), {C is -X}.
number(X) --> [C], {48=<C, C=<57, X is C-48}.

% In the last rule, C is the ASCII code of a digit.
% The question
%
%       | ?- expr(Z,"-2+3*5+1",[]).
%
% will compute Z=14. The two extra arguments here are because
% grammar rules are in fact merely 'syntactic sugar' for ordinary
% Prolog clauses. Each grammar rule takes an input string, analyses
% some initial portion and produces an output portion (the rest,
% possibly enlarged a bit) for further analysis.

h[2:102] uprolog
CProlog version 1.4d.edai
| ?- [dcg0].
dcg0 consulted 1208 bytes 0.78333 sec.

yes
| ?- expr(Z,"2-3-4",[]).

Z = 3 ;

no
| ?- Z is 2-3-4.

Z = -5 ;

no
| ?- halt.

[ Prolog execution halted ]
h[2:103]
script done on Thu Sep 19 15:18:16 1985

------------------------------

Date: Mon, 16 Sep 85 10:26:55 PDT
From: (Rick McGeer) Mcgeer%ucbkim@Berkeley
Subject: Destructive assignment

I would be extremely surprised if you could get away
without destructive assignment.  In addition to the other points
I've raised, consider this: many large programs consist of networked
data structures: that is, some data structure (say A) may be a
substructure of structures (B, C, and D).  If the only way to modify
A is to create a new structure, A', then the code that modifies A to
produce A' must also create B', C', and D', and so forth recursively.
Now, it's almost impossible to write modular code in such a
situation:  the internals of each data structure in such a network
must be apparent to the code which modifies the leaf structures in
the network,  or else the network becomes inconsistent.

This sort of problem creeps up all the time in NL understanding
programs, VLSI layout programs, and algorithm-based algebra programs.
To my knowledge, only toy systems have been written in these areas in
Prolog.  Both Dwight Hill at Bell Labs and I have independently tried
to write a real layout program in Prolog, but the mechanics of moving
the data structures around and ensuring that the ds network remains
consistent have forced him to abandon his attempt (and write in C):
I have been forced into a reappraisal.

I agree, if it were merely a matter of efficiency then we
could hope for hardware to work its magic.  But it's a matter of
sofware engineering, aswell.  A language ultimately stands or falls
on its software engineering aspects and not on its semantics (which
are only important to implementers and, to a minor extent, in
ergonomics).  I am afraid that Prolog currently does not support
good software engineering techniques, and in fact makes some
software engineering techniques (modularity, data abstraction)
impossible. Unless and until it supports those techniques at least
to the extent that Lisp does, Prolog is a toy language.

-- Rick.

------------------------------

Date: Mon, 23-Sep-85 07:40:31 PDT
From: Vantreeck%logic.DEC@DECWRL.ARPA
Subject: DCGs + Left Recursion = (:-[) !

     You are correct in your assertion that most PROLOGs'
DCGs do not handle left recursion. That's because they're
simple top down parsers. There is a PROLOG DCG parser that
uses bottom up parsing (and can handle left recursions).
The parser is called BUP. Need I say it stands for Bottom
Up Parser?

     I don't have the references here at work - the papers
are a couple of years old. I'm sure your UCLA libarary can
do a computer search on BUP. Off the top of my head -- A
brief description can be found in a book by Dahl, Natural
Language Understanding with Logic - it's a collection of
papers from a 1984 symposium in France.

     The articles I've read on BUP tend to be of the "Gee,
isn't this wonderful!" variety, rather than delving into
enough implementation detail that someone could use it as a
guide to writing a BUP. If anyone finds a paper that gives
an in depth description of the implementation details,
please post it to the net!

-- George

------------------------------

End of PROLOG Digest
********************

∂26-Sep-85  1048	@SU-SUSHI.ARPA:MAYR@SU-SCORE.ARPA 	talk by Marc Snir   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 26 Sep 85  10:44:56 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 26 Sep 85 10:38:14-PDT
Date: Thu 26 Sep 85 10:18:47-PDT
From: Ernst W. Mayr <MAYR@SU-SCORE.ARPA>
Subject: talk by Marc Snir
To: su-bboards@SU-SCORE.ARPA, aflb.su@SU-SCORE.ARPA
Message-ID: <12146369201.9.MAYR@SU-SCORE.ARPA>

                             Seminar Announcement

         Software Issues in the Design of Shared Memory MIMD Software

Speaker: Marc Snir
Location: MJH352
Time: Tuesday, Oct. 1, at 1:00pm

Abstract: The NYU Ultracomputer and the IBM RP3 machines are MIMD shared 
memory parallel machines. Processors are connected to global memory by a
global packet switched inter-connection network; the network is actively
involved in processing.

These machines pay the cost of a complex interconnnction network in order
to avoid the burden of explicit resource allocation by the user. They are
intended to provide a close approximation to an ideal model of computation,
where accesses to shared memory are atomic, and can be concurrently done by
all processors.

In reality memory accesses are not atomic, and resource allocation overheads
cannot be ignored. The gap between the ideal model and the machine reality
can be narrowed by a mixture of compile time and run time optimization
techniques, and by new hardware structures.

We shall survey some specific problems and some of the techniques that can
be used to soften their impact. How to schedule memory accesses so that
resulting code is serializable.

How to provide atomic access to structures.
How to cache or prefetch shared data.
How to avoid excessive process fragmentation.
How to synchronize.

-------

∂26-Sep-85  1539	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.16  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 26 Sep 85  15:38:26 PDT
Date: Thu 26 Sep 85 13:33:02-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.16
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-LIST: RISKS-FORUM Digest  Thursday, 26 Sep 1985  Volume 1 : Issue 16

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Intellectual honesty and the SDI (Bill Anderson)
  RISKy Stuff (Mike Padlipsky)
  Mailer Protocol Woes (Rob Austein)
  Risks in Synchronizing Network Clocks (Ann Westine for Jon Postel)
  Re: Moral vs. Technological Progress (Joel Upchurch)
  Risk Contingency Planning -- Computers in Mexico (Mike McLaughlin)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions must
  be relevant to the topic, objective, in good taste, and coherent.  Others
  will be rejected.  Please try to advance the discourse, rather than iterating
  and reiterating.  There is no fine line between what is acceptable and what 
  is not, and thus the moderator's decisions may occasionally seem arbitrary.  
  It is our clear intent to represent a diversity of views.  However, we
  cannot be expected to provide counterarguments when none are forthcoming.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: 18 Sep 85 15:48 EDT
From: WAnderson.wbst@Xerox.ARPA
Subject: Intellectual honesty and the SDI
Courtesy-of: minow%rex.DEC@decwrl.ARPA  (Martin Minow, DECtalk Engineering)

FROM AIList Digest       Friday, 20 Sep 1985      Volume 3 : Issue 125

At the recent IJCAI at UCLA I picked up a couple of papers at the GE
exhibit booth.  One of these,  entitled "A Tutorial on Expert Systems
for Battlefield Applications," (delivered at a meeting of the Armed
Forces Communications and Electronics Association last May) states that
"AI systems that incorporate human expertise may be the only way" to
fill the gap between availability of people and complexity of military
hardware.  In defense of this strategy the author states:

        - In contrast with humans, AI systems are good at handling the myriad
details of complex situations, such as often occur in military settings.

        - In contrast with other computational approaches that are more formal
and algorithmic, AI systems are more robust:  they are designed to deal
with problems exhibiting uncertainty, ambiguity, and inaccuracy.

I find it appalling (and frightening) that statements like this can be
presented in a technical paper to military personnel.   The author
(according to the references) has contributed widely to the AI field at
many conferences. It's simply ludicrous to state that current AI systems
are better in battlefield situations than humans.  What was the last AI
system that could drive a tank, carry on a conversation, and fix a
broken radio whilst under enemy fire?  The second comment is equally
misleading.  To  contrast "formal and algorithmic" with "robust" seems
to imply that algorithms and formal procedures are inherently not
robust.  On what is this claim based?  (There is no reference attached
to either statement.)  It sounds like a recipe for unreliable software
to me.

How can someone write this stuff?  I know, to make money.  But if this
is the kind of information that is presented to the military, and upon
which they make decisions, then how can we expect any kind of fair
assessment of the possible projects in the Strategic Computing (and
Defense) Initiatives?  How can this kind of misinformation be rebutted?

Bill Anderson

P.S. The full reference is available on request.

------------------------------

Date: 20 Sep 1985 18:33:30 EDT
From: PADLIPSKY@USC-ISI.ARPA
Subject: RISKy Stuff
To:   neumann@SRI-CSL.ARPA

... I suppose I might as well succumb to temptation and offer a couple 
of comments on stuffgoneby:

The most striking omission, to my mind, in the SDI discussion is (unless I
missed spotting it) the failure to draw the parallel to SAGE.  For those who
don't remember/know, the Semi-Automated Ground Environment was the
grandfather of the big, "man-machine" system, certainly in DoD and most
probably in the field as a whole.  It was intended to allow for the
interception of manned bombers.  It is widely acknowledged to have spun off
a lot of what became the state of our art (collective art, that is-- i.e.,
what I call the computer racket).  Like SAGE, SDI is probably dealing with
the wrong threat (since things don't have to go through the air to go boom
... and since things don't even have to go boom to rack up megadeaths).
Also like SAGE, SDI might have useful spinoffs (a 20-years-younger colleague
claims to be for it because it should help get him off this planet).
Unfortunately, unlike SAGE it seems to possess a real potential for
stampeding the presumed Bad Guys into doing something ... unfortunate. 

What a good thing we Men of Science know better than to reason
from analogy, eh?

...  muted cheers, map

------------------------------

Date: Fri, 20 Sep 1985  14:36 EDT
From: Rob Austein <SRA@MIT-XX.ARPA>
To:   mooremj@EGLIN-VAX
Cc:   neumann@sri-csla
Subject: Mailer Protocol Woes

I was actually a culprit in a similar mailer lossage earlier this
week.  The whole thing actually started out when I was dialed up to XX
on a noisy connection.  Improbable as it seems (although not quite on
the order of monkeys and Shakespeare), the random line noise managed
to generate the 7 character command sequence necessary to send off my
entire mail file as a single message to a major mailing list.  All 110
pages (=281600 bytes) worth of it.  Fortunately for the network (but
unfortunately for my reputation) the list happened to be the TOPS-20
maintainers' mailing list, so the message got killed off in pretty
short order.  I have since put in a couple of safeguards into my mail
reader environment so that this particular lossage can't happen again,
but since the real culprit was transmission line noise I have been
kind of nervous about reading my mail over dialups ever since....

--Rob

------------------------------

Date: 25 Sep 1985 09:07:47 PDT
Subject: Risks in Synchronizing Network Clocks (RFC956 Now Available)
From: Ann Westine <WESTINE@USC-ISIB.ARPA>
Plucked-From: Request-For-Comments-List: ;

A new Request for Comments is now available from the Network Information
Center in the <RFC> directory at SRI-NIC.ARPA.

RFC 956:

   Title:       Algorithms for Synchronizing Network Clocks 
   Author:      D. L. Mills
   Mailbox:     Mills@USC-ISID.ARPA
   Pages:       26      
   Characters:  68868

      pathname: <RFC>RFC956.TXT

   This RFC discussed clock synchronization algorithms for the 
   ARPA-Internet community, and requests discussion and suggestions for 
   improvements.  The recent interest within the Internet community in 
   determining accurate time from a set of mutually suspicious network 
   clocks has been prompted by several occasions in which errors were 
   found in usually reliable, accurate clock servers after thunderstorms
   which disrupted their power supply.  To these sources of error should
   be added those due to malfunctioning hardware, defective software and
   operator mistakes, as well as random errors in the mechanism used to 
   set and synchronize clocks.  This report suggests a stochastic model 
   and algorithms for computing a good estimator from time-offset 
   samples measured between clocks connected via network links.  
   Included in this report are descriptions of certain experiments which
   give an indication of the effectiveness of the algorithms.  
   Distribution of this memo is unlimited.

Public access files may be copied from the <RFC> directory at 
SRI-NIC.ARPA via FTP with username ANONYMOUS and password GUEST.

   The normal method for distribution of RFCs is for interested parties 
   to copy the documents from the NIC online library using FTP.  
   Requests for special distribution should be addressed to either the 
   author of the RFC in question or to NIC@SRI-NIC.ARPA.  Unless 
   specifically noted otherwise on the RFC itself, all RFCs are for 
   unlimited distribution.

Submissions for Requests for Comments should be sent to POSTEL@USC-ISIB.ARPA.

Requests to be added to or deleted from this distribution list should be
sent to NIC@SRI-NIC.ARPA.

--jon.

      [I include this item in the RISKS Forum for a very obvious reason:
       one of the nastiest of all problems in distributed computer systems
       and networks is the synchronization problem.  That so many
       seemingly correct algorithms have in fact been flawed is a very
       important consideration here.  PGN]

------------------------------

Date: Wednesday, 25 Sep 1985 16:27-EDT
To: risks@sri-csl.arpa
Subject: Re: Moral vs. Technological Progress
X-From: joel@peora.UUCP (Joel Upchurch)

   >Actually  McCarthy's  original  comment  presupposes  that  moral  and
   >technological  progress  are comparable.  It is that assumption that I
   >disagree with.  Ethics and the attendant morality provide the  context
   >within  which  all activity, and in particular technological progress,
   >exists.  Morality and technology are not substitutes for  one  another
   >and  moral  progress  is  not  dependent on technology nor vice versa.
   >There is always technological progress  attendant  to  moral  progress
   >just because there is always technological progress.

I don't think that there is any such thing as an absolute moral standard.
Morality is simply a set of customs that evolves by trial and error to
improve the survival chances of the social group (family, tribe, nation
whatever).  Notice that this has has little to do individual survival and
that a moral principle, such as patriotism, may cause the individual to get
killed.

Now I think it fairly easy to see that the capacity to put group survival
ahead of self-interest is an important genetic trait and that tribes of
people that had this trait would be more likely to survive that tribes that
didn't.  That is not to say that this moral capacity doesn't vary greatly
from one person to the next or that even that it may not be more fully
realized in one person than another because of upbringing.  It is even
possible that, because of some genetic error, some people may be born
without a moral capacity, just like they might be born without arms or legs.

The point I'm trying to make is although there will always be these survival
customs we call morality, the nature of the customs is heavily dependent on
the context they evolve in.  Thus the morals of a herding society may be
greatly different from those of an agrarian one.  Even two societies in
similar contexts may evolve different moral solutions to the same problems.
This implies that if the context in which a society operates changes, then
the morals of that society will have to change too.  In recent centuries the
most important changes in the social context have been caused by technology.
Thus the morals appropriate to a agrarian society are not always appropriate
to an industrial one or those of an industrial society to a post-industrial
one.

Moral progress means the evolution of survival customs more appropriate to
the current context.  The trouble in recent centuries has been that our
ability to evolve new technology has outstripped our capacity to evolve the
appropriate morality for it.  There is a strong tendency to stick to the
morality that one learns as a child, even if it not appropriate to the
current situation.

Our current problem is that we have a technology that supplies us with ICBMs
and a morality that includes national patriotism.  Now it is obvious to any
thinking person that this is a serious dilemma.  Some people argue that we
should adopt a new morality, more appropriate to this technology, and indeed
in the long term they are correct, although one can argue what is the most
effective moral solution to this problem.

The trouble is that moral solutions evolve slowly.  Even today much of our
morality is left over from our nomadic and agrarian heritage and has limited
relevance in our modern society.  In order to apply the long term solution,
we must first survive in the short term, and in that short term technical
band-aids, like the SDI, are appropriate.  Technology put us into this
situation and I don't see why we can't ask technology to assist us in the
solution.  To think that we can solve the problem by an overnight revolution
in human nature is wishful thinking of the most dangerous sort.

     Joel Upchurch
     Perkin-Elmer Southern Development Center
     2486 Sand Lake Road/ Orlando, Florida 32809/ (305)850-1031
     {decvax!ucf-cs, ihnp4!pesnta, vax135!petsd}!peora!joel

------------------------------

Date: Sat, 21 Sep 85 13:25:25 edt
From: mikemcl@nrl-csr (Mike McLaughlin)
To: risks@sri-csl.ARPA
Subject: Risk Contingency Planning -- Computers in Mexico

This is not info, but a call for info.  I have no idea how bad the computer
bug has bit in Mexico, but the current news & TV coverage of the disaster
suggest that whatever computing/networking is going on must have been
affected.  If any RISKS reader knows some statistics or details about
computer usage in Mexico, and can give insights as to what happened, what
backup and alternative modes were in place, how well they worked... etc.  It
would certainly help me, and perhaps many others in planning for the
disaster that we hope never comes.  Perhaps we could discuss just what we
want to find out, and then go after the answers formally, if there are no
informal contacts on the net.  - Mike

       [This item may seem a little odd for the RISKS Forum, but when you 
        realize that Mike's Navy job involves expecting the unexpected risks 
        in computer and network usage, and planning what to do, it is indeed
        relevant.  By the way, in Mexico, as usual in times of disaster, the
        ham-radio buffs lept in to help.  Computer networks and internets
        also have a role to play.  I'm ready with my battery-operated
        portable terminal.  (This issue is being sent from DC.)  PGN]

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂26-Sep-85  1704	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Monday's PLANLUNCH   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 26 Sep 85  17:04:28 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Thu 26 Sep 85 16:59:56-PDT
Date: Thu 26 Sep 85 16:58:54-PDT
From: LANSKY@SRI-AI.ARPA
Subject: Monday's PLANLUNCH
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, phayes@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    araya@XEROX.ARPA, frayman@XEROX.ARPA, suchman@XEROX.ARPA, weld@XEROX.ARPA,
    mittal@XEROX.ARPA, dekleer@XEROX.ARPA, trigg@XEROX.ARPA

	          PROCESSES, SIMULTANEITY AND CAUSALITY

        	          Michael P.  Georgeff
	              Artificial Intelligence Center
        	           SRI International

  	            11:00 AM, MONDAY, September 30
       SRI International, Building E, Room EJ228 (new conference room)


The notion of process is essential for reasoning about the behavior of
multiple agents or single agents in dynamic worlds.  In this talk, we
show why reasoning about process is so important, and contrast this
with other approaches in AI which are primarily based on the allowable
behaviors of agents.  An algebra of processes based on events is given.
We then show how events can be represented as changes of world state, 
and how state properties can be inferred from the model. Interestingly, 
no STRIPS-like assumption is involved in the definition of events, thus 
allowing a proper model-theoretic semantics.  One of the most important 
features of the model is a hiding operation.  This provides an abstraction
capability that can be used to avoid the combinatorial explosion
typical of other AI approaches.  Finally, we introduce a notion of
causality between events and processes.  This, together with the
notion of simultaneous actions and hiding operations, allows us to
avoid most of the problems associated with the frame problem. 

-------

∂26-Sep-85  2229	sinha%gmr.csnet@CSNET-RELAY.ARPA 	add sinha@gmr to NAIL
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 26 Sep 85  22:29:41 PDT
Received: from CSNET-RELAY.ARPA by su-aimvax.arpa with Sendmail; Thu, 26 Sep 85 22:29:01 pdt
Received: from gmr by csnet-relay.csnet id ai17481; 26 Sep 85 10:26 EDT
Date:     Wed, 25 Sep 85 21:43 EST
From: Sarvajit←Sinha <sinha%gmr.csnet@CSNET-RELAY.ARPA>
To: nail@su-aimvax.ARPA
Subject:  add sinha@gmr to NAIL

∂27-Sep-85  0153	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #40
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  01:53:39 PDT
Date: Thursday, September 26, 1985 10:46PM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #40
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Friday, 27 Sep 1985       Volume 3 : Issue 40

Today's Topics:
               Implementation - Destructive Assignment,
                 LP Philosophy - Hewitt's Challenge,
                       LP Library - Benchmarks
----------------------------------------------------------------------

Date: 26 Sep 85 8:35:05 PDT
From: Deutsch.PA@Xerox.ARPA
Subject: Destructive Assignment

With regard to destructive assignment: I am constantly amazed at how
people who appear to know little about the last 20 years of
language/compiler implementation technology make statements about
"language X is intrinsically inefficient".  Even today, the amount of
compiler technology invested in Lisp is miniscule compared to Fortran.
Why??  I think it is because sub-field specialization in CS has gotten
to the point where almost no one in the Lisp/AI sub-field studies the
optimizing compiler sub-field.  Look at what David Warren's compilers
did for Prolog!

Specifically with regard to the assertion that "applicative languages
are horrendously inefficient because they have to copy all the time":
static analysis, not much more subtle than performed by current
optimizing compilers, can identify patterns of sharing, and can result
in compiled code that implements the semantics of the source program
by updating in place in many cases.  Sometimes this can even be done
dynamically with reference counts -- the overhead of reference
counting is well documented at around 10% using deferred reference
counting for stack frames (also known as Deutsch/Bobrow reference
counting), climbing to around 50% if you count everything.

There is a real chicken-and-egg problem here: people think
applicative languages are inefficient, so they don't take them
seriously, so the work doesn't get done that would produce the
efficient implementations that would lead people to take them
seriously!

------------------------------

Date: Wed 25 Sep 85 11:22:32-PDT
From: Fernando Pereira <PEREIRA%sri-candide@sri-marvin>
Subject: Prolog vs. Lisp (again!)

I humbly confess my incompetence at the debating style Carl
Hewitt is using in the Prolog vs. Lisp discussions, which
seems to consist in ignoring the FACTUAL points of other
contributions and just continuing to repeat the same OPINIONS.

It is a FACT that no practical Prolog system is written entirely
in Lisp: Common, Inter or any other. Fast Prolog systems have
been written for Lisp machines (Symbolics, Xerox, LMI) but their
performance depends crucially on major microcode support (so
much so that the Symbolics implementation, for example, requires
additional microstore hardware to run Prolog). The reason for
this is simple: No Lisp (nor C, for that matter...) provides the
low-level tagged-pointer and stack operations that are critical
to Prolog performance.

The fact that Lisp is not good enough as an implementation
language for Prolog should not be considered as a weakness of Lisp,
BECAUSE Lisp was not designed for such low-level operations in the
first place. In fact, NO ``high-level'' programming language that I
know of provides those kinds of operations, and for the very simple
reason that, being high-level languages, they have no business in
exploring the recesses of particular machine architectures. ALL
fast Prolog systems that I have seen (some of which I helped
implement) rely on a careful exploitation of the underlying machine
architecture.

By the same argument, the fact that Lisp cannot be efficiently
implemented in Prolog cannot be the basis of a valid criticism of
Prolog. Prolog is not a systems programming language, and in any
case a good Lisp implementation must be carefully coupled to the
underlying machine architecture -- so much so the the fastest Lisps
rely on specialized architectures!

It seems clear to me that no single existing programming language
can be said to provide a ``foundation'' for AI. In fact, the very
notion of a programming language providing a foundation for a
scientific  subject seems to me rather misguided. Does Fortran
provide  a ``foundation'' for physics?  The relation between AI
problems, formal  descriptions and programming concepts is far too
subtle for us to  expect a ``foundation'' for AI in a mere
programming language.

The crusading tone of Hewitt's comments is also rather
unsettling.  AI researchers will use whatever language they
feel most comfortable  with for the problem they are working
on, without need for any  guidance from on high as to the
ultimate suitability of that language. If more researchers use
Prolog, is that a threat to Lisp users? If I do a piece of AI
research using Prolog, will it not be judged according to its
content, independently of programming language?

That kind of battle might be very important for AI software
companies, but surely we should not let marketing hype get in
the way of our research. I am sitting at a Sun workstation typing
this, with a Prolog window just to the right. Will my research be
useless to someone who sits at a Xerox 1109? If I walk down the
corridor and write a Lisp program on a Symbolics machine (as I
have done and surely will continue to do), will THAT work have
a different value? If I decide to use Hoare's CSP for the
navigation component of our robot Flakey, will I be then outside
AI, because I am not using an ``official'' AI language?

With respect to Rick McGeer's points: there are some elegant ways
of compiling Prolog into Lisp, particularly into those statically
scoped variety (or into other statically-scoped languages such as
Algol-68...). I have reason to believe that a compiler along these
lines would produce code considerably faster than the 100 LIPS he
reports, even though still much slower than what is attainable with
a lower-level implementation.

The idea has been around in the ``folklore'' for a long time, I do
not know to whom to attribute it. Basically, given predicate
definitions like

        p :- q, r.
        p :- s.

        q.

we get the code

        (defun p (continuation)
           (progn
              (q (function (lambda () (r continuation))))
              (s continuation)
           )
        )

        (defun q (continuation)
           (funcall continuation)
        )

This is the basic control skeleton, extra code is needed for
unification. This technique can be made more complicated (and
more efficient...) in a number of ways, and it will be somewhat
more space-efficient on a tail-recursive Lisp.

As to the speed of Franz, if all the appropriate compilation
flags are used an append function written in Lisp runs at
around 25K calls/sec. on a VAX 780, if I remember correctly.
The lower speed he listed is the result of using the (default)
slow function-calling protocol that allows breaks and stack
backtraces. As a comparison, one of the commercial compiled
Prologs does 23K calls/sec. on the equivalent test.

-- Fernando Pereira

------------------------------

Date: Wed 25 Sep 85 11:34:13-PDT
From: Fernando Pereira <PEREIRA%sri-candide@sri-marvin>
Subject: My original rebuttal of Hewitt's flame

Some combination of Carl's and editing has made it impossible
to distinguish in his message what I said and his comments on
what I said. Here is my original note, which I sent only to
the AIList, where the discussion had migrated from COG-SCI
(which I don't read, time being limited...)

Carl Hewitt's message is based on several misconceptions:

1. (the least interesting one) All the so-called commercially
viable Prolog systems in Lisp are not really Prolog systems
written IN Lisp, but rather Prolog systems written FOR Lisp
machines. Or is it that a microcode sublanguage or Lisp machine
pointer-smashing operations are part of List as we know it? Without
those machine-level operations, those Prolog systems would run too
slow and use too much memory to be useful for serious Prolog
programming. From the Prolog implementation point of view, what
is important about the Lisp machines is not that they run Lisp,
but that they can be microcoded and have good support for tagged
data types and stack operations.

2. If the decisions (actions) of a system are not determined by
its inputs, the system is nondeterministic. Nondeterminism in a
system can be either an artifact of our incomplete knowledge (or
lack of interest) of the detailed operation of the system; or it
can be ``real physical'' nondeterminism. It would take us to far
to discuss whether the second kind of nondeterminism is ``real''
or also an artifact. In any case, most uses of nondeterminism,
say in models of concurrency, are of the first kind, and can be
expressed appropriately in various temporal/dynamic logics.
Admittedly, these are not Prolog, but then Common Lisp is not Lisp
1.5! (Prolog is 13 years old, Lisp 25).

3. The first logic course dictum ``from a contradiction one can
conclude anything'' is getting in the way. Notice that the dictum
says ``can'', not ``must''. There is an enormous difference between
things that are in principle true and things that an agent knows to
be true in a way that can affect its action. An agent might know
``p'' and ``not p'', but it might well never come to infer the
dreaded totally unrelated ``q'' which IN PRINCIPLE follows from
the contradiction. This inference might not happen either because
of inference control mechanisms of the agent (eg. it uses the set
-of-support strategy) or because the agent's logic is just TOO WEAK
to conclude anything from a contradiction (vide Hector Levesque's
work in the proceedings of the last AAAI). In any case, Horn clauses
(the basis of Prolog) are too weak to represent contradictions... :-)

4. The question of whether ``taking action'' fits in a logic paradigm
tends to be answered negatively after an hour's worth of
consideration.  If you persist for several years, though, this
question becomes a source of insight on the relations between
knowledge, state and action that is not available to those who started
by dismissing the question after that initial hour. There is just too
much work on logics of knowledge and action in AI and computer science
for me to try to discuss it here. Some of this work has been applied
to logic programming, either in the form of new logic programming
languages based on temporal or dynamic logics or in representations of
temporal reasoning and decision in, say, Prolog.

5. It is curious to see someone by implication defend Lisp as a
language for expressing the taking of action! We know by now that the
most difficult issue in ``reactive systems'' is not EXPRESSING action,
but rather keeping it under control to prevent unwanted interactions.
In this area, less is REALLY more, and highly complex languages like
Lisp are just not suitable for the formal reasoning about programs
that is needed to help us believe our programs do what we intend. To
those who say ``...but we can keep to a carefully constrained subset
of Lisp, use only message passing for interactions,...'' I will answer
that the history of all large Lisp programs I know (and I think that
is a representative sample) is littered with the brutalized corpses of
constrained programming styles. Anyone who has looked at the current
flavor mechanism in Zetalisp and its use in the window system will
know what I mean...

6. Underlying Carl Hewitt's misconceptions is the old chestnut ``logic
is static, systems are dynamic''. Any language, be it first-order
logic or Lisp, is static; it is its USE which is dynamic (changes the
state of communicating agents). A good analogy here is the use of
differential equations to model dynamic systems in classical
mechanics. The differential equations themselves do not say directly
what happens when (they are ``static'' in Hewitt's jargon). It is
their SOLUTIONS that tell us the sequence of events. Even the
solutions are given as static objects (functions from an open interval
of the reals to some space). Does anyone worry that such equations do
not ``really'' describe the dynamic behavior of a system? Is it not
possible to combine such ``static'' entities with an incremental
solution procedure to build systems that actually control their
(classical mechanical) environment?

-- Fernando Pereira

------------------------------

Date: Tue, 24 Sep 85 09:43:30 edt
From: "Bruce T. Smith" <bts%mcnc.csnet@CSNET-RELAY>
Subject: A set of PROLOG benchmarks...

I received the following file late in the Summer, with permission
to post it to the net.  It's a set of Prolog benchmarks, by folks
at Tektronix and Portland State University.  The times included
are for C-Prolog on a Tektronix Magnolia-- as I understand it, a
predecessor of the 4404.

I've run them-- with minor changes-- with Quintus Prolog, on a VAX
785 at MCNC, and with C-Prolog, UNSW Prolog and MU-Prolog at UNC
Chapel Hill, also on a 785. (Both running 4.2bsd UNIX.) For those
folks with other Prologs or other machines, I will volunteer to
collect the results you mail and post a summary to the net later
in the Fall.

I'd also like to hear comments on the benchmarks themselves.  For
that, a discussion on the net is more appropriate.

[ I've left this file in the <Prolog> directory as:

                  TEKTRONIX←PSU-BENCHMARK.PL  -ed ]

------------------------------

End of PROLOG Digest
********************

∂27-Sep-85  0902	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.17  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  09:01:49 PDT
Date: Fri 27 Sep 85 07:51:56-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.17
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-LIST: RISKS-FORUM Digest  Friday, 27 Sep 1985  Volume 1 : Issue 17

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  SDI debate announcement
  Minor risk to the pocket book (Eugene Miya)
  Social Impacts of Computing: Graduate Study at UC-Irvine (Rob Kling)
  Friendly enemy test teams (John Mashey)
  More protocol goofs (Dave Curry)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: 26 Sep 1985 17:48-EST
From: genrad!teddy!lkk@mit-eddie.MIT.EDU
Subject: SDI debate announcement
To: risks@sri-csl.arpa

Dear Colleague,

Computer technology plays an ever-increasing role in the arms race.
The Strategic Defense Initiative (`Star Wars') is a leading example
of a military system in which almost complete reliance will be placed
on computerized decision making.  The feasibility and desirability of
this system are currently undergoing serious public debate.

On Monday, the 21st of October, at 8:00 pm in M.I.T.'s Kresge Auditorium, the
M.I.T. Laboratory for Computer Science and Computer Professionals for Social
Responsibility are co-sponsoring a public forum designed to raise many of the
technical issues surrounding the Strategic Defense Initiative (SDI).

Professor Michael Dertouzos, Director of the M.I.T. Laboratory for Computer
Science, will moderate a debate on the feasibility of software development
for the SDI project.  Dr. Danny Cohen (Director of the Systems Division at
the University of Southern California's Information Sciences Institute and
Chairman of the Strategic Defense Initiative Organization panel on Computing
in Support of Battle Management (SDIO/CSBM)) and Professor Charles Seitz
(Professor of Computer Science, California Institute of Technology and
member, SDIO/CSBM)) will speak in favor of the SDI proposal.  Professor
David Parnas (Lansdowne Professor of Computer Science, University of
Victoria, Canada) and Professor Joseph Weizenbaum (Professor of Computer
Science, Massachussetts Institute of Technology) will take the position that
such a software development project is infeasible.

Professor Parnas' resignation from the SDIO/CSBM panel in June of
this year, and the set of memos he wrote discussing the infeasibility
of the SDI project attracted extensive press coverage in June of this
year.

CPSR will be holding a reception for the speakers at La Groceria (853
Main Street in Cambridge) between 6:30 and 7:30.  Please join us for
dinner and an opportunity to meet some of the panelists.  The $25 per
plate donation will help us cover expenses for this forum and related
projects.  Please RSVP to Mark Vilain at (617) 648-4325.

Earlier that afternoon, the M.I.T. Technology and Culture Seminar
will sponsor a talk by Dr. James Ionson, Director of the Office of
Innovative Science and Technology for the SDIO.  After Dr. Ionson
describes the general research goals of SDI, two M.I.T. professors
will respond with varying views on why they have chosen to accept or
refuse funding for research from the SDIO.  A student representative
will report on student reaction to Star Wars projects on campus.
This talk will be held at MIT in building 9, room 150 at 4:00 p.m.

------------------------------

From: eugene@AMES-NAS.ARPA (Eugene Miya)
Date: 26 Sep 1985 1652-PDT (Thursday)
To: risks@sri-csl.ARPA
Subject: Minor risk to the pocket book

Here is a minor man-machine risk which occurred today (9/26) in the Silicon
Valley at El Torito [popular Mexician food chain] at lunch.  We arrived for
a late lunch and found that our bill was 50% over what appeared in the menu.
The cash register [one of those computer systems where they press buttons
rather than prices] was running on the dinner menu prices rather than the
lunch menu prices.  Since we arrived late, everybody else [e.g., hundreds of
people] were over-charged for lunch that day and perhaps earlier.  This
implies several things about the way customers in Si Valley treat their
bills [rich?  no verification? ...]  What about the restaurant?

From the Rock of Age Home for Retired Hackers:

--eugene miya,   NASA Ames Research Center,   eugene@ames-nas.ARPA

------------------------------

Date: Fri, 27 Sep 85 12:01:18 EST
From: munnari!mulga.oz!bjpt@seismo.CSS.GOV (Benjamin Thompson)
To: risks@sri-csl.ARPA
Subject: Re: Technology and Morality
Organization: Computer Science, University of Melbourne

Nicholas Spies writes [in RISKS-1.13]:

  >             ...                 Hitler opened the Pandora's Box of applying
  >high-tech to warfare and it worked (at least until a higher-tech response
  >prevailed).

Technology has been successfully applied to warfare for millenia.  Alexander
the Great didn't win through having a bigger army (he didn't); he had better
weapons (e.g. he found out that ballistae propel spears faster than people do).
(and he had better trained soldiers etc. - education and technology basically).

  >	     After WWII a new era was born in which global political power no
  >longer rested on moral authority but on a command of the new applied
  >sciences and scientists.

Nothing could be further removed from morality than the way in which global
political power is grabbed and maintained.  It has always been based upon
physical strength (which usually, but not always, corresponds to technological
strength).  History is a never-ending sequence of examples.

------------------------------

Date: 26 Sep 1985 0920-PDT
From: Rob-Kling <Kling%UCI-20B@UCI-ICSA>
Subject: Social Impacts of Computing: Graduate Study at UC-Irvine
To: risks@SRI-CSL
                                CORPS:
                        Graduate Education in
            Computing, Organizations, Policy, and Society
               at the University of California, Irvine

     This graduate concentration at the University of California,
Irvine provides an opportunity for scholars and students to
investigate the social dimensions of computerization in a setting
which supports reflective and sustained inquiry.

     The primary educational opportunities are PhD concentrations in
the Department of Information and Computer Science (ICS) and MS and
PhD concentrations in the Graduate School of Management (GSM).
Students in each concentration can specialize in studying the social
dimensions of computing.

     The faculty at Irvine have been active in this area, with many
interdisciplinary projects, since the early 1970's.  The faculty and
students in the CORPS have approached them with methods drawn from the
social sciences.

     The CORPS concentration focuses upon four related areas of inquiry: 

 1.  Examining the social consequences of different kinds of computerization
     on social life in organizations and in the larger society.

 2.  Examining the social dimensions of the work and organizational
     worlds in which computer technologies are developed, marketed,
     disseminated, deployed, and sustained.

 3.  Evaluating the effectiveness of strategies for managing the
     deployment and use of computer-based technologies.

 4.  Evaluating and proposing public policies which facilitate the
     development and use of computing in pro-social ways.


     Studies of these questions have focussed on complex information
systems, computer-based modelling, decision-support systems, the
myriad forms of office automation, electronic funds transfer systems,
expert systems, instructional computing, personal computers, automated
command and control systems, and computing at home.  The questions
vary from study to study.  They have included questions about the
effectiveness of these technologies, effective ways to manage them,
the social choices that they open or close off, the kind of social and
cultural life that develops around them, their political consequences,
and their social carrying costs.

     CORPS studies at Irvine have a distinctive orientation -

(i) in focussing on both public and private sectors,

(ii) in examining computerization in public life as well as within
      organizations,

(iii) by examining advanced and common computer-based technologies "in
      vivo" in ordinary settings, and

(iv) by employing analytical methods drawn from the social sciences.


         Organizational Arrangements and Admissions for CORPS

     The CORPS concentration is a special track within the normal graduate
degree programs of ICS and GSM.  Admission requirements for this
concentration are the same as for students who apply for a PhD in ICS or an
MS or PhD in GSM.  Students with varying backgrounds are encouraged to apply
for the PhD programs if they show strong research promise.

     The seven primary faculty in the CORPS concentration hold appointments
in the Department of Information and Computer Science and the Graduate
School of Management.  Additional faculty in the School of Social Sciences,
and the program on Social Ecology, have collaborated in research or have
taught key courses for CORPS students.  Our research is administered through
an interdisciplinary research institute at UCI which is part of the Graduate
Division, the Public Policy Research Organization.

Students who wish additional information about the CORPS concentration
should write to Professor Rob Kling (Kling@uci-icsa), Department of
Information and Computer Science, University of California, Irvine, Irvine,
Ca. 92717, 714-856-5955 or 856-7548, or Professor Kenneth Kraemer
(Kraemer@uci-icsa), Graduate School of Management, University of California,
Irvine, Irvine, Ca. 92717, 714-856-5246.

------------------------------

Date: Wed, 25 Sep 85 01:17:53 pdt
From: mips!mash@glacier (John Mashey)
To: risk@sri-csl.ARPA
Subject: Friendly enemy test teams

John McCarthy <JMC@SU-AI.ARPA> writes in RISKS-1.10:

  > 2. My opinion is that if the physics of the problem permits a good
  > anti-missile defense the programs can be written and verified.  However, it
  > will be quite difficult and will require dedicated work.  It won't be done
  > by people who are against the whole project.  Computer checked proofs of
  > program correctness will probably play some role. So will anticipating what
  > kind of bugs would be most serious and putting the biggest effort into
  > avoiding them.  Having many people go over and discuss all the critical
  > parts of the program will also be important.
  
Perhaps the best way to make it work WOULD be to have a test team of
people (who might be skeptics, at least) trying to break it.  Most large
complex projects that actually worked, at least that I've seen, have
succeeded at least partly because they had a large test team who didn't
believe anything worked until it could get past the worst of their tests.
I don't know what the ratio is elsewhere, but many complex ATT/BTL projects
allocated 30-50% of the staff to building test frameworks designed to stress
the system under test.  Consider the recent history of evaluation of
new military systems (like the Sergeant York).  It's very hard for the
builders of something to evaluate it well; you need a good enemy for that.

     [Tiger teams have indeed had some success in finding the more obvious
      program bugs, but in general many flaws may remain.  This topic has been 
      raised superficially in past issues.  Perhaps we are ready for some
      detailed discussions on the strengths and limitations of testing.  PGN]

------------------------------

Date: Thu, 26 Sep 85 22:29:41 EST
From: davy@purdue-ecn.ARPA (Dave Curry)
To: risks@sri-csl.arpa
Subject: More protocol goofs

    [The original message contained 8576 characters, almost exclusively
     headers.  I have pruned it to give just the flavor.  PGN]

I'm forwarding this as a wonderful example of protocols getting
completely hosed... this mail of mine bounced for some unexplained
reason.  I resent the message the same day this came back, and it
went through just fine.

Looking at the headers should make the problem more than obvious...

--Dave Curry

Return-Path: <decvax!mcnc!unc!unc!unc!unc!unc!unc!unc!unc!mailer-daemon>
Received: from ee.ECN by ec.ECN; Thu, 6 Sep 84 19:48:01 est (4.12/5.20)
From: decvax!mcnc!unc!unc!unc!unc!unc!unc!unc!unc!mailer-daemon
Received: by ee.ECN; Thu, 6 Sep 84 19:47:35 est (4.12/5.20)
Return-Path: <decvax!mcnc!unc!unc!unc!unc!unc!unc!unc!unc!mailer-daemon>
Received: by decvax.UUCP (4.12/1.0)
	id AA24764; Thu, 6 Sep 84 02:25:55 edt
Received: by mcnc (4.12/4.7) id AA00613; Wed, 5 Sep 84 18:51:54 edt
Original-From: <unc!unc!unc!unc!unc!unc!unc!unc!mailer-daemon@mcnc>
     [... and so on iteratively, ad nauseum, down to... ]
Received: by unc (4.12/4.7) id AA06045; Wed, 5 Sep 84 09:07:16 edt
Original-From: <mailer-daemon@mcnc>
Received: by mcnc (4.12/4.7) id AB15479; Wed, 5 Sep 84 08:01:18 edt
Date: Sat, 1 Sep 84 10:03:51 est
Original-From: Mail Delivery Subsystem <MAILER-DAEMON@mcnc>
Subject: Returned mail: Unable to deliver mail
To: mcnc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!pur-ee!davy@unc

   ----- Transcript of session follows -----
554 sendall: too many hops (30 max)

   ----- Unsent message follows -----
Received: by mcnc (4.12/4.7) id AA15479; Wed, 5 Sep 84 08:01:18 edt
From: <mcnc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!pur-ee!davy@unc>
Received: by unc (4.12/4.7) id AA03055; Wed, 5 Sep 84 07:04:54 edt
Original-From: <mcnc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!unc!mcnc!pur-ee!davy@mcnc>
   [...]

   [I am reminded of the tale of the unattended British power station
    equipped with an automatic calling unit to report troubles.  When it
    finally had to dial the emergency reporting number, it received a
    recorded message that the number it had dialed was not valid.  PGN]

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂27-Sep-85  0930	JOHN@SU-CSLI.ARPA 	Philosophy 287A 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  09:30:21 PDT
Date: Fri 27 Sep 85 09:26:02-PDT
From: John Perry <JOHN@SU-CSLI.ARPA>
Subject: Philosophy 287A
To: folks@SU-CSLI.ARPA
cc: philosophy@SU-CSLI.ARPA

		
Philosophy 287A: Topics in Language and Information.

Fall Quarter, Mondays at 1:15, Ventura Hall Seminar Room.

The topic will be folk psychology and cognitive science.  We will
discuss Stephen Stitch's recent book FROM FOLK PSYCHOLOGY TO
COGNITIVE SCIENCE: THE CASE AGAINST BELIEF, in which he argues
that our ordinary psychological concepts, such as belief, are
pretty confused and have not place in cognitive science.  I
expect to develop a line of argument opposed to Stitch, based on
the point of view developed in Part D of Barwise&Perry SITUATIONS
AND ATTITUDES, maintaining that folk psychology isn't such a
mess, and might have some role in cognitive science.  In addition
to Stitch's book, we will examine a couple of key articles by
Fodor and also Dan Dennett's attack on belief, "Beyond Belief". 
Background reading in recent philosophy of mind and philosophy of
psychology will also be assigned.
≠
-------

∂27-Sep-85  0934	BETSY@SU-CSLI.ARPA 	1986 Site Visit
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  09:34:45 PDT
Date: Fri 27 Sep 85 09:30:25-PDT
From: Betsy Macken <BETSY@SU-CSLI.ARPA>
Subject: 1986 Site Visit
To: folks@SU-CSLI.ARPA




CSLI's 1986 SDF Site Visit will be on Tuesday, June 24.  SDF is
going to invite some of the same people who came last year, but
may include some others as well.

We don't know any other details, but the visit will play an important
role in determining whether or not SDF gives additional money to CSLI.
We'd like to have our researchers well represented so that our
presentation can show off all aspects of our research, so we hope that
with this early notice many of you can arrange to be here on that day.
We know we can't expect everyone to be here, but we hope you will take
the site visit into account in making your long-term plans.

Thanks.
Betsy


-------

∂27-Sep-85  1554	BSCOTT@SU-SCORE.ARPA 	Lost Articles, Yesterday's Reception  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  15:54:48 PDT
Date: Fri 27 Sep 85 15:49:06-PDT
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: Lost Articles, Yesterday's Reception
To: CSD@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12146691477.25.BSCOTT@SU-SCORE.ARPA>



The caterer for yesterday's reception has reported that one large black
lacquer tray and three serving spoons are missing from the serving 
paraphernalia they picked up from the MJH kitchen today.  If anyone has
seen any of the missing items, I will appreciate hearing from her or him.
If they can't be located, the Department will have to pay for the loss.

Thanks for your help.

Betty
-------

∂27-Sep-85  1606	LIBRARY@SU-SCORE.ARPA 	Math/CS Library Fall Hours 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  16:06:21 PDT
Date: Fri 27 Sep 85 15:53:06-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library Fall Hours
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12146692204.49.LIBRARY@SU-SCORE.ARPA>

The Math/CS Library has started its fall hours.  Monday through Thursday
8am to 10pm.  Friday 8am to 6pm. Saturday 10am to 5pm. Sunday 1pm to 
10pm.

H.Llull
-------

∂27-Sep-85  1615	LIBRARY@SU-SCORE.ARPA 	Socrates: New Version To Be Activated On Sunday, September 29th    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Sep 85  16:15:36 PDT
Date: Fri 27 Sep 85 16:06:21-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates: New Version To Be Activated On Sunday, September 29th
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12146694618.49.LIBRARY@SU-SCORE.ARPA>

The new version of Socrates should be available this Sunday.  The new
Socrates will have a Guided mode and Command mode that will work a little
differently then the old Lookup and Command modes.  There will be less of
a barrier between the two modes.  I will be sending out more messages next
week concerning this new feature and other new features.

For those faculty and students in Computer Science, you will be most interested
in the fact that the new Technical Reports File includes all the reports 
collected in the Math/CS Library.   The Technical Reports File is a separate
file and must be selected before one can search it either in the Guided mode
or Command mode.

If you have any questions or concerns about the new Socrates, you can use
the SUGGEST command to express your opinions.  I would also be interested
in hearing from you if you are experiencing any diffculties in searching
the new version of Socrates.

For those of you who do not have an individual Socrates account, you may
come to the Math/CS Library to fill out a form and pick up your account
number.

I will announce on the bulletin board when I have the guide sheets to the
new version of Socrates.

Harry Llull
-------

∂29-Sep-85  1439	@SU-SCORE.ARPA:lantz@su-gregorio.arpa 	Re: speaking at orientation day
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Sep 85  14:39:44 PDT
Received: from su-gregorio.arpa by SU-SCORE.ARPA with TCP; Sun 29 Sep 85 14:36:30-PDT
Received: by su-gregorio.arpa with Sendmail; Sun, 29 Sep 85 14:36:40 pdt
Date: 29 Sep 1985 1436-PDT (Sunday)
From: Keith Lantz <lantz@su-gregorio.arpa>
To: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Cc: MikeDixon.pa@XEROX.ARPA, faculty@SU-SCORE.ARPA, WIEDERHOLD@SUMEX-AIM.ARPA,
        REGES@SU-SCORE.ARPA
Subject: Re: speaking at orientation day
In-Reply-To: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA> / 
		Tue 17 Sep 85 09:12:14-PDT.
             <12143997789.10.TAJNAI@SU-SCORE.ARPA>

Although I'm at least two weeks removed from the bulk of this
conversation (having been away during that period), I would strongly
favor moving the CS Colloquium back to the vicinity of Margaret Jacks
(or somewhere on a somewhat direct line between there and the CIS bldg.)
whether or not that meant keeping it on TV.  I am convinced that in the
first two years I was at Stanford, when it was in Jordan, many more 
faculty and staff attended than do now.  As for the effect of TV
itself, my own experience is not positive -- whether for the colloquia
or for courses -- but I appreciate the financial benefits.

Keith

∂30-Sep-85  0157	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #41
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  01:57:41 PDT
Date: Sunday, September 29, 1985 4:59AM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #41
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Monday, 30 Sep 1985       Volume 3 : Issue 41

Today's Topics:
                    Query - Prolog for Apple IIe,
                        Challenge - Ken Law's,
                 LP Philosophy - Hewitt's Challenge,
           Implementation - DCG's & Destructive Assignment
----------------------------------------------------------------------

Date: Wed, 25-Sep-85 14:38:38 PDT
From: (Craig D. Singer) mcnc!duke!cds@UCB-Vax.ARPA
Subject: Prolog for Apple IIe

Does anybody know where we might purchase a non-copy-protected
version of Prolog for the Apple IIe?  This software is strictly
for educational use, and we have about three or four Apples we'd
like to be able to run it on simultaneously; hence, the need for
a copyable program.  One or two of the Apples are enhanced and
one or two of them are not, but that is not a major concern.
We simply need software which our AI grads can use without having
to handle the disk with extreme delicacy.  Alternatively, we would
consider purchasing (or, ahem, graciously accepting) the printed
or downloaded source.  If you can help, please send mail or post
something here...we read this newsgroup daily.

Thank you now for good deeds later!

-- Craig Singer

------------------------------

Date: Fri 27 Sep 85 16:14:09-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Connected-Component Extraction

My thanks [to Allen VanGelder] for considering the connected
-component extraction problem.  After sending it off to Phil-Sci,
I realized thatI was really asking for an automated programming
solution rather than a logic programming solution -- and I don't
expect anyone to come up with one.  The question remains, though,
whether expressing a problem in a logic formalism will help the
programmer develop an algorithmic solution.  I'm sure the answer
depends on the problem, the logic formalism, and the programmer.

The problem statement was a little imprecise.  I did indeed have a
2-dimensional array in mind, although I believe that any algorithm
would be essentially the same for higher dimensionalities.  I also
had just rectilinear neighbors in mind, although permitting diagonal
neighbors makes just as good a problem.  (There is a quirk in the
computational geometry, though; allowing diagonal connectedness
permits connected components that intersect each other.  This
plays havoc with containment relationships.  Some pattern
recognition systems use diagonal connectedness for foreground
objects and rectilinear connectedness for the background, but
that won't work for other than binary arrays.)

Shortly after submitting the problem, I ran across an algorithm
for solving it (in the binary case) using a pyramid of
interconnected processors.  The algorithm was so clever that I
can't remember it now. Maybe we do need automated programming
systems to derive optimal algorithms for any hardware or
circumstance, but some of the procedural solutions people have
invented (often for ill-defined problems) are awe-inspiring.

-- Ken Laws

------------------------------

Date: Thu, 26 Sep 85 10:29 EDT
From: Tim Finin <Tim%upenn.csnet@CSNET-RELAY.ARPA>
Subject: Lisp and Prolog

                           PROLOG/LISP = LISP/C

I'd like to amplify a point of view Clocksin put forth in V3
#39 in the great LISP vs. PROLOG debate.  Prolog (and Logic
Programming in general) and Lisp are both tools which are suited
for different tasks. Luckily, none of us is being forced to use
one to the exclusion of the other.

I've had to give more than my share of introductory AI talks over
the years. When the discussion gets around to Lisp I usually point
out that Lisp is especially attractive for experimental programming
situations, i.e. where you know what you want to accomplish but do
not yet have all of the details as to algorithms, data structures,
etc. worked out.  Once you've worked out the last detail, you can
re-implement your system in C, if you like, and gain the benefits
of a faster and smaller system.

Along these lines, I think that the slogan "Prolog is to Lisp as
Lisp is to C" is not too inaccurate.  I think that Prolog is even
better suited to initial, experimental and exploratory attempts to
attack a problem computationally.  I find that it is a very
convienent paradigm in which to get started.  Once I have a better
idea of how  to reprent a problem and how to manipulate the
representation, I can  re-implement it in Lisp and gain a faster
more steamlined system.

-- Tim

------------------------------

Date: Thu 26 Sep 85 15:45:15-PDT
From: Fernando Pereira <PEREIRA%sri-candide@sri-marvin>
Subject: DCGs and parsing strategy

First, let's get rid of one misconception about DCGs. DCGs
are not ``top-down'' any more than, say, context-free grammars.
I have top-down, bottom-up, left-corner, Earley, and various
chart-parsers for DCGs, all implemented in Prolog.

Now for the minimal bottom-up (actually, left-corner) DCG parser.
The parser is itself a DCG (a DCG metainterpreter, if you wish)
that takes rules given by the binary predicate ==>. To parse a
string S as having category C, call

        ?- down(C,S,[]).

Here is the code:

        down(Phrase) --> leaf(SubPhrase), lc(SubPhrase,Phrase).

        leaf([Word]) --> [Word].
        leaf(Phrase) --> { Phrase ==> [] }.

        terminals([]) --> [].
        terminals([Word|Words]) --> [Word], terminals(Words).

        lc(Phrase,Phrase) --> [].
        lc(SubPhrase,SuperPhrase) -->
           { Phrase ==> [SubPhrase|Rest] },
           right(Rest),
           lc(Phrase,SuperPhrase).

        right([]) --> [].
        right([Phrase|Phrases]) -->
           down(Phrase),
           right(Phrases).

Note that the sequence of symbols to the right of the arrow is
written as a list (rather than as a `,'-separated sequence) in
the object grammar. A few modifications are needed to handle
{...} conditions and terminal sequences [...]. It is possible
to make the code much more efficient by partial evaluation
with the object grammar and other compile-time optimizations.

-- Fernando Pereira

------------------------------

Date: Thu 26 Sep 85 15:20:03-PDT
From: Fernando Pereira <PEREIRA%sri-candide@sri-marvin>
Subject: Destructive Assignment

Rick McGueer seems to be confused about the differences between
structure sharing and non-structure sharing methods of term
representation in Prolog.  In both methods, if we want to ``change''
an argument of a functor, say `a' by `b' in f(X,a,Y), what Prolog
copies is a chunk of c*n+k cells where c and k are implementation
-defined constants and n is the arity of the functor.  When replacing
a subterm at depth m of a term t, the number of cells copied is
sum(i=0..m-1,c*n←i+k) where n←i is the arity of the functor at depth
i. If the data structures are kept well balanced, it follows that
the cost  of updating a term of height m is O(log m). Of course,
this is just a particular case of the fact that a random-access
memory of size s can be simulated by assignment-free data structures
at an overhead of  O(log s).  In applications I have developed, this
overhead is a small cost to pay for ease of programming and the
space recovery efficiencies inherent in the stack allocation used
by any reasonable Prolog.

-- Fernando Pereira

PS: Cf. discussions on arrays in Prolog in earlier Digests, the
    paper ``Incorporating Mutable Arrays into Logic Programming''
    by Lars-Henrik Eriksson and Manny Rayner, and David H. D.
    Warren's logarithmic array code in the collection at
    SCORE:<PROLOG>

------------------------------

Date: 26 Sep 85 23:58:16 PDT
From: Deutsch.PA@Xerox.ARPA
Subject: Destructive Assignment

Again, I would like to suggest a different view of assignment
than Rick McGeer's last comment.  If A is embedded in B, C, and
D, the routines manipulating B, C, and D don't need to know
anything at all about the structure of A: all they need to know
is that they have to embed the modified A in their own structure
in the appropriate place, namely the place they got it from in
the first place to be able to tell it to modify itself.  Whether
a language has abstraction is independent of whether it has
assignment.  I consider Prolog's lack of abstraction and
modularity a much more serious problem for its use in systems
than its lack of destructive assignment.

------------------------------

Date: Fri, 27 Sep 85 10:32:44 PDT
From: (Rick McGeer) Mcgeer%ucbkim@Berkeley.EDU
Subject: Destructive Assignment, again...

Your argument is well-founded, but the principal argument for
destructive assignment is modularity, not efficiency.  Consider
two data structures, A and B, that share some common substructure
C.  Now, if C is modified, a new copy of C, C' is created.  In
most applications, one wants the data structures A and B to
contain the new substructure C'. Am I wrong in believing that A'
and B' must be created in the absence of destructive assignment,
since both A and B are now outdated?  And if I'm correct, isn't
it also the case that the parents of A and B must also be modified,
and so on recursively.

If program data structures were treelike rather than networklike
(in short, if no two data structures shared common substructure),
then copying is reasonable since one assumes that the only method
of accessing substructure is through the root of the tree, and
superstructures can be appropriately modified through the accessing
traversal.  If data structures are networklike, the problem is harder,
since the access to the substructure can come through one of a number
of parents, which means that the code which modifies the substructure
must be able to access all the parents of the substructure, and so on
through the data structure.

My own programs tend to shared structure a fair amount, which is why
I'm banging away at this problem.  It turns out that there's a way
to hack the moral equivalent of destructive assignment into Prolog
(using lists with unbound cdrs), but this requires time O(n) for a
set or access, where n is the number of times that this variable has
been set...

Anyway, if I'm confused about this, I'd appreciate any pointers that
you have to programming techniques to avoid the problem of shared
structure.

-- Rick.

------------------------------

Date: Fri, 27 Sep 85 15:39:51 PDT
From: (Rick McGeer) Mcgeer%ucbkim@Berkeley.EDU
Subject: Destructive Assignment, again

Yes, I thought of that, too.  I rejected it because it seemed
to defeat one of the central purposes of symbolic languages,
namely getting rid of explicit pointers.  Further, after examing
the Warren PLM I couldn't think of a good way to do this invisibly
without major surgery to the PLM.  Destructive assignment, however,
required only oen new instruction (rplacarg), and modifying the
trail to contain 2 longwords per entry rather than one.

-- Rick

------------------------------

Date: Fri, 27 Sep 85 13:44:33 PDT
From: Deutsch.pa@Xerox.ARPA
Subject: Destructive Assignment, again...

Thanks for your thoughtful reply.  As you point out, the natural
underlying structure of the data may be an unrestricted graph,
for which I don't know how to construct a purely functional
representation that models edges as pointers.  However, it's
easy to construct a functional representation that doesn't model
edges as pointers: divide the structure up into a set of nodes
represented by unique ids, and a separate dictionary that maps a
node id to the contents of the node it represents.  This
representation caters nicely to the problem of updating an
interior node and having the update visible from anywhere that
references it.  In effect, a level of indirection has been
introduced.  While the dictionary itself must be global to the
structure involved, it doesn't have to know anything about the
semantics of the individual nodes or edges, so logical modularity
is still preserved.  And the indirection can be covered with
syntactic sugar to eliminatemost of the burden on the programmer:
all that is necessary is that node references be represented as
pairs <containing structure, node identifier>, and that all node
updates return a new containing structure with appropriate
modifications.  All that remains is to make sure the garbage
collector knows it can remove any dictionary entries where the
id isn't accessible -- in effect, knows that the dictionaries are
being used to simulate a memory system.

In circumstances where the natural structure of the data is not an
unrestricted graph, pointers can be modelled as pointers, and
reconceptualizing the algorithms to use paths or indices rather
than direct references may yield more transparent programs.  (It's
been my experience that a lot of the bugs that arise in systems are
the result of implicit transmission of side-effects through sharing
of references to mutable structures.)

There have been a number of papers in the Lisp / Functional
Programming symposia on the subject of effective programming
without side effects. That would be the first place I would
look for references.

------------------------------

Date: 27 Sep 85 18:05:34 PDT
From: Deutsch.pa@Xerox.ARPA
Subject: Destructive Assignment, again...

Wait a minute.  Surely destructive assignment implies the
existence of "reference" as a concept that is visible to
the programmer?  And how is this different from "explicit
pointers"?  In a functional language, the concept of
"reference" disappears -- you can't tell the difference
between eq and equal (and in fact the implementor is free
to adopt strategies that may copy or share, ad lib.)  The
dodge I suggested simply makes references explicit in those
(and only those) places where their semantics are needed.

In Prolog, logical variables create a kind of sharing that
is pretty much just like a reference.  In this respect Prolog
feels quite different to me than a functional language.  On
the other hand, there may be a way to think of Prolog clauses
as functions from environments to environments that is more
like what I suggested.  Also, in principle the Prolog
implementor is free to use a mix of copying and sharing
strategies (e.g. Prolog on a multi-processor will probably have
to do some copying), as long as the equality constraints between
occurrences of a given variable are extended to include any
copies.

------------------------------

End of PROLOG Digest
********************

∂30-Sep-85  0221	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #19  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  02:21:09 PDT
Date: 30 Sep 85 0144-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #19
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 30 Sep 1985       Volume 1 : Issue 19

Today's Topics:

            Recent posting on Japan trip by Charles Watson
                      Parallel Lisp environments
                Simulators for hypercube architectures
                 Response to the First PARSYM Survey
            Electronic clearinghouse for technical reports

----------------------------------------------------------------------

Date: Mon, 23 Sep 85 10:34:58 pdt
From: eugene@AMES-NAS.ARPA (Eugene Miya)
Subject: Recent posting by Charles Watson

As a quick addition to the note by Charles Watson on his
recent visit to Japan:

We are trying to use on of the Ames gateways to UUCP and the ARPAnet
to set up an informal groups of people to monitor foreign language
technical information.  Our principal interest right now is Japanese
because as Charles found out, a lot is being done there.

We don't want to form a new bulletin board [there are too many now],
but we want to identify individuals capable of translating foreign
technical material, identify significant non-translated journals,
and then ask people to post significant papers as they emerge to speed
translation or interest.

The Usenet has a newsgroup, net.mag, where people post the TOCs of
significant new issues.  What I am proposing is posting such TOCs
of new non-English material to existing bulletin boards.  Where
interest exists, interested parties can get together and pay for
the cost of translation, thus identifying significant papers quickly.

If you are interested and can help translate some non-English
foreign material, please contact me.  We have interest from people
all over Europe, and a couple of people in Japan.

--eugene miya
  From the Rock of Ages Home of Retired Hackers
  eugene@ames-nas

------------------------------

Date: Tue, 27 Aug 85 15:37:18 EDT
From: Deepak Kumar <kumard%buffalo.csnet@csnet-relay.arpa>
Subject: Parallel LISP Environments

[Forwarded from AIList by Laws@SRI-AI.ARPA]

        We are planning to implement a parallel-environment
        simulator in LISP for trying out various control
        metaphors that require parallelism.

        Anyone already having any implementations or
        experiences in using such environments?

        We would appreciate any kind of responses that could
        help in highlighting various aspects of design as well
        as implementation.

        Thanx.

        Deepak.


UUCP   : {cmcl2,hao,harpo}!seismo!rochester!rocksvax!sunybcs!kumard
         ...{allegra,decvax,watmath}!sunybcs!kumard
CSNET  :  kumard@buffalo
ARPA   :  kumard%buffalo@csnet-relay
BITNET :  kumard@sunybcs

------------------------------

Date: 30 Aug 85 13:54:23 EDT (Fri)
From: duke@mitre.ARPA
Subject: Simulators for hypercube architectures

[Forwarded from AIList by Laws@SRI-AI.ARPA]

Lisa Sokol (sokol@mitre) and I are investigating the use of parallel
architectures for object-oriented simulations.  We have a large simulation
called the BEM (Battlefield Environment Model) which is written in the
simulation language ROSS, which runs on top of Franz LISP on a VAX.  We
would like to have a hypercube simulator which would help us investigate
strategies for distributing objects in a hypercube.  Our main problem seems
to be the integration of LISP code with the simulators.  Translating the
Lisp into C would be messy at best, even with a translator program.  Another
possibility would be to have Franz LISP running in another process on the
VAX and communicating with the simulator process.  That sounds pretty messy
also, particularly for analyzing the timing since the Lisp is running
outside of the simulator process.  For our purposes, it may be better to
write a Lisp simulator of a hypercube architecture.  Do you know of anyone
who already has such a simulator?  We expect that Lisp will be available on
some hypercube in the next year, and our simulator studies should allow us
to determine how to distribute the objects by that time.


Duke Briscoe
duke@mitre

------------------------------

Date: Thursday, 26 September 1985  16:24-PDT
From: zorn%ucbrenoir at Berkeley.EDU (Ben Zorn)
Re:   Response to the PARSYM survey... (a little late, I'm afraid)

[Thank you -- it's never too late.  For newcomers, the First PARSYM
Survey was devoted to goals for research in parallel computing.  The
results of the survey were distributed as V1 #11 of the PARSYM digest,
which is available from PARSYM-Request@SUMEX-AIM.ARPA.  -- BD]

The SPUR Project at Berkeley is interested in building a workstation
with 10-12 VLSI RISC-style processors specifically designed to run Lisp.
One phase of the project is to implement Common Lisp for SPUR and provide
primitives for multiprocessing.  Here is our response to the survey.

	Ben Zorn
	Jim Larus
	George Taylor
	Prof. Paul Hilfinger

 --------

1.  What, if any, is your target application for parallel computing?
    (Theoreticians aren't required to reply to this question.)

    [Symbolic computation including expert systems, rule-based systems,
    and algebraic manipulation systems.  Other applications that might
    benefit from LISP and multiprocessors.]

2.  Is performance improvement your main goal in exploring parallel
    computing?

    [Yes.  We are hoping to develop a programming style and a
    sophisticated compiler and runtime system that will permit programmers
    to utilize the multiple processors to speed up their applications.]

3.  (a) If so, what degree of performance improvement does your target
       	application require (in, say, orders of magnitude)?  What
        degree of performance improvement does your current approach
        promise?

    [Given that the SPUR workstation will have at most 10-12 processors,
     we expect a performance improvement of at most an order of magnitude.
     We would be happy with a 4 or 5 times speed-up for most applications.]

    (b) If not, what else do you expect to achieve through the use of
        parallelism?

	[Higher performance without increased chip complexity.]

------------------------------

Date: Fri, 27 Sep 85 13:38:21 cdt
From: Laurence Leff <leff%smu.csnet@CSNET-RELAY.ARPA>
Subject: Electronic clearinghouse for technical reports

I have volunteered to organize an electronic mechanism for the
distribution of technical report lists from Universities and R&D labs.
Some (and hopefully all) of the people producing technical reports
would send a copy of the list to me.  I would then send these to a
moderated group on USENET as well as a mailing list for those sites on
the INTERNET who do not get news (ARPANET, CSNET, etc.).

I need two things from you:
  1) if your organization prepares technical reports and sends them
      out to interested parties (perhaps for a fee), please arrange
      to have electronically readable copy of your lists sent
      to trlist%smu@csnet-relay.
  2) if people at your organization would like to receive lists
     of tech reports produced by universities and R&D labs, please
     provide me an electronic address to send them to (if you are not
     on USENET).  Send such administrative mail to trlist-request%smu@
     csnet-relay.

Some frequently asked questions:

1. What are the advantages of sending my lists to you?

a. Most of the people to whom you are sending printed lists will be
receiving this list, either through the INTERNET as a mailing list or
as a moderated news group on the USENET distributed bulletin board
system.  Thus you can save the postage and printing costs in mailing
these lists.  I would be happy to provide you with a list of
institutions receiving this list as a mailing list as well as those
institutions on USENET who would be receiving it that way.  You can
use this to prune the mailing list you use to send out printed copies
of your technical report lists.

b. Many people at the Universities are not aware of technical
report lists.  I have been sending out lists of AI tech reports
to the AIList, an electronic newsletter on AI, for some time.
Every time I do so, my electronic mailbox fills up with requests on
how to obtain the tech reports.  Many of these requests come from
the most prestigious AI organizations in the country.

c. Many companies, particularly those on the USENET, would not
otherwise be aware of your research.  There are hundreds of small
companies on USENET who have no other access to the wealth of
information represented by University and other tech reports.

2. What is a technical report?

Most universities and big company R&D labs publish reports about their
research.  Some are highly research oriented (like new results in
automata theory).  Others are manuals for their public domain software
or tutorials.  For example NASA/Ames published a tutorial on SETUID
programs under UNIX.  These lists are currently sent out by mail to
other schools and R&D labs.

Some of the technical reports will later get turned into journal
articles while other items will never be more formally published.
Thus looking at these lists would give you information on new research
results before they would appear in journals or would let you know of
material you would not otherwise be aware of.

3. What format should the tech report lists be in?

Please see to it that there is some info indicating how people can
order the tech reports (whether sending you a check to cover costs,
requests via electronic mail or the reports can be electronically
available for Arpanet FTP transfer).

If you are already producing the list in some format, feel free to use
that format.  If you are preparing the list just for this purpose, I
would prefer that you use the input format for bib/refer, a common
bibliography tool.  This way people can dump the lists into a file on
their machine and be able to do keyword searches.  Also bib/refer will
automatically include and format references in documents to be
formatted or typeset.  However, I would prefer the material in some
weird format than not to have it at all!

For those not familiar with bib/refer, here is a brief tutorial.  Each
report or other item should be a sequence of records which are not
separated by blank lines.  Each report should be separated by the
others by one or more blank lines.  Each report entry consists of a
label consisting of a % followed by a capital letter and then a space.
Then include the information.  If the information for a field (such as
an abstract) requires more than one line, just continue the field on a
new line with no initial space.

The labels needed for tech reports are:

%A Author's name  (this field should be repeated for each author).
%T Title of report
%R report number
%I issuer, this will be the name of your institution.   This may
be ommited if implied by the report number
%C City where published (not essential)
%D Date of publication
%X Abstract


Here is an example of some tech report listings in the appropriate
format:

%A D. Rozenshtein
%A J. Chomicki
%T Unifying the Use and Evolution of Database Systems: A Case Study in
PROLOG
%R LCSR-TR-68
%I Laboratory for Computer Science Research, Rutgers University
%K frame control

%A C. V. Srinivasan
%T CK-LOG, A Calculus for Knowledge Processing in Logic
%R DCS-TR-153
%I Laboratory for Computer Research, Rutgers University
%K MDS

4. I already have exchange agreements with other Universities.
How does this affect them?

The only change would be how the information on what technical reports
you have for them to request gets transferred.  Instead of them
receiving a piece of paper by U. S. Mail, they would look at the
appropriate notes group (if this is a USENET site) or at the item
received in the mail, request the reports they want and send the
request to you.  You would probably request that the free technical
report order came from a specific person or account in case some
student seeing the list decided to order the tech reports.  You should
do that with the printed lists anyway since at some schools, technical
report lists are frequently left around for graduate students and
faculty to look at and check the ones they want.  Any person could
send in the form themselves if they chose.

5. I need to charge for my tech reports to cover costs.

Fine.  Just include the prices for your reports next to each report (you can
use the %X field for that too).  At the beginning of the list you send me,
state where checks should be sent and to whom they should be made payable.

6. What about non-CS reports?

I am happy to handle reports for other departments.  If the volume of
non-CS reports becomes significant, I will split the list into tr-cs,
tr-math, tr-ee etc.  I would suspect that the majority of the people
receiving this list would be CS researchers since CS departments are
quick to join networks, etc.  However, some CS researchers (myself
included) are working in applications of computers and would like to
receive information in those areas as well.

7. I am already on USENET.  What should I do?

I anticipate a USENET moderated group in a time frame of one to two
weeks which will contain the same information as the technical report
lists.  If you indicate that you will get the information via USENET,
I will remove your name when the list is established.  If you want to
wait a week or two to see if the list comes up, that is OK too.  I can
send back copies of the TR Lists that get sent out in the first few
batches of the mailing.  I will also send out on the USENET group,
everything that got sent out in the mailing list so you won't miss
anything either way.

8. I am on Arpanet, BITNET, etc.

I can get to Arpanet sites through csnet-relay so there is no problem
there.  Otherwise, send me your address as best you know it.  I will
get through to you if at all possible.

------------------------------

End of PARSYM Digest
********************

∂30-Sep-85  0829	PATASHNIK@SU-SUSHI.ARPA 	Abstracts 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  08:29:37 PDT
Date: Mon 30 Sep 85 08:26:26-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Abstracts
To: aflb.all@SU-SUSHI.ARPA

Here are abstracts for the first two AFLB talks this quarter.

	****************************************************

3-Oct-85 - Umesh Vazirani (MSRI)

	Sampling a Population, with a Slightly-random Source

Consider the following question: Given a box containing red and blue
tickets, we wish to estimate the percentage of red ones.  A result
from elementary probability theory is that a constant number of
samples suffice, independent of the size of the population. This
simple but very surprising fact has obvious applications in gallup
polls and consumer surveys.  In computation, this fact forms the basis
of Monte-Carlo or Randomizing Algorithms.

The problem we consider is: How do we pick a random sample?  The
difficulty is that the available sources of randomness such as Zener
diodes, and geiger counters are imperfect. They do not output
unbiased, independent coin flips.  An imperfect source of randomness
can be modelled as an adversary source, the slightly-random source: An
adversary with infinite computing power, and complete knowledge of our
algorithm generates the successive bits of output of the source.  The
only constraint on the adversary is that each successive bit must be 1
with probability at least delta, and 0 with probability at least
delta, for any fixed constant 0 < delta < 1/2.

We prove that O(n logn) samples suffice to sample a population of size
2**n, using a slightly-random source to generate the sample points.
As a consequence of this sampling theorem we strengthen the SRp = Rp
result in two ways:

Theorem 1: SBPP = BPP.
Theorem 2: RTIME(T) is contained in SRTIME(O(T logT)).

Theorem 2 says that any probabilistic algorithm that runs in time T
can be converted into an algorithm that runs in time O(T logT), and is
guaranteed to work even if its source of random bits is a
slightly-random source instead of a fair coin.

***** Time and place: October 3, 12:30 pm in MJ352 (Bldg. 460) ******

10-Oct-85 - Eli Shamir (Hebrew University)

		New results about graph coloring

The chromatic number of a random graph with expected degree d(n) is
sharply concentrated around b*d(n)/logd(n).  There are provably good
heuristics to unravel the coloring of a dense graph with an unusually
small number of color classes.

***** Time and place: October 10, 12:30 pm in MJ352 (Bldg. 460) ******
-------

∂30-Sep-85  0935	RICHARDSON@SU-SCORE.ARPA 	Tuesday Lunch Series    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  09:35:45 PDT
Date: Mon 30 Sep 85 09:33:53-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA, bureaucrats@SU-SCORE.ARPA
Message-ID: <12147409604.12.RICHARDSON@SU-SCORE.ARPA>

Tomorrow, Oct. 1, at 12:15 in MJH Room 146 the Tuesday Lunch Series
continues with Dwain Fullerton on "Fund Raising".

Anne
-------

∂30-Sep-85  1359	DAVIES@SUMEX-AIM.ARPA 	No meeting this week  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  13:58:54 PDT
Date: Mon, 30 Sep 1985  13:59 PDT
Message-ID: <DAVIES.12147457986.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No meeting this week
cc:   Davies@SUMEX-AIM.ARPA

There will be no meeting of the Advanced Architectures Project this
week.  Both Ed and Penny are out of town, and no one expressed any
great interest in having a meeting.

There will be a meeting next week on Wednesday, Oct. 9 at 9:30 am.
The topic will be announced by this coming Friday.

	-- Byron

∂30-Sep-85  1417	NILSSON@SU-SCORE.ARPA 	Course Schedule Policy
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  14:16:52 PDT
Date: Mon 30 Sep 85 14:02:01-PDT
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Course Schedule Policy
To: ac@SU-SCORE.ARPA
cc: berg@SU-SCORE.ARPA
Message-ID: <12147458414.37.NILSSON@SU-SCORE.ARPA>

I have had some complaints that there is a general lack of consistency
between the CS course listings in "Courses and Degrees" and the final
quarterly time schedule.  There are also apparently several unapproved
changes in course times that occur near the beginning of a
quarter--after the time schedule comes out.  These changes are all very
difficult for our students who rely heavily on the information in
"Courses and Degrees" to plan their schedules.

We should remind ourselves of the department policy on course
scheduling--established at the Faculty Meeting on 10 January 1984:

"Once a year, the Curriculum Committee is charged with updating Courses
and Degrees.  Subsequently, each professor will receive a notice with
information about courses, quarters to be taught, and time slots.

Prior to final galley reading of Courses and Degrees, requests for changes,
additions, and deletions can be submitted to the Publications Coordinator
who will in turn refer them to the Curriculum Committee and to the Chairperson
for final approval.  Changes will be considered only if they do not cause
conflicts, and if appropriate classroom space is available.  Once the final
galley reading in the Editorial Office has taken place, no course can be
changed from one quarter to another.  Instructors must hold the first class
meeting at the prescribed time and place.  Upon completion of Courses and
Degrees, there is a commitment to teach the courses as stated.  No changes
will be accepted for quarterly time schedules, since these must reflect 
information contained in Courses and Degrees."

I assume that the faculty had good reasons for adopting this policy and
that those reasons still obtain.  Please let me know if there are
situations in which exceptions must occur.  Thanks,  -Nils
-------

∂30-Sep-85  1430	MAYR@SU-SCORE.ARPA 	topics for prob. theory course
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  14:30:15 PDT
Date: Mon 30 Sep 85 14:26:22-PDT
From: Ernst W. Mayr <MAYR@SU-SCORE.ARPA>
Subject: topics for prob. theory course
To: faculty@SU-SCORE.ARPA
Message-ID: <12147462849.29.MAYR@SU-SCORE.ARPA>

Brad Efron from the Statistics Dept. has offered to teach a course in
Probability Theory/Statistics (say at a level comparable to Stat116) which
could be particularly aimed at CS students, or topics we'd like our students
to know about. Right now, Stat116 is taught three times a year (fall, spring,
and summer), and this new course could replace one of these offerings.

I am trying to compile a list of topics in Probability Theory and Statistics
that are relevant to our CS students, and I am asking your help. Of course,
there are some basic things every such course must contain (random variables,
event spaces, conditional probabilities, the main distributions, basic facts
about tests, a.m.m.). What else would you like to see? Like:

- distributions of sums/max/functions of many r.v.'s
  and estimates for those
- Chernoff bounds
- more extensive treatment of binary r.v.'s
- construction of r.v.'s with certain distribution from others (computational
  methods)
- more treatment of dependent r.v.'s (whatever one can do here)

- ???

ernst
-------

∂30-Sep-85  1828	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  18:26:34 PDT
Received: from ames-vmsb.ARPA by SU-SCORE.ARPA with TCP; Mon 30 Sep 85 18:15:15-PDT
Date: 30 Sep 85 17:51:00 PST
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA


               ASSOCIATION FOR COMPUTING MACHINERY
                San Francisco Golden Gate Chapter
               "SIGBIG" Special Interest Committee
                 For Large High Speed Computers

 Meetings on  the first Wednesday of each month at 7:30 PM.   Speakers 
 who  can give insights to various aspects of  SUPERCOMPUTING are 
 featured each month.

 Next meeting:
     Wednesday, October 2,1985,  7:30 PM

	 A SIGBIG general business meeting will be held at
	 ZeroOne Systems,Inc.	SIGBIG goals will be discussed
	 and tasks for new volunteers will be discussed.  
	 	Please bring your Ideas.

     Location:  ZeroOne Systems, Inc. "01"
		2431 Mission College Blvd.
		Santa Clara, CA.

     Directions:  From Hwy 101, North onto Great America Parkway.
		  Right,East on Mission College Blvd. Left,North
		  at second light into parking.
 ---------------------------------------------------------------
 Tape-recordings  of  most of the previous  may  be obtained
 in exchange for a tape cassette or $5.00 by contacting: 
                Mary Fowler (415)261-4058 (rec)
                Supercomputing  #192, BOX 2787
                Alameda, CA. 94501-0787

 For information contact Mary Fowler, Chairperson (415) 839-6547
                     or  Mike Austin, Publ. Chair (415) 423-8446

------

∂30-Sep-85  2209	ACUFF@SUMEX-AIM.ARPA 	Explorer meets TCP
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 30 Sep 85  22:09:03 PDT
Date: Mon 30 Sep 85 22:08:17-PDT
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorer meets TCP
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12147546937.74.ACUFF@SUMEX-AIM.ARPA>

   Beta TCP is now running on the Explorer.  Unfortunately, TCP comes
with an unstable system (an kludged together mid-release development
system), so be sure to read the notes tacked to the wall beside the
console, or look in <Acuff>TI-TCP.DOC before using the Explorer.
These notes cover how to use the new system, as well as caveats and
known bugs, though I doubt it is complete.  Release 1.0 is still
available also.

	-- Rich
-------

∂01-Oct-85  0855	RICHARDSON@SU-SCORE.ARPA 	Tuesday Lunch Series    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  08:55:45 PDT
Date: Tue 1 Oct 85 08:52:31-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA, bureaucrats@SU-SCORE.ARPA
Message-ID: <12147664216.23.RICHARDSON@SU-SCORE.ARPA>

Don't forget lunch today at 12:15 in MJH 146!

Anne
-------

∂01-Oct-85  0903	JF@SU-SUSHI.ARPA 	Still need stanford speaker for 10/11/85 bats  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  09:03:02 PDT
Date: Tue 1 Oct 85 08:55:50-PDT
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: Still need stanford speaker for 10/11/85 bats
To: aflb.su@SU-SUSHI.ARPA
cc: ragde%ucbernie@UCB-VAX.ARPA

No one from Stanford has volunteered to speak at BATS in Berkeley on 10/11.
I'm still waiting...
-------

∂01-Oct-85  0922	@SU-SCORE.ARPA:SHORTLIFFE@SUMEX-AIM.ARPA 	Re: topics for prob. theory course    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  09:22:36 PDT
Received: from SUMEX-AIM.ARPA by SU-SCORE.ARPA with TCP; Tue 1 Oct 85 09:18:36-PDT
Date: Tue 1 Oct 85 09:21:59-PDT
From: Ted Shortliffe <Shortliffe@SUMEX-AIM.ARPA>
Subject: Re: topics for prob. theory course
To: MAYR@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
In-Reply-To: <12147462849.29.MAYR@SU-SCORE.ARPA>
Office: Room TC-135, Stanford Med Center; Phone: (415) 497-6979
Message-ID: <12147669580.39.SHORTLIFFE@SUMEX-AIM.ARPA>

	I am VERY interested to hear about this development.  We require a
course in probability (Stat 116 or EES 221) for the MS or PhD in Medical
Information Sciences, and there has been some dissatisfaction with the way
those courses relate both to the DA (decision analysis) portions of our
curriculum and to the CS requirements which form the bulk of the required
courses.  I sent out your message to our current students, including those
who have taken Stat 116 or EES 221, and asked them to make suggestions based
on their experiences in those courses.  Will forward you the responses that
I receive.  One of our better students, with a particular interest in
math and statistics (working on problems of evidential reasoning for expert
systems) has already replied and I'll forward his message to you right now.
	Please keep me informed on how this develops.  We will almost
certainly change our required curriculum to include this new course if it
is indeed implemented.
	Regards,
	   Ted Shortliffe

-------

∂01-Oct-85  1017	CLT  	Seminar on Logic and the Foundations of Mathematics   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA, su-bboards@SU-AI.ARPA 


Speaker: Prof. Shaughan Lavine, Dept. of Mathematics, Stanford

Title: Prewellordering and the generalized reduction property.

Time: Monday, Oct.7, 4:15-5:30 P.M.

Place: Room 383N (faculty lounge, 3d floor, Math. Bldg.).

                                     S. Feferman


PS

If you see this message on one of the bulletin boards and
would like to be on the regular mailing list for logic
and theory of computation type announcements 
or
if you receive this message in your electronic mail box and 
would like to be removed from the mailing list
then
send a message to CLT@SU-AI.  

∂01-Oct-85  1039	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar on Logic and the Foundations of Mathematics    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  10:30:16 PDT
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 1 Oct 85 10:29:26-PDT
Date: 01 Oct 85  1017 PDT
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Seminar on Logic and the Foundations of Mathematics   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA, su-bboards@SU-AI.ARPA 



Speaker: Prof. Shaughan Lavine, Dept. of Mathematics, Stanford

Title: Prewellordering and the generalized reduction property.

Time: Monday, Oct.7, 4:15-5:30 P.M.

Place: Room 383N (faculty lounge, 3d floor, Math. Bldg.).

                                     S. Feferman


PS

If you see this message on one of the bulletin boards and
would like to be on the regular mailing list for logic
and theory of computation type announcements 
or
if you receive this message in your electronic mail box and 
would like to be removed from the mailing list
then
send a message to CLT@SU-AI.  

∂01-Oct-85  1150	ullman@su-aimvax.arpa 	meeting tomorrow 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  11:50:24 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 1 Oct 85 11:47:21 pdt
Date: Tue, 1 Oct 85 11:47:21 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting tomorrow
To: nail@diablo

Let's meet at 11AM.
Therer is no agenda, but I have some funny stories
to tell, and perhaps other topics will come up.

∂01-Oct-85  1227	avg@su-aimvax.arpa 	Re:  meeting tomorrow    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  12:23:12 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 1 Oct 85 12:20:19 pdt
Date: Tue, 1 Oct 85 12:20:19 pdt
From: Allen VanGelder <avg@diablo>
Subject: Re:  meeting tomorrow
To: nail@diablo, ullman@diablo

I can continue on NC/P and present the non-trivial NC example and or the
P-complete example.

∂01-Oct-85  1340	HAUNGA@SUMEX-AIM.ARPA 	Title for Siglunch - Friday Oct 4    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  13:40:15 PDT
Date: Tue 1 Oct 85 13:37:33-PDT
From: Ana Haunga <HAUNGA@SUMEX-AIM.ARPA>
Subject: Title for Siglunch - Friday Oct 4
To: siglunch@SUMEX-AIM.ARPA
Message-ID: <12147716104.31.HAUNGA@SUMEX-AIM.ARPA>

This week's SIGLunch talk is "Introspection" by Mike Genesereth.
The abstract will follow.  SIGLunch is held in the Chemistry Gazebo
at 12:05-1:00.



-----a
-------

∂01-Oct-85  1525	LIBRARY@SU-SCORE.ARPA 	Socrates: SELECT 9 for Technical Reports  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  15:25:07 PDT
Date: Tue 1 Oct 85 15:21:26-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates: SELECT 9 for Technical Reports
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12147735017.15.LIBRARY@SU-SCORE.ARPA>

Now that the new version of Socrates is up, I hope you will take advantage
of the Technical Reports file which has been added.  The Technical Reports
file contains records from all of the science libraries.  However a majority
of the records currently in the file are records from the technical reports
collection in the Math/CS Library.  Our printed index in the library is 
no longer up to date for our collection and the file on Socrates should
be used to determine what we have.  New reports are being added to the
Socrates file directly and printed indexes will not be generated.  However
we are continuing to announce our New Reports List online through the
Bulletin Board.

The Technical Reports file is a separate file on Socrates and is not 
included in the Catalog Headings file.  To call up the Technical Reports
file you can say SELECT Technical Reports or SELECT 9.  Once you are
in the file, you can search the file as you would any other file on
Socrates either in the GUIDED or COMMANS modes.  You may also BROWSE
authors, subjects etc.

If you have any comments or questions about the Technical Reports file or
other Socrates issues, I would be interested in hearing them.  You may
make suggestions directly to the group developing Socrates by using
the SUGGEST command.  Anyone needing a Socrates account can come by the
library to file out the form and pick up an account number.

Harry Llull
-------

∂01-Oct-85  1600	RICHARDSON@SU-SCORE.ARPA 	Pre-Colloquium Cookie Time   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  16:00:31 PDT
Date: Tue 1 Oct 85 15:47:30-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Pre-Colloquium Cookie Time
To: csd@SU-SCORE.ARPA
Message-ID: <12147739761.11.RICHARDSON@SU-SCORE.ARPA>

CS 500 Pre-Colloquium Cookie Time is now in progress in the student lounge
of Margaret Jacks Hall.

-------

∂01-Oct-85  1636	MARJORIE@SU-CSLI.ARPA 	key    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  16:36:38 PDT
Date: Tue 1 Oct 85 16:34:12-PDT
From: Marjorie Maxwell <MARJORIE@SU-CSLI.ARPA>
Subject: key
To: folks@SU-CSLI.ARPA

Has anyone lost a key - if so check with me at G-4.
Marj
-------

∂01-Oct-85  1723	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Monday Planlunch Reminder!
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  17:22:47 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 1 Oct 85 17:15:45-PDT
Date: Sun 29 Sep 85 19:05:29-PDT
From: LANSKY@SRI-AI.ARPA
Subject: Monday Planlunch Reminder!
To: aic-associates@SRI-AI.ARPA, folks@SU-CSLI.ARPA, mugs@SUMEX-AIM.ARPA,
    val@SU-AI.ARPA, phayes@SRI-KL.ARPA, carnese@SRI-KL.ARPA,
    alpert@SU-SUSHI.ARPA, frayman@SUMEX-AIM.ARPA, dmrussell@XEROX.ARPA,
    araya@XEROX.ARPA, frayman@XEROX.ARPA, suchman@XEROX.ARPA, weld@XEROX.ARPA,
    mittal@XEROX.ARPA, dekleer@XEROX.ARPA, trigg@XEROX.ARPA

	          PROCESSES, SIMULTANEITY AND CAUSALITY

        	          Michael P.  Georgeff
	              Artificial Intelligence Center
        	           SRI International

  	            11:00 AM, MONDAY, September 30
       SRI International, Building E, Room EJ228 (new conference room)


The notion of process is essential for reasoning about the behavior of
multiple agents or single agents in dynamic worlds.  In this talk, we
show why reasoning about process is so important, and contrast this
with other approaches in AI which are primarily based on the allowable
behaviors of agents.  An algebra of processes based on events is given.
We then show how events can be represented as changes of world state, 
and how state properties can be inferred from the model. Interestingly, 
no STRIPS-like assumption is involved in the definition of events, thus 
allowing a proper model-theoretic semantics.  One of the most important 
features of the model is a hiding operation.  This provides an abstraction
capability that can be used to avoid the combinatorial explosion
typical of other AI approaches.  Finally, we introduce a notion of
causality between events and processes.  This, together with the
notion of simultaneous actions and hiding operations, allows us to
avoid most of the problems associated with the frame problem. 

-------

∂01-Oct-85  1726	POLLARD@SU-CSLI.ARPA 	new entity   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Oct 85  17:26:10 PDT
Date: Tue 1 Oct 85 17:23:12-PDT
From: Carl Pollard <POLLARD@SU-CSLI.ARPA>
Subject: new entity
To: folks@SU-CSLI.ARPA

Lucretia Claire Pollard was born at 1:39 Saturday afternoon at Stanford
Hospital. 9 pounds 2 ounces, 21-1/2 inches long, and doing fine. Parents
& big sister doiing fine too (but our recollections of a night's sleep
grow dimmer by the hour).
-------

∂02-Oct-85  0920	JOHN@SU-CSLI.ARPA 	Wheels for Wheels    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  09:20:02 PDT
Date: Wed 2 Oct 85 09:17:16-PDT
From: John Perry <JOHN@SU-CSLI.ARPA>
Subject: Wheels for Wheels
To: Folks@SU-CSLI.ARPA

Thanks to the hard work of Jamie Marks and David Brown, we have solved
our transportation needs to and from the Inner Quad for this era.

There are now 10 beautifully painted bicyles -- see if you can pick
them out -- available for trips between CSLI and the Philosophy/
Linguistics corner.  The following instant traditions govern the use
of these bicyles:

1.  Any CSLI-ite can get a key that will work on the locks of all of
    these bikes from Bach-Hong.

2.  Anyone may ride any CSLI bike parked at Ventura to the Philosophy/
    Linguistics corner with no responsibility to return it.

3.  Vice versa.

4.  One can ride a CSLI bike to some other part on Campus but has the
    responsibility of returning it to one of the above areas the same 
    day.

5.  Anyone who takes a CSLI bike off Campus is behaving outrageously 
    and other CSLI-ites who see this happening should throw things.

Corollary, you cannot be sure when you ride a CSLI bike to Ventura or
to the Philosophy/Linguistics corner for some event, that it will be
waiting for you after the event, i.e., plan accordingly.

After we have had some experience with this we will consider changes
in the rules, if necessary.  Let us know how it is working.

Report repair problems to Tom Yamarone or Jamie Marks.

We are negotiating for 10 Alfa Romeos to be painted similarly for
transportation between Ventura and SRI and PARC.


					 John Perry
-------

∂02-Oct-85  1030	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ucbernie.Berkeley.EDU    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  10:30:40 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 2 Oct 85 10:27:21-PDT
Received: from UCB-VAX.ARPA by SU-SCORE.ARPA with TCP; Wed 2 Oct 85 10:25:10-PDT
Received: by UCB-VAX.ARPA (5.28/5.11)
	id AA18296; Wed, 2 Oct 85 08:17:05 PDT
Received: by ucbernie.ARPA (4.24/5.11)
	id AA19894; Wed, 2 Oct 85 08:17:00 pdt
Date: Wed, 2 Oct 85 08:17:00 pdt
From: karp@ucbernie.Berkeley.EDU (Richard Karp)
Message-Id: <8510021517.AA19894@ucbernie.ARPA>
To: aflb.all@su-score.arpa

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley  (415)642-0143

COMPUTATIONAL COMPLEXITY SEMINARS

Wednesday, October 2   2:00  MSRI Lecture Hall
Lenore Blum
A Tutorial on Karmarkar's Algorithms

MSRI-EVANS WEDNESDAY LECTURE
Wednesday, October 2   4:15  60 Evans
Juris Hartmanis
On The Structure of Feasible Computations

Thursday, October 3    4:00  MSRI Lecture Hall
Neil Immerman
Expressibility and First-Order Reductions

Tuesday, October 8     2:00   MSRI Lecture Hall
Helmut Alt
Simulation of Idealized Models of Parallel Computers on (More) Realistic Ones

Tuesday, October 8     4:00   MSRI Lecture Hall
Eric Kostlan
Complexity Theory of Numerical Linear Algebra

Week of October 14-18
Workshop on Computation in Algebra and Number Theory
Schedule to be announced.

∂02-Oct-85  1248	admin@ucbcogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 8 (Terry Winograd, Stanford) 
Received: from UCB-VAX.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  12:48:36 PDT
Received: by UCB-VAX.ARPA (5.28/5.11)
	id AA23805; Wed, 2 Oct 85 12:45:01 PDT
Received: by ucbcogsci.ARPA (5.5/5.7)
	id AA22854; Wed, 2 Oct 85 12:46:31 PDT
Date: Wed, 2 Oct 85 12:46:31 PDT
From: admin@ucbcogsci.Berkeley.EDU (Cognitive Science Program)
Message-Id: <8510021946.AA22854@ucbcogsci.ARPA>
To: cogsci-friends@ucbcogsci.Berkeley.EDU
Subject: UCB Cognitive Science Seminar--Oct. 8 (Terry Winograd, Stanford)

                     BERKELEY COGNITIVE SCIENCE PROGRAM
                                 Fall 1985
                   Cognitive Science Seminar -- IDS 237A

       TIME:       Tuesday, October 8, 11:00 - 12:30
       PLACE       240 Bechtel Engineering Center
       DISCUSSION: 12:30 - 1:30 in 200 Building T-4

       SPEAKER:    Terry Winograd, Computer  Science,  Stanford University

       TITLE:     "What Can Cognitive Science Tell Us About Computers?"

       Much work in cognitive science rests on  the  assumption  that
       there is a common form of "information processing" that under-
       lies human thought and language and that also  corresponds  to
       the  ways we can program digital computers.  The theory should
       then be valid both  for  explaining  the  functioning  of  the
       machines  (at whatever level of "intelligence") and for under-
       standing how they can be integrated into human situations  and
       activities.

       I will argue that theories like  those  of  current  cognitive
       science  are  based  on  a "rationalistic" tradition, which is
       appropriate for describing the mechanics of machine operation,
       but  is  inadequate for understanding human cognitive activity
       and misleading as a guide to the  design  and  application  of
       computer  technology.   The  emphasis  will  be  on looking at
       alternatives to this tradition, as a starting point for under-
       standing what computers really can do.
       ---------------------------------------------------------------
       UPCOMING TALKS
       Oct 15:     Ron Kaplan, Xerox PARC
       Oct 22:     Lotfi Zadeh, Computer Science, UCB
       Oct 29:     Mardi Horowitz, Psychiatry, UCSF
       Nov 5:      Edward Zalta, CSLI, Stanford
       Nov 12:     Robert Wilensky, Computer Science, UCB
       Nov 19:     Richard Alterman, Computer Science, UCB
       Nov 26:     Eve Clark, Linguistics, Stanford
       Dec 3:      Bernard Baars, Langley Porter, UCSF
                                 * * * * *
       ELSEWHERE ON CAMPUS
       Edward De Avila, of Linguametrics, Inc., will be  speaking  on
       "The  Status of Language Minority Students in the U.S.: Scho-
       lastic Performance in Math and Science" on Monday, October  7,
       1985, at 4:10 p.m. in 2515 Tolman Hall, UCB.

       Boris Gasparov, of the UCB Slavic  Languages  and  Literatures
       Dept.,  will  be speaking on "Stylistic `Shifters' in Russian"
       on Tuesday, Oct. 8, 1985, at 8:00 p.m.  in  the  Tilden  Room,
       Student Union Bldg., UCB.

       William Cole, Cognitive Science, will be speaking on  "Medical
       Cognitive  Graphics" on Friday, October 11, 1985, at 4:00 p.m.
       in the Beach Room, 3105 Tolman Hall, UCB.
       --------------------------------------------------------------

∂02-Oct-85  1248	@SU-CSLI.ARPA:admin@ucbcogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 8 (Terry Winograd, Stanford)  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  12:48:39 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Wed 2 Oct 85 12:45:30-PDT
Received: by UCB-VAX.ARPA (5.28/5.11)
	id AA23805; Wed, 2 Oct 85 12:45:01 PDT
Received: by ucbcogsci.ARPA (5.5/5.7)
	id AA22854; Wed, 2 Oct 85 12:46:31 PDT
Date: Wed, 2 Oct 85 12:46:31 PDT
From: admin@ucbcogsci.Berkeley.EDU (Cognitive Science Program)
Message-Id: <8510021946.AA22854@ucbcogsci.ARPA>
To: cogsci-friends@ucbcogsci.Berkeley.EDU
Subject: UCB Cognitive Science Seminar--Oct. 8 (Terry Winograd, Stanford)

                     BERKELEY COGNITIVE SCIENCE PROGRAM
                                 Fall 1985
                   Cognitive Science Seminar -- IDS 237A

       TIME:       Tuesday, October 8, 11:00 - 12:30
       PLACE       240 Bechtel Engineering Center
       DISCUSSION: 12:30 - 1:30 in 200 Building T-4

       SPEAKER:    Terry Winograd, Computer  Science,  Stanford University

       TITLE:     "What Can Cognitive Science Tell Us About Computers?"

       Much work in cognitive science rests on  the  assumption  that
       there is a common form of "information processing" that under-
       lies human thought and language and that also  corresponds  to
       the  ways we can program digital computers.  The theory should
       then be valid both  for  explaining  the  functioning  of  the
       machines  (at whatever level of "intelligence") and for under-
       standing how they can be integrated into human situations  and
       activities.

       I will argue that theories like  those  of  current  cognitive
       science  are  based  on  a "rationalistic" tradition, which is
       appropriate for describing the mechanics of machine operation,
       but  is  inadequate for understanding human cognitive activity
       and misleading as a guide to the  design  and  application  of
       computer  technology.   The  emphasis  will  be  on looking at
       alternatives to this tradition, as a starting point for under-
       standing what computers really can do.
       ---------------------------------------------------------------
       UPCOMING TALKS
       Oct 15:     Ron Kaplan, Xerox PARC
       Oct 22:     Lotfi Zadeh, Computer Science, UCB
       Oct 29:     Mardi Horowitz, Psychiatry, UCSF
       Nov 5:      Edward Zalta, CSLI, Stanford
       Nov 12:     Robert Wilensky, Computer Science, UCB
       Nov 19:     Richard Alterman, Computer Science, UCB
       Nov 26:     Eve Clark, Linguistics, Stanford
       Dec 3:      Bernard Baars, Langley Porter, UCSF
                                 * * * * *
       ELSEWHERE ON CAMPUS
       Edward De Avila, of Linguametrics, Inc., will be  speaking  on
       "The  Status of Language Minority Students in the U.S.: Scho-
       lastic Performance in Math and Science" on Monday, October  7,
       1985, at 4:10 p.m. in 2515 Tolman Hall, UCB.

       Boris Gasparov, of the UCB Slavic  Languages  and  Literatures
       Dept.,  will  be speaking on "Stylistic `Shifters' in Russian"
       on Tuesday, Oct. 8, 1985, at 8:00 p.m.  in  the  Tilden  Room,
       Student Union Bldg., UCB.

       William Cole, Cognitive Science, will be speaking on  "Medical
       Cognitive  Graphics" on Friday, October 11, 1985, at 4:00 p.m.
       in the Beach Room, 3105 Tolman Hall, UCB.
       --------------------------------------------------------------

∂02-Oct-85  1305	CHEADLE@SU-SCORE.ARPA 	Hopeful Ph.D. Grads   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  13:05:43 PDT
Date: Wed 2 Oct 85 13:00:45-PDT
From: Victoria Cheadle <CHEADLE@SU-SCORE.ARPA>
Subject: Hopeful Ph.D. Grads
To: faculty@SU-SCORE.ARPA
Office: Margaret Jacks 258, 497-1519
Message-ID: <12147971551.26.CHEADLE@SU-SCORE.ARPA>


Listed below are the names of the Ph.D. students in the Computer Science
Department who plan to graduate sometime in the next year.
@begin(verbatim)
Abadi, Martin
Anderson, Richard
Asente, Paul
Beigel, Richard
Celoni SJ, James
Demetrescu, Stefan
Foulser, David
Feigenbaum, Joan
Fraley, Christina
Kenniston, Michael
Lowry, Michael
Mackinlay, Jock
Mitchell, Chad
Mogul, Jeffrey
Moses, Yoram
Russell, Stuart
Spreitzer, Michael
Tang, Wei Pai
Theimer, Marvin
Treitel, Richard
VanGelder, Allen
Weening, Joseph
Winslett, Marianne
@end(verbatim)

I thought you find this information useful.

Victoria


-------

∂02-Oct-85  1323	LIBRARY@SU-SCORE.ARPA 	Socrates:  Display to account--SUSHI problem   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  13:23:03 PDT
Date: Wed 2 Oct 85 13:19:14-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates:  Display to account--SUSHI problem
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12147974915.33.LIBRARY@SU-SCORE.ARPA>

John Lamping has brought to my attention that Display To Account on Socrates
does not work.  He has send a message to the Socrates people about this.
Display To Account does work with SCORE and I assume other computers.
Are users of Socrates on other computers having problems with the Display
To Account command?

Harry Llull
-------

∂02-Oct-85  1352	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  13:52:36 PDT
Date: Wed 2 Oct 85 13:49:13-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12147980373.33.LIBRARY@SU-SCORE.ARPA>

Ada For Specification: Possibilities And Limitations. ed. by Goldsack.
Cambridge Univ. Press. QA76.73.A35A284 1985

Combinatorics For Computer Science. by Gill Williamson. Computer Science
Press. QA164.W55 1985

The Second Scandinavian Conference On Image Analysis. Finland 1981.
TA1632.S27 1981

Surveys In Combinatorics 1985 Invited Papers for the 10th British
Combinatorial Conf.  ed. by Ian Anderson. Cambridge Univ. Press.
QA164.B74 1985.

TEX For Scientific Documentation. Proceedings of the first European Conf.
1985 Italy.  Z253.4T47E97 1985

Progress In Artificial Intelligence. ed. by L. Steels and J.A. Campbell.
Ellis Horwood Series AI.  Q335.P76 1985

Introduction To Computer Science Using The Turing Programming Language.
by R.C. Hold and J.N.P. Hume.  Reston Publ. QA76.H623 1984.

Dynamical Systems And Cellular Automata. ed. by J. Demongeot, E. Goles,
and M. Tchuente.  Academic Press.  QA267.5C45D96 1985

Numerical Methods For Scientific And Engineering Computation.  by M.K.Jain
S.R. Klyengar and R.K Jain. Halsted Press Book.  QA297.J28 1985.

Programming With APSE Software Tools. By Roy Freedmand. Petrocelli Books.
QA76.73A35F74.

The Analysis and Design Of Computer-Based Information Systems. by Joan
Nordbotten.  Univ. of Bergen.  QA76.9.S88N67 1985.

Application Systems in A.P.L.; How to build them right.  by Gibson, Levine,
Metzger. I.P. Sharp Associates. LTD.  QA76.73.A27G53 1985.

Writing and Analyzing Effective Computer System Documentation. by Ann
Stuart. Holt, Rinehart and Winston.  QA76.9.D6S78 1984.

Forth; Tools and Applications by Gary Feierbach and Paul Thomas.  
QA76.73.F24F45 1985.

The Biology Of Computer Life; Survival, Emotion and Free Will. Geoff
Simons, National Computing Centre, England.  Birkhauser. QA76.9.P75S57
1985.

Back To Basic; The History, Corruption, and Future of the Language.
by Kemeny and Kurtz. QA76.73.B3K44 1985

H. Llull
-------

∂02-Oct-85  1410	LIBRARY@SU-SCORE.ARPA 	Math/CS Library: New Reference Books--Guide to CS literature, Industrial Robotics Handbook, Computer-Readable Databas
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  14:10:21 PDT
Date: Wed 2 Oct 85 14:06:26-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library: New Reference Books--Guide to CS literature, Industrial Robotics Handbook, Computer-Readable Databases, German/English Dict.
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12147983508.33.LIBRARY@SU-SCORE.ARPA>

Computer Information Directory.  compiled and ed. by Darlen Myers 
Hilebrandt.  Revised and expanded. 1985   Z5640.H54 1985
   Excellent up-to-date guide to information resouces in Computer Science.

Handbook Of Industrial Robotics. by ed. Shimon Y. Nof.  with a forward by
Isaac Asimov. 1985. 1358 pages.  TS191.8H36 1985.

Data Systems Dictionary. English/German, German/English.  QA76.15.B74 1985.

Computer Readable Databases; A directory and data sourcebooks.  two volumes
vol. 1 Business, Law, Humanities, Socieal Sciences  Vol. 2 Science, 
Technology and Medicine.  Martha Williams  Editor-in-chief.  
Z699.22C66 1985

H. Llull
-------

∂02-Oct-85  1524	STUCKY@SU-CSLI.ARPA 	Bibliography Files 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  15:24:02 PDT
Date: Wed 2 Oct 85 15:20:34-PDT
From: Susan Stucky <Stucky@SU-CSLI.ARPA>
Subject: Bibliography Files
To: folks@SU-CSLI.ARPA

Dear All, 

A number of folks met yesterday to talk about the feasibility of setting
up bibliography files that would be accessible to everyone for
formatting in Latex or Scribe.  The idea is to keep files in a public directory
that can be used (though each subject file is maintained by a "Keeper
who is interested enough to enter materials sent to him/her and check
for consistency.)  We decided that such files would be feasible and are
presently working out the details.  

The following files and keepers are in the works:

Pragmatics: Dietmar Zaefferer
Phil. of Lg./Logic/Mathematics: Etchemendy, Menzel
Semantics: Link
Syntax: Rooth (and Hans if we can get him interested)

These are flexible (we decided on subject files rather than big ones
even though it may be that Latex cannot access more than one.)  If you
are interested in maintaining a file, let me know.  

If you are interested in getting the mail relevant to this endeavor,
send a message to Requests. This will be the last public message until
the system is up and running. If you are interested in the mail that has
been sent already or if yu have further questions, send a message to Stucky. 

-Susan
-------

∂02-Oct-85  1713	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #20  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  17:13:15 PDT
Date:  2 Oct 85 1656-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #20
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 2 Oct 1985      Volume 1 : Issue 20

Today's Topic:

   Second PARSYM Survey: Data Abstractions for Parallel Programming

----------------------------------------------------------------------

Date: Wed, 2 Oct 1985  16:50 PDT
From: DAVIES@SUMEX-AIM.ARPA
Subject: Second PARSYM Survey


              Data Abstractions for Parallel Programming


In this Second PARSYM Survey, I would like to find out what data
abstractions people find useful or are now developing to support
parallel computation.  From what I've seen, most published work is
devoted to control structures and communication rather than to data
structures.  To illustrate what I mean by "data abstractions for
parallel programming", I'll use some examples from multiprocessing
Lisp implementations.

Speaking in dichotomies, there are two approaches to introducing
parallelism to Lisp.  The first -- the control structure approach --
is exemplified by QLISP's QLET and Multilisp's PCALL: parallelism is
invoked through the lambda binding or function calling control
constructs.  QLET computes the values of LET parameters in parallel;
PCALL computes the arguments to a function in parallel.  (Both QLISP
-- formerly QLAMBDA -- and Multilisp are described in the Proceedings
of the 1984 Lisp and Functional Programming Conference.)

The second approach is what I call the data structure approach, and is
exemplified by the QLAMBDA construct in QLISP or the FUTURE construct
in Multilisp or the XECTOR on the Connection Machine (see Danny
Hillis's excellent new book "The Connection Machine" from MIT Press).
These constructs allow the programmer to deal with processes as first
class objects, and hence build structures of processes.  Gabriel and
McCarthy created pipelines in QLISP by stringing together QLAMBDA
processes with the output of one feeding the input of the next.  The
pipeline object so created is also a first class object which can be
passed around to provide pipeline processing whereever needed.  A
XECTOR is a generalized vector.  It can be used to represent
structures of data as we normally think of them or structures of
processes (or rather a combination of the two since the Connection
Machine closely identifies data and processes).

I don't want to use the questionnaire to exhaust all possible answers
to itself, but I will point out one more published example of a data
structure -- of quite a different type -- specifically created for
parallel computing.  This is the FRONS of Friedman and Wise's IAP
system (described in Filman and Friedman's book Coordinated
Computing).  A FRONS is a list the order of whose elements is
determined by the order in which they were computed.  It's roughly a
stream which freezes as the objects are created: the identity of the
CAR of a FRONS is nondeterministic until it is accessed, but remains
the same forever once it is accessed.


Please fill in the following questionnaire and return it to
PARSYM-Survey@SUMEX-AIM.ARPA.  As with the first survey, I plan to
distribute to PARSYM a summary of the replies as well as the text of
the replies themselves.  If you would rather not have your reply
distributed in full, please let me know.  Replies are particularly
welcomed from those whose systems I've mentioned: I'd appreciate any
clarifications.  Many thanks to Vineet Singh for help in formulating
the questionnaire.


               Second PARSYM Survey: Data Abstractions

              (Replies to PARSYM-Survey@SUMEX-AIM.ARPA)

1.  Are you using or have you developed data structures or
    abstractions specialized for parallel computing?  Please
    describe.

2.  In particular, do you use specialized data abstractions to
    represent configurations of multiple processes?  Please
    describe.

3.  Do you use specialized data structures, a la FRONS, for
    representing non-deterministic collections of data?  Please
    describe.

4.  What are some useful references on the topic of data abstraction
    in parallel computing (including your own)?

5.  Other comments welcome.

------------------------------

End of PARSYM Digest
********************

∂02-Oct-85  1726	EMMA@SU-CSLI.ARPA 	addendum to the newsletter
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  17:26:06 PDT
Date: Wed 2 Oct 85 17:25:28-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: addendum to the newsletter
To: friends@SU-CSLI.ARPA
Tel:  497-3479

Re: the tentative Fall calendar

  Each project is responsible for ONE colloquium sometime during the
two to three week period listed.

-Emma Pease
-------

∂02-Oct-85  1744	EMMA@SU-CSLI.ARPA 	Newsletter October 3, No. 48   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Oct 85  17:43:59 PDT
Date: Wed 2 Oct 85 16:55:24-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter October 3, No. 48
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
October 3, 1985                 Stanford                       Vol. 2, No. 48
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, October 3, 1985

   12 noon		TINLunch
     Ventura Hall       ``Idealized Cognitive Models'' and ``Metonymic Models''
     Conference Room    Sections 4, 5 of ``Women, Fire, and Dangerous Things''
			by George Lakoff
			Discussion led by Douglas Edwards

   2:15 p.m.		CSLI Seminar
     Redwood Hall	``Notes from the STASS Underground''
     Room G-19		David Israel, CSLI and SRI

   3:30 p.m.		Tea
     Ventura Hall		
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *NEXT* THURSDAY, October 10, 1985

   12 noon		TINLunch
     Ventura Hall       ``Artificial Intelligence Meets Natural Stupidity''
     Conference Room    by Drew McDermott
			Discussion led by Roland Hausser, U. of Munich

   2:15 p.m.		CSLI Seminar
     Redwood Hall	``Ontology and Intensionality''
     Room G-19		Edward Zalta, CSLI
			Discussion led by John Perry
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←
                  TENTATIVE FALL SCHEDULE FOR THURSDAYS

   THURSDAY SEMINARS

   Date                    Person or Group responsible

   10-3			Situation Theory and Situation Semantics
   10-10		Zalta
   10-17		Sells
   10-24		Discourse, Intention and Action
   10-31		Foundations of Document Preparation
   11-7			Phonology and Phonetics
   11-14		Finite State Morphology
   11-21		Computational Models of Spoken Language 
   12-5			Winograd
!
Page 2                     CSLI Newsletter                    October 3, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
   THURSDAY COLLOQUIA

   10-3 to 10-17:	Situation Theory and Situation Semantics
   10-24 to 11-7:	Discourse, Intention and Action 
   11-14 to 11-20:	Phonology & Phonetics, Finite State Morphology, &
                        Computational Models of Spoken Language 
   11/21		Joe Traub, CS Dept., Columbia
                              ←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S SEMINAR
                     ``Ontology and Intensionality''

      The foundations of semantics require more than just a theory of
   properties, relations, and propositions.  Such theories do show that
   logically equivalent relations and propositions are not necessarily
   identical, but they do not provide us with an explanation of modality
   and tense (for which we need something like worlds and times), nor
   with an explanation of the truth conditions, entailments, and
   substitutivity failures involving codesignative names and descriptions
   of important varieties of intensional sentences (for which we need
   something like intentional objects).  The theory which I have been
   developing has logical axioms which generate properties, relations,
   and propositions, and proper axioms which generate abstract
   individuals, some of which have just the features worlds have and some
   of which can help us explain intensionality by serving as intentional
   objects.  In the seminar, I'll show how to extend the theory to define
   times and account for the many similarities between worlds and times.
   Then I'll show that, given this ontology, the traditional
   understanding of intensionality must be revised and that certain
   classic puzzles involving modality and descriptions have a simple
   solution.					--Ed Zalta
                              ←←←←←←←←←←←←
                               LOGIC LUNCH

      On Mondays there will be an informal brown bag logic lunch in the
   Philosophy Lounge, building 90, from 12 to 1, starting October 7.  If
   you are interested in logic, please come any time.  Send questions to
   Jon Barwise (barwise@su-csli).
                              ←←←←←←←←←←←←
                              LOGIC SEMINAR

      The Logic Seminar will resume October 7 in the mathematics seminar
   room.  It will meet every Monday at 4:15.  Contact Sol Feferman
   (SF@su-csli) for details.  Information on the first seminar follows.

       ``Prewellordering and the Generalized Reduction Property''
          Prof. Shaughan Lavine, Dept. of Mathematics, Stanford
                      Monday, Oct. 7, 4:15-5:30 P.M.
           Room 383N (faculty lounge, 3d floor, Math. Bldg.).
-------

∂03-Oct-85  0750	HAUNGA@SUMEX-AIM.ARPA 	title and abstract for SIGLUNCH Friday 4th.    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 3 Oct 85  07:50:50 PDT
Date: Thu 3 Oct 85 07:50:26-PDT
From: Ana Haunga <HAUNGA@SUMEX-AIM.ARPA>
Subject: title and abstract for SIGLUNCH Friday 4th.
To: Siglunch@SUMEX-AIM.ARPA
Message-ID: <12148177201.13.HAUNGA@SUMEX-AIM.ARPA>


Here is the title and abstract for tomorrow's SIGLUNCH.


                            Introspection


                        Michael R. Genesereth


                             Logic Group
                     Knowledge Systems Laboratory
                         Stanford University


Abstract: Introspection is a significant part of human mental
activity.  We introspect whenever we think about how to solve problem,
whenever we decide what information we need to solve a problem,
whenever we decide that a problem is unsolvable.

By its nature, the process of introspection involves both descriptive
and prescriptive metaknowledge.  Over the past years, logicians and AI
researchers have devoted considerable attention to autoepistemic
sentences (involving terms like KNOW).  By comparison, little attention
has been paid to prescriptive metaknowledge (involving terms like OUGHT).

This talk introduces a semantics for such knowledge in the form of
constraints on the process of problem solving.  It demonstrates the
computational advantages of introspection, and analyzes the
computational fidelity and cost of various introspective
architectures.  Finally, it discusses the potential for practical
application in logic programming and building expert systems.
-------
-------
-------

∂03-Oct-85  1038	NILSSON@SU-SCORE.ARPA 	5-year Plan 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 3 Oct 85  10:30:41 PDT
Date: Thu 3 Oct 85 10:26:24-PDT
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: 5-year Plan
To: faculty@SU-SCORE.ARPA
Message-ID: <12148205595.27.NILSSON@SU-SCORE.ARPA>

I am in the process of writing a "five-year plan" for the Department.
This plan has been requested by Dean Gibbons and will form a part of the
plan for Engineering requested by Provost Jim Rosse.  The plan will
play a particularly important role in future events for the CSD.  It will
influence how planned new chairs are distributed within the SOE; it will
influence space decisions regarding the new CSD building; it will influence
new faculty position assignments and other resource allocations.  One 
important component of the plan will involve the resource needs projected
for the new undergraduate major.  So, in developing this plan, we need
to think carefully about what we want the department to be like during the
next five years and beyond. Do we have weaknesses that need to be corrected
by adding faculty in particular areas?  What do we think about how research
will be done (more research associates per faculty member and things like
that)?  I'm not even sure we have all the right questions yet, but I need
to get some preliminary draft material written in the next few weeks. So--
please send me any comments you might have that you think might be helpful
in this process--either questions we need to ask ourselves or answers.

I will certainly circulate drafts of anything I write before they ascend up
the system.  Thanks,  -Nils
-------

∂03-Oct-85  1236	WINOGRAD@SU-CSLI.ARPA 	Environments group - Monday 12:00pm  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 3 Oct 85  12:36:52 PDT
Date: Thu 3 Oct 85 12:35:46-PDT
From: Terry Winograd <WINOGRAD@SU-CSLI.ARPA>
Subject: Environments group - Monday 12:00pm
To: friends@SU-CSLI.ARPA

--sorry for the separate distribution but this didn't make it
   in time for the newsletter this week. --t
----------------------
COMING WEEK (Monday Oct. 7): 12:00 to 1:15 in the Ventura trailer
classroom (NOTE NEW REGULAR TIME), David Levy (Xerox PARC and CSLI)
will describe his work on a theoretical foundation for document
preparation environments.

------

PREVIOUS WEEK (Sept. 30):  At the first meeeting of the environments
group we set out the general directions for our discussions.  We
identified some major dimensions along which to compare and examine
environments and made an initial list of examples that might be
presented.  This list is very sketchy -- the random result of what
happened to come up in conversation.  We are eager for further details
and suggestions (either systems for general consideration, or about
which specific people would like to talk):

Programming environments: Interlisp, Smalltalk, Cedar, [all 3 Xerox],
(Linton) [Berkeley/Stanford], Gandalf [CMU], Mentor [INRIA], ZetaLisp
[Symbolics], Kee [Intellicorp], HPRL, HPLisp [last 2 Hewlett-Packard]

Grammar development environments: LFG [CSLI], HPSG [HP], BLT [CSLI],

Specification environments: Aleph [CSLI], (Balzer)[ISI]

Language development environments: MUIR [CSLI]

Document preparation environments:  (Levy) [CSLI], Notecards [Xerox]

Data access and manipulation envrionments: ?

Mathematical and logical deduction environments: MACSYMA [MIT], FOL
[Stanford]

There is a variety of application areas not as central to CSLI concerns,
but in which enviornments are built.  These include VLSI design,
CAD/CAM, image manipulation, mail systems, etc. In addition, most
operating systems take on the functions of an environment, either for
use outside of applications programs or as a base within them.
So-called "intelligent agents" are one attempt to provide a uniform
environment for a particular user interacting with multiple systems.

For each kind of environment there are specific problems dealing with
the particular structures being worked with (programs, proofs, grammars,
formatted documents, etc.).  There is also a framework of common
problems having to do with the basic structure of items being
manipulated (text, trees, databases, etc.), their representation on a
screen or hardcopy, interfaces for operating through that
representation, storage on one or more devices, consistency between
aspects (e.g., source and compiled code, specifications and proofs),
change over time (versions, releases, etc.), coordination of access
among a group, etc.

Our plan is to address the basic conceptual issues by looking at one
particular envrionment or group of related environments in each session.
Next week's topic will be a discussion of the theoretical foundations of
document preparation, by David Levy.
-------

∂03-Oct-85  1354	@SU-SUSHI.ARPA:broder@decwrl.ARPA 	Plane ticket swap wanted 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 3 Oct 85  13:53:55 PDT
Received: from decwrl.ARPA by SU-SUSHI.ARPA with TCP; Thu 3 Oct 85 13:49:59-PDT
Received: from magic.ARPA by decwrl.ARPA (4.22.01/4.7.34)
	id AA04315; Thu, 3 Oct 85 13:53:14 pdt
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA26503; Thu, 3 Oct 85 13:53:41 pdt
From: broder@decwrl.ARPA (Andrei Broder)
Message-Id: <8510032053.AA26503@magic.ARPA>
Date:  3 Oct 1985 1353-PDT (Thursday)
To: aflb.all@sushi.ARPA
Cc: 
Fcc: sent
Subject: Plane ticket swap wanted


I have a ticket to Portland on Oct 19 (Sat) at 12:20 p.m. from San Jose
on Air-Cal.  I'd like to swap it for an earlier flight the same day,
either from San Jose or from San Francisco.  Anyone interested?  -

Thanks, Andrei

∂04-Oct-85  1014	ACUFF@SUMEX-AIM.ARPA 	Symbolics Machines available at some ACM Conference  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 4 Oct 85  10:12:48 PDT
Date: Fri 4 Oct 85 10:12:18-PDT
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Symbolics Machines available at some ACM Conference
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12148465172.14.ACUFF@SUMEX-AIM.ARPA>

   There is supposed to be some ACM conference 13-16 of this month in
Denver, about which I know nothing.  The Symbolics folks asked me to
make it known that they would have machines there which they would be
willing to let us use for demos if anyone would like to do so.  Please
let me know if you are interested, and I'll put you in touch with the
correct entities.

	-- Rich
-------

∂04-Oct-85  1214	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	Next week's PLANLUNCH --- WEDNESDAY (not monday) OCT.9  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Oct 85  12:14:36 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Fri 4 Oct 85 12:09:30-PDT
Date: Fri 4 Oct 85 12:09:03-PDT
From: LANSKY@SRI-AI.ARPA
Subject: Next week's PLANLUNCH --- WEDNESDAY (not monday) OCT.9
To: planlunch.dis: ;

A quick note to planlunch attendees, and particularly FOLKS@CSLI:

Because of several complaints about duplicated messages, I have decided
to stop sending planlunch announcements to FOLKS@CSLI.  Instead, I will
be building my own local list of CSLI people who want to receive planlunch
announcements.  SO, if you receive your planlunch announcements via
FOLKS@CSLI and want to continue to do so, please send me your net address
and I will add it to this list.

Also, two more announcements:  
1) If you would like to give a planlunch talk, let me know.
2) Note the date change to WEDNESDAY for Lucy Suchman's talk (below).

Thanks, 
Amy Lansky
-----------------------------------------------------------------------

	                   WHAT IS A PLAN?

                             Lucy Suchman
    	          Intelligent Systems Lab, Xerox PARC

 	        11:00 AM, WEDNESDAY, October 9  (NOTICE DATE CHANGE)
       SRI International, Building E, Room EJ228 (new conference room)

Researchers in AI have equivocated between using the term "plan" to
refer to efficient representations of action, and to the actual data and
control structures that produce behavior.  But while these two uses of
the term have been conflated, they have significantly different
methodological implications.   On the first use, the study of plans, as
internal representations of actions and situations, is an important
companion to the study of situated actions, but essentially derivative.
On the second use, plans as the actual mechanisms that produce behavior
are foundational to a theory of situated actions. 

In this talk I will argue in support of the first use of "plans," to
refer simply to efficient representations of actions.  Situated actions,
on this view, are the phenomena to be modelled, whereas the function of
plans in the generation of situated actions is taken to be an open
question.  The interesting problem for a theory of situated action is to
find the mechanisms that bring efficient representations and particular
environments into productive interaction.   The assumption in classical
planning research has been that this process consists in filling in the
details of the plan to some operational level.  In contrast with this
assumption, I will present evidence in support of the view that situated
action turns on local interactions between the actor and contingencies
of his or her environment that, while they are made accountable to a
plan, remain essentially outside of the plan's scope.
-------

∂04-Oct-85  1358	JAMIE@SU-CSLI.ARPA 	graduate student offices 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Oct 85  13:58:13 PDT
Date: Fri 4 Oct 85 13:54:08-PDT
From: Jamie Marks <JAMIE@SU-CSLI.ARPA>
Subject: graduate student offices
To: folks@SU-CSLI.ARPA



We are trying to make more effective and equitable use of our space
this year, and are consequently doing some remodeling and
rearrangement of offices.  Part of the purpose is to provide more
study space for graduate students than we've been able to in the past.

We've planned an open door policy for a set of offices in the trailers
(E-1, E-2, and E-3) which provide semi-private desks for 7 students
and open onto a common room which provides 7 additional workstations.
Any unoccupied desk or workstation will be available to any graduate
student granted office privileges at CSLI.  We'll also provide lockers
in the common room for storage of private papers and belongings.

Please contact Jamie Marks (Jamie@CSLI) if you're interested in using
graduate student space.
-------

∂04-Oct-85  1508	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Oct 85  15:08:22 PDT
Date: Fri 4 Oct 85 15:03:29-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12148518182.22.LIBRARY@SU-SCORE.ARPA>

Programming Expert Systems In OPS5; An Introduction To Rule-Based Programming.
by Lee Brownston, Robert Farrell, Elaine Kant, and Nancy Martin. 
QA76.9.E96P76 1985.

Ada In Use. Proceedings of the Ada International Conference Paris 1985.
ed. by John Barnes, Gerald Fisher Jr. QA76.73.A16A25 1985

Algorithmic Graph Theory. by Alan Gibbs. Cambridge University Press.
QA166.G53 1985

Editing In A UNIX Environment: The vi/ex Editor. by Mohamed el Lozy.
QA76.6.L69 1985.

Handbook Of Systems Analysis: Overview Of Uses, Procedures, Applications
and Practice. by Hugh Miser and Edward Quade.  T57.6.H365 1985

H. Llull
-------

∂04-Oct-85  1554	NEUMANN@SRI-CSLA.ARPA 	RISKS-1.18  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 4 Oct 85  15:54:37 PDT
Date: Fri 4 Oct 85 14:07:29-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSLA.ARPA>
Subject: RISKS-1.18
Sender: NEUMANN@SRI-CSLA.ARPA
To: RISKS-LIST@SRI-CSLA.ARPA

RISKS-LIST: RISKS-FORUM Digest  Friday, 4 Oct 1985  Volume 1 : Issue 18

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
   Lack of a backup computer closes stock exchange (Marty Moore)
   DPMA survey on computer crime offenses (J.A.N. Lee)
   Ethics vs. morality (Marty Cohen)
   The Mythical Man-Month of Risk (Stavros Macrakis)
   Risk Assessment by real people (Mike McLaughlin)
   CRTs again, solution to one eye-problem (Mike McLaughlin)
   Failure of Mexican Networks (Dave Flory)
   Technical Reports Lists (Laurence Leff)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

[Note: Despite SRI-CSL having had a nasty disk controller wipe-out [Sun-Mon]
which prevented this issue from coming out on Monday, and despite my being
East, here is RISKS-1.18.]

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Mon, 30 Sep 85 11:44:49 CDT
From: mooremj@EGLIN-VAX
Subject: Lack of a backup computer closes stock exchange
To: risks@sri-csl.arpa

When Hurricane Gloria was approaching the New York area, the New York and 
American Stock Exchanges did not open.  The Midwest Exchange, located in
Chicago, opened on schedule; unfortunately, it had to close 40 minutes later,
when its nationwide computer system failed.  Where is the central computer
of that system located?  New York, of course.  The Director of the Exchange
was quoted as saying, "Well, this has got to change."

------------------------------

Date:     Wed, 2 Oct 85 15:39 EST
From:     Jan Lee <janlee%vpi.csnet@CSNET-RELAY.ARPA>
To:       RISKS@sri-csl.arpa
Subject:  DPMA survey on computer crime offenses

Peter ... here are the proper figures on Computer Crime Offenses as reported
by the DPMA from a survey taken by COMP-U-FAX (TM) and reported 1985 May 27:

(It doesn't say how many people were surveyed -- just that DPMA is an
organization of 50,000 members.)

21% reported one or more abuses in past 3 years in their workplace.  2% of
these offenses were committed by outsiders (so much for the Hacker myth!!).
The reasons for the abuses were:

  27% ignorance of proper professional conduct
  26% playfulness
  25% personal gain
  22% maliciousness

Only 45% of respondents worked for a company who had a full-time
or part-time data security person.

Only 2.2% of abuses were reported publicly (does that mean
reported to the news media or the legal authorities?).

The surveyor was Detmar Straub, Grad. School of Business Admin.,
Indiana University.

(In a note I got from Donn Parker, Donn seems to cast some aspersions
on the validity of this survey, but I haven't had chance to do anything
other than read the press release myself.)

JAN

------------------------------

Date:     Fri, 27 Sep 85 13:18:47 PDT
From: Marty Cohen <mcohen%NRTC@USC-ECL.ARPA>
To: risks@SRI-CSL.ARPA
Subject:  Ethics vs. morality

  "Morals: Society's code for individual survival"
  "Ethics: An individual's code for society's survival"

From Theodore Sturgeon, "More Than Human" ("Baby is Three" is one 
part of the book), page 177 of the Ballentine books paperback.

------------------------------

Date: Thu, 3 Oct 85 19:14:56 EDT
From: macrakis@harvard.ARPA (Stavros Macrakis)
To: risks@sri-csl
Subject: The Mythical Man-Month of Risk

In reference to the discussion on the Wilson article:
   > ... formula that measures risks [as] ... seconds of life lost ...

In discussing risks, whether from computer systems or other causes, it
would surely be desirable to have some reasonable guidelines.
Wilson's basic point was that we should be able to calculate costs and
benefits; a point with which I am in fundamental agreement.  However,
this calculation has many difficulties.  In this note, I should like
to sketch (in a somewhat disorganized way) some of these difficulties.

It is often useful to reduce complicated facts to simple unidimensional
measures for comparison.  But such reduction loses a great deal of
information in general; and also ignores variation in personal utility
functions.  For that matter, statistical measures should differentiate
between associations and causation.

Among the figures cited, there were several misleading measures along
these lines.

For instance, although perhaps cigarette smokers ought to allocate 12
minutes of their lives per cigarette, a lung cancer death is typically far
harder than an automobile accident death.  On the other hand, car accidents
subtract years by killing fewer, younger, people; cancer by killing more,
but older, people.  How to compare 5 x (lifespan 25-70) with 25 x (lifespan
63-70)?  Is each 175 man-years?  I have no answer, although I have certain
intuitions--in particular, `losing' the 69 man-years 1-70 seems far less
tragic than losing the 69 man-years 3 x (47-70) (`struck down in the prime
of life'): but the Wilson extract did explicitly talk only of 25+-year-olds.
Economics has various answers: discounted productivity contribution; market
value attached to hazardous jobs; ....  None is satisfactory, but all are
useful for separating grossly different risks.

As for correlations, why is it that living in New York is hazardous?
Perhaps it is the pollution.  But if it is the poverty or the street crime,
then poverty or bad neighborhoods (regardless of city) probably relate far
better statistically and surely causally than does residence.  New York just
has many poor people and bad neighborhoods.  A statistical analysis that
excludes such correlated hazards is surely non-trivial.

`Post hoc ergo propter hoc' seems especially implicated in the case of
unmarried males.  There are likely advantages to being married, but
perhaps inability to find or keep a wife indicates other problems.

Then there is the presumption of linear additivity.  Even the risks which
are not strongly correlated may combine: consider asbestosis and smoking.

Of course, in other cases the unidimensional metrics may be far more useful.
In the case of the costs of unreliable funds transfer, the cost to the bank
can be calculated quite precisely.  In cases where these errors affect
customers, it may also be reasonable to estimate that damage (e.g., you may
lose $100 of annual profits per error of type X if a retail customer takes
his business elsewhere).  If you're a materials supplier to a just-in-time
manufacturer, the monetary consequences may be far more serious: still, a
monetary measure may be meaningful.

In conclusion, it seems to me useful to develop a range of measures of cost
and benefit, and not try to reduce them to single numbers.  If one wishes to
be ultra-cautious, one will then weigh the minimum expected benefit against
the maximum expected cost.  If one is a `rational' gambler, perhaps the
average benefit against average cost.  If one is an optimist or a gambler,
perhaps the maximum benefit against the minimum cost.  I believe Howard
Raiffa (among others) discusses such issues (although I'm afraid I can't
provide a reference).

Risk-free systems are unlikely.  We need good ways of evaluating risks
and benefits.
	-s

------------------------------ 

Date: Sat, 28 Sep 85 16:05:48 edt
From: mikemcl@nrl-csr (Mike McLaughlin)
To: RISKS@SRI-CSLA.ARPA
Subject: Risk Assessment by real people

Ellen Goodman's column in The Washington Post, Saturday, 28 Sep 85, (c) 1985,
The Boston Globe Newspaper Company, is worth reading.  Title is: "AIDS: The 
Risks We'll Take."  No, it doesn't mention computers.  It really isn't much 
about AIDS, either.  

What it is about is people, and how risks are assessed by real people - not
by computers, or calculators, but by the folks that would die of boredom
reading this forum... or might die of something else if *we* do not
understand how *they* evaluate risks.

"In California, members of a family cut back on sugar in the decaffeinated 
coffee they drink in their house - on the San Andreas fault."  

"Another friend drinks only bottled water these days, eats only meat un- 
touched by steroids and spends weekends hang-gliding."  

"The AIDS story... is a tale about experts and the public, about the gap 
between our skepticism and our longing for certainty."  

Ms. Goodman also quotes an article in October's Science '85: "We may be much 
more willing to accept higher risks in activities over which we have control, 
such as smoking, drinking, driving or skiing, than things over which we have 
little control, such as industrial pollution, food additives and commercial 
airlines." 

Her summation is very relevant to *us*, who read and write about SDI, the use 
and non-use of computers, and so on: "This is, after all, a country that bans 
saccharin and builds nuclear bombs.  We argue and will go on arguing about 
risk in two different languages: numbers and emotions, odds and anxieties." 

If *we* cannot make ourselves understood by *them* when we discuss what 
matters for the survival of *all of us*, then this forum is just a
modern form of omphaloskepsis.  

- Mike 

------------------------------

Date: Sat, 28 Sep 85 16:25:57 edt
From: mikemcl@nrl-csr (Mike McLaughlin)
To: RISKS@SRI-CSLA.ARPA
Subject: CRTs again, solution to one eye-problem

CRTs can cause eye strain in users who wear glasses.  Distance lenses won't 
focus at closer than arm's length.  Reading lenses focus too close.  Bifocals 
require you to hold your head in one of two specific positions... neither of 
which works.  What to do?  

I already done it.  Several years ago.  So much a part of my computerized 
life that I didn't think of it while reading/writing via computer about CRT
usage.  Got "intermediate lenses" - an Rx for glasses optimized for my CRT/
VDT viewing habits.  Got comfy in front of the tube & keyboard, had a friend 
measure distance from bridge of nose to CRT screen.  Averaged several tries. 
Gave the distance to my opthamologist - he turned out an Rx in short order.  

Bought frame & lenses.  Expensive.  Use them *only* at office computer - don't 
take them home (don't have a computer at home).  Declared them as a non-
reimbursed business expense.  IRS content, helped reduce cost.  

Did NOT get tri-focals, because: 
  1.  They force you into an even more rigid head position - and I believe 
that rigid posture is a major cause of computer-fatigue.  
  2.  They confused the IRS issue, which was quite clear with intermediate-
only glasses.  
  3.  I'm not old enough to wear TRI-focals, for heaven's sake!  

Suggestion: 
  Employers requiring use of CRT should consider paying for intermediates; 
should resist paying for trifocals (or even the incremental cost of tri- over 
bi-focals).  
  An acceptable alternative might be bifocals, distance/intermediate or inter- 
mediate/reading, depending upon the user's eye condition and job content.  

	- Mike 

------------------------------

Date: Wed 2 Oct 85 13:43:03-EDT
From: Shadow <Z.HXWY-FLORY-DAVID%CRNL20A.BITNET@UCB-VAX.Berkeley.EDU>
Subject: Failure of Mexican Networks
To: Risks@sri-csl.arpa

This doesn't refer to computer networks, but it is similar.

According to Fred Goldstein on Telecom Digest (Telecom-Request@MIT-MC.ARPA),
phone service from most of the world to Mexico City was destroyed by the
collapse of the building containing the switches, frames, etc. for Mexico
City's international gateway switch.

Sites which are major network nodes collapsing due to earthquake/etc could
result in a similar effect.

David Flory

ARPANET ---> flory@CORNELL-GVAX.ARPA or shadow@RU-AIM.ARPA
BITNET ----> z.hxwy-f@CRNL20A.BITNET

    [Good engineering tends to avoid sensitivity to single-point failures 
     and to avoid singly connected nodes.  Designing for massive failures
     and disasters is of course much more difficult.  PGN]

------------------------------

Date: Fri, 27 Sep 85 13:38:21 cdt
From: Laurence Leff <leff%smu.csnet@CSNET-RELAY.ARPA>
Subject: Technical Reports Lists
To:  [... all sorts of lists...]

   [Here is what could be a useful service if suitable indexing occurs.  I
    have stripped Laurence's message down.  SEND to him for the original.
    This is of course more general in scope than just RISKS, but seemed
    worth including. -- in case you missed it elsewhere.  Respond to him,
    not me.  PGN]

I have volunteered to organize an electronic mechanism for the distribution
of technical report lists from Universities and R&D labs.  Some (and
hopefully all) of the people producing technical reports would send a copy
of the list to me.  I would then send these to a moderated group on USENET
as well as a mailing list for those sites on the INTERNET who do not get
news (ARPANET, CSNET, etc.).

I need two things from you:
  1) if your organization prepares technical reports and sends them
      out to interested parties (perhaps for a fee), please arrange
      to have electronically readable copy of your lists sent
      to trlist%smu@csnet-relay.  
  2) if people at your organization would like to receive lists
     of tech reports produced by universities and R&D labs, please
     provide me an electronic address to send them to (if you are not
     on USENET).  Send such administrative mail to trlist-request%smu@
     csnet-relay.

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂05-Oct-85  1030	JF@SU-SUSHI.ARPA 	Abstracts and other Information about 10/11/85 BATS 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 5 Oct 85  10:30:20 PDT
Date: Sat 5 Oct 85 10:27:29-PDT
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: Abstracts and other Information about 10/11/85 BATS
To: aflb.su@SU-SUSHI.ARPA
cc: su-bboards@SU-SUSHI.ARPA

Fellow theorists:  I have received only 7 email responses about next week's
BATS.  Do you want to go or not?  I enclose the talk abstracts with this
message; hence there is NO EXCUSE for further delay.  I need at least several
firm committments from drivers (several people said they might drive but
weren't sure). 

In case you delete them, the abstracts are in

SUSHI:<JF>bats.abstracts

Wearily yours,
Joan


	J. Rissanen (IBM San Jose):
	Stochastic Complexity

Modifying the algorithmic notion of complexity we define the stochastic
complexity I(x) of a string of symbols x = x(1),..., x(n),
relative to a class of parametrically defined probabilistic models
to be the infinum of the code length when the data are encoded with
prefix codes, designed suitably for the given class of models. There is
an element of arbitrariness in specifying exactly the "suitable" design,
but for long strings the complexities given by
all such choices tend to a common bound, for
which an asymptotic formula can be given.
 
Stochastic complexity with its associated inequality has notable
consequences. First, it generalizes
Shannon's information and the associated inequality.
Secondly, stochastic complexity also sets a tight lower bound for the
prediction errors. Thirdly, the main inequality provides a way to
assess the goodness of any estimator for the parameters including their
number. Accordingly, unless too presumptuous, we submit the computation
of the stochastic complexity and the associated optimal model to be the
fundamental problem in all statistics and signal processing.

	Joel Friedman (UC Berkeley):
	Non-Blocking Networks

     A "switching network" is a model for a telephone communication
network.  A switching network is a directed acyclic graph whose set 
of sources (i.e. vertices with indegree 0) is disjoint from the set
of sinks.  The sources are called "callers," and the sinks are called
"receivers."  At any time, a caller may request to talk to a receiver,
provided they are both idle, i.e. not engaged in another conversation.
We must then connect the caller to the receiver via a path in the graph
which is vertex disjoint from any other conversation path currently
in use.  At any time the caller can request to hang up and, perhaps,
talk to a different receiver.  A network that can handle any such 
sequence of requests is called "non-blocking."  In short, a non-blocking
network can dynamically handle all hookup and hangup requests from
callers to receivers.

     Two variants of the concept "non-blocking" appear in the literature:
"strictly" non-blocking, and "wide-sense" non-blocking.  When a 
caller requests to talk to a receiver, there may be many paths available
to hookup the two.  A strictly non-blocking network is one where
hookup requests can be handled by any available path.  A wide-sense
non-blocking network is one with a strategy for choosing paths (given
the state of the network and the caller and receiver) which makes the
network non-blocking.  A wide-sense network can be blocked if an 
adversary chooses the connecting paths, a strictly non-blocking network
can't.

     In this talk I will describe new results which, for the first time,
show that wide-sense non-blocking networks can be less expensive that
strictly non-blocking networks.  In the simplest case we show that depth
2 strictly non-blocking must have n↑2 edges, while wide-sense networks
exist with n↑1.5 (log n)↑.5 edges.


	Allen Van Gelder (Stanford):
	Parallel Complexity of Function-Free Logic Programs
	(joint work with Jeffrey D. Ullman)

We consider the parallel time complexity of logic programs
without function symbols.
We give a PRAM algorithm for computing the minimum model of a function-free
Horn logic program, and show that for programs with the
``polynomial fringe property,'' this algorithm runs in poly-log time.
As a result, the ``linear'' and ``piecewise linear'' classes of
logic programs, as well as the word problem for CFLs, are in NC.
(Using ideas of Miller and Reif that were recently presented at MSRI,
we can actually show log↑1 PRAM time and NC↑2 membership.)

Then we examine several non-linear cases in which the program has
a single recursive rule that is an ``elementary chain.''
We demonstrate a connection between such programs and context free grammars.
Among such ``elementary single rule'' programs that are non-linear,
we exhibit cases, analogous to balanced parentheses languages,
that have the ``polynomial fringe property,'' hence are in NC.
Also, we exhibit an elementary single rule program that is P-complete.


	Janet Blumer (UCSC):
	Minimal Automata For The Set Of All Suffixes And The Set Of All Subwords

The Directed Acyclic Word Graph (DAWG) is a directed acyclic graph that
can be viewed as the minimal DFA for the set of all suffixes of a word.
It will be proved linear in the size of the word, and an algorithm will
be described which builds the DAWG on-line in linear time.  With a
different assignment of accepting states, the DAWG can be viewed as a DFA
for the set of all subwords of a word.  As such, it is not minimal, but
closely related to the minimal subword automaton.  This relationship
is examined, producing tighter bounds on the size of the minimal subword
automaton and indicating modifications to the DAWG construction algorithm
that allows the minimal subword automaton to be built on-line in linear 
time.
-------

∂05-Oct-85  1701	@SU-CSLI.ARPA:barwise@csli-whitehead 	Linguistics Positions 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 5 Oct 85  17:01:23 PDT
Received: from csli-whitehead ([36.9.0.8].#Internet) by SU-CSLI.ARPA with TCP; Sat 5 Oct 85 16:56:31-PDT
Received: by csli-whitehead with TCP; Sat, 5 Oct 85 14:59:14 pdt
Date: Sat, 5 Oct 85 14:59:14 pdt
From: Jon Barwise <barwise@csli-whitehead>
Subject: Linguistics Positions
To: folks@csli

CMU is starting a new program in linguistics in the philosophy department
and is advertising two tenure track positions.  This seems like a great
opportunity, since the Pittsburgh
area is so strong in CS, Phil and Logic.  Ingrid has copies of the
announcemnet if you want one.
Jon

∂07-Oct-85  0032	DAVIES@SUMEX-AIM.ARPA 	This week's meeting   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  00:32:22 PDT
Date: Mon, 7 Oct 1985  00:33 PDT
Message-ID: <DAVIES.12149146214.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: This week's meeting
cc:   Davies@SUMEX-AIM.ARPA

It has been difficult to find a technical topic for this week's
meeting.  Some people were out of town last week, so they couldn't
contribute to the selection of a topic.  Others will be out of town
this week, so they had little to say about the topic of a meeting they
wouldn't be attending.  Nonetheless, there are some business matters
to be discussed, so there will be a meeting Wednesday at 9:30 am.
Suggestions for a technical topic are welcomed.

	-- Byron

∂07-Oct-85  0632	PATASHNIK@SU-SUSHI.ARPA 	Next AFLB 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  06:32:23 PDT
Date: Mon 7 Oct 85 06:28:50-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Next AFLB
To: aflb.all@SU-SUSHI.ARPA

We have no speaker for next week, October 17, yet.  Please let me
know if you'd like to speak.

		********************************

10-Oct-85 - Eli Shamir (Hebrew University)

		New results about graph coloring

The chromatic number of a random graph with expected degree d(n) is
sharply concentrated around b*d(n)/logd(n).  There are provably good
heuristics to unravel the coloring of a dense graph with an unusually
small number of color classes.

***** Time and place: October 10, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352 (Bldg.
460).

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

For more information about future and past AFLB meetings and topics
are in the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂07-Oct-85  0815	EMMA@SU-CSLI.ARPA 	Logic Lunch
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  08:15:04 PDT
Mail-From: BARWISE created at  6-Oct-85 14:45:14
Date: Sun 6 Oct 85 14:45:14-PDT
From: Jon Barwise <BARWISE@SU-CSLI.ARPA>
Subject: Logic Lunch
To: AFA@SU-CSLI.ARPA, clt@SU-AI.ARPA, paolo@SU-CSLI.ARPA
cc: SF@SU-CSLI.ARPA, emma@SU-CSLI.ARPA
ReSent-Date: Mon 7 Oct 85 08:13:51-PDT
ReSent-From: Emma Pease <Emma@SU-CSLI.ARPA>
ReSent-To: "@ps:<csli-lists>finterest.dis"@SU-CSLI.ARPA

Just in case any of you missed the announcement in the Newsletter, we
are starting a Logic Lunch tomorrow (Monday).  Ordinarily we will meet
in the philosophy lounge, but tomorrow it will be in the philosophy
seminar room.  Tell all your friends.
-------

∂07-Oct-85  0819	HAUNGA@SUMEX-AIM.ARPA 	Title & abstract for the SIGLUNCH, Friday 11th.
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  08:18:52 PDT
Date: Mon 7 Oct 85 08:16:34-PDT
From: Ana Haunga <HAUNGA@SUMEX-AIM.ARPA>
Subject: Title & abstract for the SIGLUNCH, Friday 11th.
To: SIGLUNCH@SUMEX-AIM.ARPA
cc: Haunga@SUMEX-AIM.ARPA
Message-ID: <12149230536.25.HAUNGA@SUMEX-AIM.ARPA>


The SIGLUNCH will be held at the Chemistry Gazebo at
12:05 p.m. - 1:00 p.m.

-------------

		TWO EFFORTS TOWARD SOFTWARE ICAI
		
				BY

			Lewis   Johnson

Software systems are a rich domain for intelligent computer-
based training.  In this talk I will discuss two efforts which
relate to software ICAI.  The first is PROUST, a system which
finds bugs in programs written by novice Pascal programmers. 
PROUST analyzes bugs by identifying each programmer's intentions, and
their reflection in their programs.  This results in a detailed
understanding of the bugs and the misconceptions that probably
caused them, information which is indispensable to an intelligent
tutoring system for programming.  Results of classroom tests of
PROUST will be presented.

The second effort is the development of a high-level software
design language, called ODETTE.  Software is difficult to understand
and maintain because programming languages describe program behavior, and
do note relate it to the user's behavior, nor to the designer's and
user's goals.  ODETTE enables designers to make all of these
explicit in the design.  Our long-term goal is to build tools to
support enhancement of ODETTE designs, and to generate intelligent
interfaces with built-in explanation and training facilities.


------ relate it to the user's behavior, nor to the designer's and
-------

∂07-Oct-85  0921	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  09:21:06 PDT
Date: Mon 7 Oct 85 09:09:24-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA, bureaucrats@SU-SCORE.ARPA
Message-ID: <12149240155.17.RICHARDSON@SU-SCORE.ARPA>


The title and speaker for tomorrow's lunch in MJH 146 at 12:15 is:
Dean Gibbons on "CSD and the School of Engineering".

Anne
-------

∂07-Oct-85  0949	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  09:48:49 PDT
Received: by su-aimvax.arpa with Sendmail; Mon, 7 Oct 85 09:46:47 pdt
Date: Mon, 7 Oct 85 09:46:47 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

I'm going to be away this Wednesday, so I
suggest we cancel the meeting unless
anyone volunteers to lead a discussion.
In general, we need volunteers to present
their own ideas or discuss any of a number of
papers that I have received recently and look interesting.
For example:

Termination of rewriting (Dershowitz)

Parallel Algs. for term matching (Dwork, Kanellakis, and Stockmeyer)

Bounded Recursion in deductive DB's (Ioannidis)

Ordering conjunctive queries (Smith and Genesereth)

Finding all solutions to a problem (Smith and Genesereth)

Controlling recursive inference (Smith, Genesereth, and Ginsberg)

Volunteers??

∂07-Oct-85  1026	ullman@su-aimvax.arpa 	paper just received   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  10:26:27 PDT
Received: by su-aimvax.arpa with Sendmail; Mon, 7 Oct 85 10:22:27 pdt
Date: Mon, 7 Oct 85 10:22:27 pdt
From: Jeff Ullman <ullman@diablo>
Subject: paper just received
To: nail@diablo

WOuldn't you know it: 2 minutes after I sent out my list
of papers, the following arrived:
Recursive axioms in deductive databases: the querry-subquery
approach, by Laurent Vieille

It seems to be another algorithm in the Lozinskii/McKay-Shapiro
class, in fact, probably equivalent to Van Gelder's algorithm,
in that he seems to distinguish "masks," which are patterns
in which constants appear.

∂07-Oct-85  1444	WASHINGTON@SUMEX-AIM.ARPA 	Siglunch location 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  14:44:37 PDT
Date: Mon 7 Oct 85 14:41:53-PDT
From: Rich Washington <washington@SUMEX-AIM.ARPA>
Subject: Siglunch location
To: sjg@SU-AI.ARPA, siglunch@SUMEX-AIM.ARPA
Message-ID: <12149300682.68.WASHINGTON@SUMEX-AIM.ARPA>

Judging from attendance the past two weeks, the chemistry gazebo
is painfully inadequate for the Siglunches.  I agree that it is
a much more desirable location than Braun Auditorium, but I think
it's only facing reality to admit that we are in a popular field
of research/study.  We should be welcoming people in to see what
we're working on, and staying in the present cramped conditions
can only have the opposite effect.

				Rich
-------

∂07-Oct-85  1841	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar on Logic and the Foundations of Mathematics    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 7 Oct 85  18:35:02 PDT
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 7 Oct 85 18:32:27-PDT
Date: 07 Oct 85  1822 PDT
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Seminar on Logic and the Foundations of Mathematics   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    



Speaker: Marian Pour-El, University of Minnesota, visiting MSRI 

Title: Computability of standard processes in analysis and physics.

Time: Monday, Oct.14, Noon to 1:15 

Place: Ventura Hall Seminar Room

                                     S. Feferman



Note the change of time and place!!

The regular meeting time of this seminar has been changed to Friday noon.
We will meet alternate weeks beginning Friday October 25th.

∂07-Oct-85  1850	CLT  	Seminar on COMMON SENSE AND NON-MONOTONIC REASONING   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    


A series of seminars on


		COMMON SENSE AND NON-MONOTONIC REASONING


will explore the problem of formalizing commonsense knowledge and reasoning,
with the emphasis on their non-monotonic aspects.

	It is important to be able to formalize reasoning about physical objects
and mental attitudes, about events and actions on the basis of predicate logic,
as it can be done with reasoning about numbers, figures, sets and probabilities.
Such formalizations may lead to the creation of AI systems which can use logic
to operate with general facts, which can deduce consequences from what they know
and what they are told and determine in this way what actions should be taken.

	Attempts to formalize commonsense knowledge have been so far only
partially successful. One major difficulty is that commonsense reasoning often
appears to be non-monotonic, in the sense that getting additional information
may force us to retract some of the conclusions made before. This is in sharp
contrast to what happens in mathematics, where adding new axioms to a theory
can only make the set of theorems bigger.

	Circumscription, a transformation of logical formulas proposed by John
McCarthy, makes it possible to formalize non-monotonic reasoning in classical
predicate logic. A circumscriptive theory involves, in addition to an axiom set,
the description of a circumscription to be applied to the axioms. Our goal is
to investigate how commonsense knowledge can be represented in the form of
circumscriptive theories.

	John McCarthy will begin the seminar by discussing some of the problems
that have arisen in using abnormality to formalize common sense knowledge about
the effects of actions using circumscription.  His paper Applications of
Circumscription to Formalizing Commmon Sense Knowledge is available from Rutie
Adler 358MJH.  This paper was given in the Non-monotonic Workshop, and the
present version, which is to be published in Artificial Intelligence, is not
greatly different. The problems in question relate to trying to use the formalism
of that paper.

	The seminar will replace the circumscription seminar we had last year.
If you were on the mailing list for that seminar then you will be automatically
included in the new mailing list. If you would like to be added to the mailing
list (or removed from it) send a message to Vladimir Lifschitz (VAL@SAIL).

	The first meeting is in 252MJH on Wednesday, October 30, at 2pm.

∂08-Oct-85  0011	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	ICDT 86 - Call For Papers 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  00:11:27 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 8 Oct 85 00:04:12-PDT
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Mon 7 Oct 85 23:10:54-PDT
Received: by rsch.wisc.edu; Tue, 8 Oct 85 00:48:27 cdt
Received: from crys.wisc.edu by rsch.wisc.edu; Mon, 7 Oct 85 20:05:06 cdt
Received: from sri-iu.arpa by crys.wisc.edu; Mon, 7 Oct 85 20:04:48 cdt
Date: Mon 7 Oct 85 18:06:40-PDT
From: Moshe Vardi <VARDI@SRI-IU.ARPA>
Subject: ICDT 86 - Call For Papers
To: theory@WISC-CRYS
Message-Id: <VAX-MM(162)+TOPSLIB(114)+PONY(0) 7-Oct-85 18:06:40.SRI-IU.ARPA>
Sender: udi@rsch.wisc.edu
Resent-From: udi@rsch.wisc.edu (Udi Manber)
Resent-To: theory:;








                      Call For Papers
                          ICDT 86
        International Conference on Database Theory
                 Rome, September 8-10, 1986


     The conference is intended to provide a European  forum
for  the  international  research  community  on theoretical
issues related to database systems.   It  will  be  held  in
downtown  Rome,  in  the  main  building of CNR (the Italian
Research Council), sponsored by the European Association for
Theoretical  Computer  Science (EATCS) and jointly organized
by Consorzio per la Ricerca e la Applicazioni in Informatica
(CRAI),  Dipartmento  di Informatica e Sistematica - Univer-
sita di Roma "La Sapienza", and  Instituto  di  Analisi  dei
Sistemi   ed   Informatica  del  Consiglio  Nazionale  delle
Ricerche (IASI-CNR).

←λT←λo←λp←λi←λc←λs

     Major themes to be covered are (this is not meant to be
an  exclusive list): relational theory, logic and databases,
conceptual models, knowledge representation  and  databases,
deductive  databases,  theory  of  distributed databases and
concurrency control, analysis and design of data structures,
database interfaces, and query processing.

←λP←λa←λp←λe←λr←λs

     Authors should submit six copies of a full draft before
March 15, 1986, to the Chairman of the Program Committee:

          Giorgio Ausiello
          Dipartmento di Informatica e Sistematica
          Universita di Roma "La Sapienza"
          Via Eudossina, 18
          00184 Roma, Italy

     Authors will be notified of the program committee deci-
sion  on  their papers by June 1, 1986.  Final versions will
be due July 15, 1986.  Proceedings will be available at  the
Conference  in  the form of preprints, and will be published
in a book form a few months later.

←λP←λr←λo←λg←λr←λa←λm ←λC←λo←λm←λm←λi←λt←λt←λe←λe

     S. Abiteboul (France); G.  Ausiello  (Italy);  F.  Ban-
cilhon, (France, USA); A. D'Atri (Italy); W. Lipski (Poland,
USA); M. Moscarini (Italy),  J.  Mylopoulos  (Canada);  J.M.
Nicolas  (France,  FRG);  J.  Nievergelt (Switzerland); C.H.
Papadimitriou (Greece, USA);  J.  Paredaends  (Belgium);  D.
Sacca (Italy); N. Spyratos (France); J.D. Ullman (USA); M.Y.
Vardi (USA).

←λO←λr←λg←λa←λn←λi←λz←λi←λn←λg ←λC←λo←λm←λm←λi←λt←λt←λe←λe

     P. Atzeni (IASI-CNR); A. D'Atri (Universita  di  Roma);
M. Moscarini (IASI-CNR); D. Sacca (CRAI).

←λS←λt←λu←λd←λe←λn←λt ←λA←λw←λa←λr←λd

     A $500 prize for the best paper authored solely by stu-
dents  will  be  awarded  to commemorate our friend and col-
league Witold Lipski (1949-1985), who departed after  agree-
ing to serve on the program committee.  Those who wish to be
considered for the prize are invited to formally state their
student status in a letter accompanying the paper.

←λF←λu←λr←λt←λh←λe←λr ←λI←λn←λf←λo←λr←λm←λa←λt←λi←λo←λn
          Paolo Atzeni
          IASI-CNR
          Viale Manzoni, 30
          00185 Roma, Italy
          Tel. (39 6) 77.00.31
-------

∂08-Oct-85  0111	NEUMANN@SRI-CSL.ARPA 	RISKS-1.19   
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  01:10:56 PDT
Date: Mon 7 Oct 85 23:58:57-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSL.ARPA>
Subject: RISKS-1.19
Sender: NEUMANN@SRI-CSL.ARPA
To: RISKS-LIST@SRI-CSL.ARPA

RISKS-LIST: RISKS-FORUM Digest  Monday, 7 Oct 1985  Volume 1 : Issue 19

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Emanations and interference in the civil sector (Peter Neumann,Jerry Saltzer)
  Administrivia -- Escaped Mail and Delays (Mark S. Day)
  Computer databases (Andy Mondore)
  Re: Friendly test teams (John Mashey)
  Re: CRTs again, solution to one eye-problem (Brint Cooper)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions
  should be relevant to the topic, technically sound, objective, in good
  taste, and coherent.  Others will be rejected.  Diversity of viewpoints is
  welcome.  Please try to avoid repetition of earlier discussions.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Sun 6 Oct 85 15:16:38-PDT
From: Peter G. Neumann <Neumann@SRI-CSLA.ARPA>
Subject: Emanations and interference in the civil sector
To: RISKS@SRI-CSLA.ARPA

I have had several queries about risks in the civil sector concerning
electronic emanations from and electronic interference upon computer systems
and networks -- and of course also about what can be done to protect oneself
or one's company.  For example, Martin Lee Schoffstall <schoff%rpi.csnet
@csnet-relay.arpa> wondered along these lines:

	If you were building a hospital from scratch, would you consider
	shielding for your computer room, how many electron volts would
	you shield for, etc.?

	In general I would like some feedback for us civilians...

This subject is generally a technically intricate one, but some guidance is
clearly necessary for the civil sector.  Thus, it seems worthwhile to note
several examples that represent varying degrees of risk to the public.
(Since microprocessor-controlled systems are becoming ubiquitous, related
problems are likely to recur in other guises.  But let us not quibble about
whether THESE examples are sufficiently computer-related.)  The first three
examples involve interference; all but the third involve emanations.

  Transmit: Microwave oven emanations
  Receive:  "Externally reprogrammable" heart pacemaker --  interference;
            pacemaker reset by microwaves to 214 beats per minute
  Result:   Dead patient (See Software Engineering Notes vol 5 no 1 Jan 1980.)

  Transmit: Anti-theft device emanations
  Receive:  Heart pacemaker -- interference
  Result:   Patient OK (See Software Engineering Notes vol 10 no 2 Apr 1985,
            but I have seen nothing more recent.)

  Transmit: Active radar jammer (in speeder's auto)
  Receive:  Police radar receiver
  Result:   In one currently popular device, the jammer simulates a fault
            mode common in the design of many police radars systems.
            (... a program bug or an electronic interface problem?)

  Transmit: Police radar transmitters
  Receive:  Radar signals (received by transmitter, and by targetted autos)
  Result:   With passive detector, driver can avoid arrest.

  Transmit: Microwave telephone transmitters (telephone company)
  Receive:  Capture telephone conversations and data (observer)
  Result:   Compromise

  Transmit: Radiating CRT or keyboard (unsuspecting computer user)
  Receive:  Recreate screen image or typed input remotely (observer)
  Result:   Compromise  (Unclassified technology for doing this has
            recently been described in a European defense magazine.)
            [The RISKS Forum has discussed CRT radiation with respect
            to possible health hazards, so I won't list that again.]

The radar detector and jammer are marginally computer-relevant, and are
included here primarily because they are illustrative of deeper problems --
fielding a computer system and its surrounding environment that can be
defeated in some relatively simple way.  [By the way, this forum does not
endorse or promote illegal activities -- we merely need to point out their
existence.]  (I have not included in this list the garage-door openings and
closings triggered by the orbiting Sputnik, which happened to be on the
right frequency.)

Emanations and interference may be accidental or intentional.  Passive
techniques for detection may require some computing as in the case of
unscrambling multiplexed communications.  Active techniques (e.g., for
intentional jamming) are at this point much less common, but are likely to
present greater risks in the future.  There are all sorts of more or less
relevant laws, but they are probably neither complete enough nor concise
enough.  There are also all sorts of commonly available devices for those
who want to break the laws.

This note is intended to help raise the general level of awareness.  With
pretenses of corporate secrecy being what they are, it would be nice to be
able to assess the real risks.  In the past, many of these risks have seemed
obscure, but that seems to be changing.  Suggestions on how to avoid those
risks are welcome.  (There are of course also nonelectronic forms of
emanations; various penetrations are reported to have begun with information
-- including passwords -- gained by reading the contents of dumpsters.)

The answers to Marty Schoffstall's hospital query, and other such questions,
depend on the perceived risks against which you think you are defending.
For Marty's example, are you trying to provide survival of the hospital
computers and communications against nuclear attack? or something less
serious such as intentional jamming or accidental interference?  Might you
be worried about compromises of privacy resulting from wire-taps and
microwave pickups of computer information?  Each threat suggests a variety
of possible measures or countermeasures.   PGN

------------------------------

Date:  Fri, 4 Oct 85 18:02 EDT
From:  Saltzer@MIT-MULTICS.ARPA <Jerry Saltzer>
Subject: Emanations and interference in the civil sector
To:  Neumann@SRI-CSL [in response to a query]

Concern for Electromagnetic Compatibility is indeed beginning to become an
important design consideration in consumer products.  These days, TV sets
are beginning to clean up their act, but the average FM tuner just can't
cope with being in a substantial RF field.  As consumers start to collect a
walkman, TV, cable converter, FM tuner, stereo amplifier, VCR, CD player,
cordless phone, remote control light switches, microwave oven, and
garage-door opener under one roof, more and more people are becoming aware
of the problems, and discovering that some manufacturers didn't put the
right effort in.

------------------------------

Date: Thu 3 Oct 85 20:07:38-EDT
From: Mark S. Day <MDAY@MIT-XX.ARPA>
Subject: Administrivia -- Escaped Mail and Delays

[ Excerpted-From: Soft-Eng Digest    Sat,  5 Nov 85    Volume 1 : Issue  34 ]

XX was a victim of Hurricane Gloria; it had multiple head crashes when it
was restarted after the storm.  The heroic efforts of the staff here brought
the machine back to life after a marathon of restoring files, which
unfortunately left the alias for this list in a strange state.  Instead of
going into my mailbox, everything sent to "Soft-Eng" was immediately
redistributed.  Fortunately, only one message got out between the time XX
came up and the time I noticed the problem.  Anyway, sorry for the
difficulties.  No doubt this will now appear in the RISKS mailing list as an
example of an unreliable computer system...  

   [SURE.  WHY NOT??!! Recovery and reinitialization are a vital part of
    keeping a system running properly.  How many times have you put in a
    patch or fix only to find that it somehow disappeared, e.g., not 
    surviving a crash or not getting propagated back into the source code?  
    But in this case you got left in an unsafe state!  PGN]

------------------------------

Date: Sat, 28 Sep 85 16:20:46 EDT
From: Andy←Mondore%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
To: RISKS@sri-csl.arpa
Subject: Computer databases

One topic I have not seen discussed here is that of computer databases.  I
am Systems Coordinator for the Registrar's Office here so I am in charge of
a fairly large database containing (obviously) student grade and course
information as well as addresses, demographic information, etc.  I'd like to
see a discussion of the risks of having incorrect information in a database,
information being seen or accessed by the unauthorized individuals, etc.
Thanks.

    [Ah, yes.  This is a wonderful topic.  The state of the art of database
     management systems that can handle sophisticated privacy/compromise and
     data integrity problems is rather abysmal.  However, the risks of
     people gleaning information by drawing inferences from a database are
     considerable.  For starters, see Dorothy Denning's book, Cryptography
     and Data Security, Addison Wesley, 1982.  As to risks, Software
     Engineering Notes has had a bunch of stories on the effects of misuse
     or mininterpretation of police data.  The Air New Zealand catastrophe
     was an example of what can happen if a change is not propagated
     properly.  As always, contributions are welcome.  PGN]

------------------------------

Date: Sat, 28 Sep 85 22:31:18 pdt
From: mips!mash@glacier (John Mashey)
To: glacier!risks@sri-csl.ARPA
Subject: Re: Friendly test teams

It might be good to ask for pointers to published data on bug histories,
effort levels, robustness in large hardware/software systems.  I suspect
these may be hard to find for SDI-like systems; I couldn't dig up any old
Safeguard info.  Although not in the same class of difficulty, ATT's new #5
ESS switch is fairly complex (300+ engineers).  A good reference is:  H.A.
Bauer, L.M. Croxall, E.A. Davis, "System Test, First-Office Application, and
Early Field Experience", ATT Technical Journal, vol 64, No 6, Part 2
(Jul-Aug 1985), 1503-1522.

------------------------------

Date:     Sun, 6 Oct 85 12:59:18 EDT
From:     Brint Cooper <abc@BRL.ARPA>
To:       mikemcl@NRL-CSR.ARPA
cc:       RISKS@SRI-CSL.ARPA
Subject:  Re:  CRTs again, solution to one eye-problem

     [We started out keeping one eye on this problem, but it does not
      want to stay out of sight.  Will this be the last message?  PGN]

A cheaper but similar solution was suggested by my opthalmalogist when I
attained that stage of life wherein my arms are too short.

Since I needed a small, positive correction (about +1.0) in each eye, I
purchased, at his suggestion, "reading glasses" from the local pharmacy for
about $12.00.  Since then, my eyes have worsened a little and I need about
+1.25 to +1.5 diopters for reading.  But this is too strong for the terminal
(an AT&T 5620 with rather small font), so I retained the old +1.0 diopter
lenses for the terminal at work.  At $12.00 each, I can afford to have a
pair at the office, a pair at home, and a pair to carry.

Note:  This won't work if one has astigmatism or if one needs widely
different corrections in each eye.  But ask your doc.  You can buy a lot of
OTC glasses for $200.

Oh yes, it is a small nuisance to switch glasses from terminal lenses to
reading lenses, but one learns quickly to minimize the hassle.

Brint

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂08-Oct-85  0912	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  09:11:56 PDT
Date: Tue 8 Oct 85 09:09:30-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA, bureaucrats@SU-SCORE.ARPA
Message-ID: <12149502317.16.RICHARDSON@SU-SCORE.ARPA>

Don't forget lunch today at 12:15 in MJH 146!

-------

∂08-Oct-85  0936	@SU-SCORE.ARPA:TW@SU-AI.ARPA 	Reminder about forum talks    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  09:36:46 PDT
Received: from SU-AI.ARPA by SU-SCORE.ARPA with TCP; Tue 8 Oct 85 09:34:09-PDT
Date: 08 Oct 85  0935 PDT
From: Terry Winograd <TW@SU-AI.ARPA>
Subject: Reminder about forum talks    
To:   faculty@SU-SCORE.ARPA
CC:   tajnai@SU-SCORE.ARPA   

This is a reminder that I have requested names of students who might speak
at this year's forum, to be sent to me by Oct 11 (this Friday).  So far
only five faculty members have responded to my request.  One of them
proposed four candidates and another said he had enough for a whole
session.  We would like a broader representation, so please respond if you
haven't. --t

∂08-Oct-85  1043	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  10:42:31 PDT
Date: Tue 8 Oct 85 10:33:33-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA, bureaucrats@SU-SCORE.ARPA
Message-ID: <12149517616.16.RICHARDSON@SU-SCORE.ARPA>

Unfortunately, Dean Gibbons will be unable to attend the CSD Lunch today.
Instead, Nils Nilsson will lead an informal discussion about the policies
for hiring and promoting Research Associates.

-------

∂08-Oct-85  1424	@SU-CSLI.ARPA:LANSKY@SRI-AI.ARPA 	PLANLUNCH mailing list    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  14:22:58 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 8 Oct 85 14:16:41-PDT
Date: Tue 8 Oct 85 14:02:13-PDT
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH mailing list
To: folks@SU-CSLI.ARPA

This is a reminder to those of you who:
1) want to continue getting PLANLUNCH notices
2) are NOT on sri's aic-associates list
3) have not yet sent a message to me concerning remaining on the list

to send me a notice that you WANT to remain on the mailing list.
I am constructing my own subset of of FOLKS@CSLI to send PLANLUNCH notices to.

I hope you could parse this message!,
-Amy Lansky  (LANSKY@SRI-AI)
-------

∂08-Oct-85  1724	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  17:24:05 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 8 Oct 85 17:20:52 pdt
Date: Tue, 8 Oct 85 17:20:52 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

remember--no meeting tomorrow.

∂08-Oct-85  1756	BSCOTT@SU-SCORE.ARPA 	No-Smoking Policy 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  17:56:04 PDT
Date: Tue 8 Oct 85 17:50:28-PDT
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: No-Smoking Policy
To: CSD@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12149597154.11.BSCOTT@SU-SCORE.ARPA>

Nils Nilsson has asked me to remind everyone about the Stanford No-Smoking
Policy.

Smoking is prohibited in indoor locations where smokers and non-smokers occupy
the same area.  Such areas include:

     Academic areas:  classrooms, lecture halls, seminar rooms, laboratories,
     libraries, computing facilities.

     Conference rooms, auditoria, exhibition areas, indoor athletic facilities,
     theaters, pavilions, and retail stores.

     Health facilities not subject to Stanford University Hospital or Stanford
     University Medical Center no-smoking regulations.

     Open areas (shared spaces not fully enclosed by floor to ceiling partitions
     and doors).

     Other common/public areas including:  stairwells, elevators, escalators,
     lobbies, hallways, waiting rooms, reception areas, restrooms, and 
     customer service areas.

     Any area in which a fire or safety hazard exists.

-------

∂08-Oct-85  1910	NEUMANN@SRI-CSL.ARPA 	RISKS-1.20   
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 8 Oct 85  19:10:25 PDT
Date: Tue 8 Oct 85 17:45:26-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSL.ARPA>
Subject: RISKS-1.20
Sender: NEUMANN@SRI-CSL.ARPA
To: RISKS-LIST@SRI-CSL.ARPA

RISKS-LIST: RISKS-FORUM Digest  Wednesday, 9 Oct 1985  Volume 1 : Issue 20

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Risks using robots in industry (Bill Keefe)
  Re: Computer databases (Matt Bishop)
  Registrar's databases; Database risks - census data (Hal Murray, 2 messages)
  The winners of evolution... (William McKeeman)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

Warning:  Dave Poole will be at it again tonight -- risk of multiple copies!

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Tuesday,  8 Oct 1985 07:26:03-PDT
From: keefe%milrat.DEC@decwrl.ARPA  (Bill Keefe)
To: risks@sri-csl.ARPA, keefe%milrat.DEC@decwrl.ARPA
Subject: Risks using robots in industry

I don't know who to credit with typing this in.  I was going to summarize,
but it's too easy to take some points out of context. It brings up many 
questions as to who bears the responsibility (liability?) to protect people 
from such occurrences.
 
                   --------------------------------

            In The Lion's Cage"      [Forbes  Oct. 7, 1985]

On July 21, 1984, at about 1 p.m., a worker at Diecast Corp. in Jackson, Mich.
found Harry Allen, 34, a diecast operator pinned between a factory pole and the 
back of an industrial robot. But Allen's co-worker couldn't come to his aid. 
Using the robot's controller, the company's director of manufacturing finally 
unpinned Allen, who was alive but in cardiac arrest. He died in a hospital 
five days later.

Allen had entered a restricted area,  presumably to clean up scrap metal from 
the floor. While there, he got in the way of the robot's work, and thus became 
the first - and so far only - U.S. victim of an industrial robot-related 
accident.

That's not a bad safety record, considering that 17,000 robots are now 
installed in the U.S. But the bet is he won't be the last. The Japanese, who 
lead the world in robot installations, also lead in robot-related fatalities: 
There have been reports of at least 5, and possibly as many as 20, such deaths 
in Japan.

That's only fatalities. In this country, companies are not required to report 
injuries related to specific equipment, so no reliable data are available. But 
in Sweden, a pioneer in the use of industrial robots, one study estimates that 
there is 1 accident a year for every 45 robots. By 1990, when the number of 
robots installed in American Industry could climb as high as 90,000, the 
number of injuries could climb accordingly. That's because robots move quickly 
and are programmed to go through a series of motions without stopping. A 
worker who gets in the way can be struck, pushed aside, crushed or pinned to 
a pole as Allen was.

How will industry minimize the risk to its workers? Probably with difficulty. 
Robots don't easily accommodate safeguards. Whereas most machinery operates 
within a fixed set of boundaries, robots have a large "striking distance" - the 
reach of their mobile arms within three dimensions. In automotive assembly 
plants, maintenance workers often collide with robots adjacent to the ones 
they're servicing because they don't realize they are in another robot's work 
area. A robot may perform one task five times and then start on a completely 
different  activity, and with it a different set of motions. Also, a robot can 
sit idly for a time and then come to alive again, threatening injury to a 
worker who mistakenly thought it was shut down.

What's being done to make robots safer? Right now, not much. "The extent of 
most safety precautions are signs saying, 'Restricted Area: Keep Out,' or 
maybe a guardrail," says Howard Gadberry of the Midwest Research Institute in 
Kansas City, Mo. Indeed, the most common safeguards - perimeter barriers such 
as guardrails and electric interlocked gates, which automatically shut down 
the robot when opened - don't protect those maintenance workers and programmers 
who must enter the lion's cage. Presence-sensing devices, such as 
pressure-sensitive mats and light curtains, both of which automatically cut 
off a robot's power, also don't seem to offer as much protection as is needed, 
if only because workers are even more unpredictable in their movements than 
robots. They may not step on the mat when feeding parts to a robot, or they 
may not break a light curtain's beam.

That's not to say that robots can't be made safer. Researchers at the 
Renssalaer Polytechnic Institute, for example, recently completed a research 
prototype for several large U.S. companies of a four-sensor safety system that 
continuously monitors the area around a robot. Using ultrasonic, infrared, 
capacitance and microwave sensors, the RPI system is designed to stop a robot 
in its tracks if a worker gets too close. Cost?  Five thousand dollars 
in production, according to Jack Meagher, a senior project manager at RPI.

The National Bureau of Standards has also been working with ultrasonic sensors 
on robot arms similar to the system at RPI. They both have developed a 
secondary, or watchdog, computer to monitor the actions of the robot and its 
microprocessor. After all, if the robot's computer goes berserk, how can it 
monitor itself? That's more important than you might think, 30% of robot 
accidents seem to be caused by runaways, according to John Moran, director of 
research at the National Institute for Occupational Safety & Health.

While such systems slowly make the transition form research to the factory 
floor, industry is trying to put basic safety standards into practice. 
Recently, the Robotic Industries Association proposed a set of national safety 
standards for robots that could go into effect as early as next summer.

Would such standards have prevented Harry Allen's death? Maybe not. The robot 
at the Diecast plant was surrounded by a safety rail with an electric 
interlocked gate that automatically shut down the robot when the gate was 
opened. However, there were two gaps in the rail that allowed workers to 
easily bypass the safeguard; that has since been corrected by the company.

Says Allan Harvie, deputy director of the Michigan Department of Labor's 
bureau of safety and regulation, "I could only presume Harry Allen thought he 
could go in and do what he intended to do without having to shut the robot 
down."

------------------------------

Date:  8 Oct 1985 1123-PDT (Tuesday)
From: Matt Bishop <mab@riacs.ARPA>
Organization: Research Institute for Advanced Computer Science
Address: Mail Stop 230-5, NASA Ames Research Center, Moffett Field, CA  94035
Phone: (415) 694-6363 [main office], (415) 694-6921 [my office]
To: risks@sri-csl.ARPA
Subject: Re: Computer databases

   I guess I'll start the ball rolling on this discussion.

   I think the greatest risk is not from the technological end but the
human end.  For instance, there was a case a couple of weeks back where
someone got stopped for a traffic ticket.  Call this gentleman John Lee
Jones (I've forgotten his real name.)  A routine computer check showed
James Lee Jones was a fugitive from an LA warrent, and the description
of James Lee Jones was pretty close to what John Lee Jones looked like.
So the SFPD hauled him downtown, and ran a fingerprint check to see
if there was anything else they could find out about John Lee Jones.
Turned out he had used several aliases in the past -- so the SFPD
notified the LAPD they had arrested James Lee Jones, and would the LAPD
please come up and get him?  The LAPD obliged, took him down to LA,
and notified the prosecutors.

   Throughout all this, Mr. Jones was (vehemently) denying he was James
Lee Jones.  About a week after he had first been locked up, his public
defender persuaded the judge to order the police to compare John Lee
Jones' fingerprints with James Lee Jones' fingerprints.  They didn't
match.  End of case.

   What's so surprising is that the people throughout the whole
proceeding did not question whether the data the computer gave them was
relevant.  True, it was accurate (so far as I know.)  But it was used
incorrectly.  In other words, in this case the technology didn't fail;
the human safeguards did.  (Incidentally, in defense of the police,
when this came out an investigation was begun to see why the
fingerprint comparison was not made immediately; according to police
procedure it should have been.)  And no amount of database security can
guard against this type of breach of security.

[Caveat -- I read the newspaper story I outlined above a couple
 of weeks ago in the S.F. Chronicle.  I have undoubtedly misremembered
 some of the details, but the thrust of it is correct.]

Matt Bishop

    [Add that to the database-related cases of false arrest reported in 
     RISKS-1.5.   PGN]

------------------------------

Date: Tue, 8 Oct 85 06:59:25 PDT
From: Murray.pa@Xerox.ARPA
Subject: Registrar's databases
To: RISKS@sri-csl.arpa

Just mentioning grades, computers and risks, all in the same paragraph
instantly brings to my mind visions of hackers who are flunking freshman
English smiling anyway, knowing that they have figured out how to get an A.

I've always assumed that everybody "knew" that students and grades couldn't
really coexist on the same machine. Does anybody know of a school
brave/silly enough to do it?

It seems like a great opportunity for somebody who makes a secure system to
get a LOT of publicity, one way or the other. Has anybody ever been
confident enough to try it? What happened?

Changing the topic slightly... Security on an ethernet is clearly
non-existent unless you encrypt everything you care about. Our personnel
people upstairs take the problem seriously. The solution is simple. They
have their own section of coax. It's not even gatewayed to the rest of our
network.

------------------------------

Date: Tue, 8 Oct 85 07:15:28 PDT
From: Murray.pa@Xerox.ARPA
Subject: Database risks - census data
To: RISKS@sri-csl.arpa

The census bureau distributes their data broken down to quite small areas. I
don't know the details, but I'm pretty sure it gets down to "neighborhood"
sized regions, and it may even get down to the block.  When the sample size
gets small enough, there are obviously opportunities for gleaning
non-statistical information by using carefully crafted querys to read
between the lines.

I remember somebody telling me that they worked pretty hard to make sure
this couldn't happen. Anybody know what they actually do? Is it written up
someplace? Does it work well enough? Any war stories? Are the techniques
simple once you know the trick? ....

   [As I noted in RISKS-1.19, Dorothy Denning's book is a good source.
    The Census Bureau tries to add phony data that preserves all of the
    overall statistics but that prevents inferences...   PGN]

------------------------------

Date: Tue 8 Oct 1985 10:00:15 EST
From: William McKeeman <mckeeman%wang-inst.csnet@CSNET-RELAY.ARPA>
Subject: The winners of evolution...
To: risks@sri-csl

A recent submission included the following paragraphs on evolution
of morals...

	 Now I think it fairly easy to see that the capacity to put
	 group survival ahead of self-interest is an important genetic
	 trait and that tribes of people that had this trait would be
	 more likely to survive that tribes that didn't.  That is not
	 to say that this moral capacity doesn't vary greatly from one
	 person to the next or that even that it may not be more fully
	 realized in one person than another because of upbringing.  It
	 is even possible that, because of some genetic error, some
	 people may be born without a moral capacity, just like they
	 might be born without arms or legs.

	 Moral progress means the evolution of survival customs more
	 appropriate to the current context.  The trouble in recent
	 centuries has been that our ability to evolve new technology
	 has outstripped our capacity to evolve the appropriate
	 morality for it.  There is a strong tendency to stick to the
	 morality that one learns as a child, even if it [is] not
	 appropriate to the current situation.

Evolution is being used with its Darwinian meaning but with an
interpretation that includes the more ordinary progress of mankind.  The
central mechanism of evolution is the failure of the less successful forms
to reproduce -- often for failing to live long enough.  Evolution is never
fast enough to avoid bloodshed -- it is bloodshed that activates it.  Until
disaster strikes the adapted and unadapted survive undifferentiated.

My point is that if we treat the present sad state of affairs as a problem
in evolution rather than politics or technology, we are implicitly planning
on rebuilding a world with the (apparently) adapted survivors of WWIII.

W. M. McKeeman            mckeeman@WangInst
Wang Institute            decvax!wanginst!mckeeman
Tyngsboro MA 01879

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂09-Oct-85  0008	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #21  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  00:05:47 PDT
Date:  8 Oct 85 2347-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #21
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 9 Oct 1985      Volume 1 : Issue 21

Today's Topics:

                              Administrivia
                 Communication and Control Structures
                     Reference: Coordinated Computing
        Seminar: Parallel Algorithms for Form Perception (GMR)
   Seminar: Recognition Algorithms for the Connection Machine (MIT)


----------------------------------------------------------------------

Date: Tue, 8 Oct 1985  23:42 PDT
From: Byron Davies <PARSYM-Request@SUMEX-AIM.ARPA>
Subject: Administrivia

So far there have been only three responses to the PARSYM Survey on
Data Abstractions for Parallel Programming.  Since the survey was
highly technical in nature, I didn't expect a large number of
responses, but I was hoping for a few more than I've seen.  The first
respondent, Tony Li, admitted that he wasn't doing any work on data
abstractions, but went ahead to report on some work on communications
and control structures.  I've included Tony's response in this digest,
as a normal -- and much appreciated -- contribution to PARSYM.  The
other two responses, by Bert Halstead of the Multilisp project at MIT
and by Dave Touretzky of the Boltzmann Machine project at CMU, will
appear when I receive at least two more responses.

By the way, in the week since the questionnaire came out, PARSYM
hasn't received many messages of any kind.  Please don't let the
survey prevent you from asking questions or passing along interesting
information about parallel computing.  There are lots of people (about
250 individuals plus over 30 local mailing lists or bboards) waiting
to hear from you.

While I'm in moderator mode, I'll respond publicly to a few of the
notes I've received recently.  Most of the messages addressed to
PARSYM-Request are administrative requests, to add a name or remove
one or consolidate a group of names into a bboard.  I've only received
a few complaints.  A few weeks ago, I was asked to be more careful
about inserting editorial comments because I was breaking an
undigestifier.  Never having undigestified anything, I wasn't aware of
the problem, but I've learned my lesson and I hope that undigestifiers
are gurgling happily across the country.

When Laurence Leff sent a l-o-n-g message about his establishing a
resource of technical abstracts, I had some doubts about including it
in the PARSYM digest because it wasn't about parallel computing.  I
decided to go with it, however, because parallel computing, being so
new to most people, is one domain that could particularly benefit from
good communication.  While I only received one complaint about this, I
want PARSYM readers to know that PARSYM is still and only the netwide
mailing list for parallel symbolic computing.

I encourage PARSYM readers to let me know what they like and dislike
about PARSYM.  I'm happy to entertain suggestions about how PARSYM can
be improved or about what sorts of information you'd like to see.  If
you have some suggestions for future PARSYM surveys, please let me
know.

Keep those messages and survey responses coming in!

	-- Byron

------------------------------

Date: 3 Oct 1985  00:47 PDT (Thu)
From: Tony Li <Tli@Usc-Eclb>
Subject: PARSYM Digest   V1 #20

    1.  Are you using or have you developed data structures or
        abstractions specialized for parallel computing?  Please
        describe.

We're following the more conventional path of communications and
control structures.  This may not be particularly relevant, but
anyhow...

We've developed two basic ideas.  The first is bus communications.
This is distinctly similar to Channels in CSP or Occam, but with
multiple senders and receivers.  Further, there is a tagging mechanism
so that processes can restrict their communications to a subset of the
possible correspondents.  As with CSP, our buses are synchronous.

Our other basic idea is that of the parallel process call.  This is
implemented in two separate constructs (PAR-AND and PAR-OR) which
encapsulate a number of process calls and communications statements.
PAR-AND is basically similar to a restricted COBEGIN.  Multiple
processes are created and the original process blocks until all
processes have terminated and returned values.  Assignment of return
values is done non-deterministically.

The PAR-OR statement terminates when any single process call or
communications statement completes.  Only the completing statement is
allowed to return values (or complete an I/O operation).

Obviously, we're not too far along, and none of this is set in
concrete.  I'd be happy to answer questions or listen to comments.

Cheers,
Tony ;-)

------------------------------

Date: Thursday, 3 October 1985  08:15-PDT
From: marick at GSWD-VMS
To:   DAVIES at SUMEX-AIM
Re:   Coordinated Computing

I hadn't heard of this book.  Who's the publisher?  When was it published?

[Coordinated Computing: Tools and Techniques for Distributed Software.
 New York:McGraw-Hill, 1984.  I highly recommend it as an introduction
 to and overview of a variety of work in parallel computing. -- BD]

Thanks.

Brian Marick
Gould CSD - Urbana

------------------------------

Date:     Wed, 2 Oct 85 16:00 EST
From: "S. Holland" <holland%gmr.csnet@CSNET-RELAY.ARPA>
Subject:  Seminar - Parallel Form Perception (GMR)

[Forwarded from Vision-List]

                 General Motors Research Laboratories
                           Warren, Michigan


                   Parallel Algorithms for Form Perception

                               Dana H. Ballard
                         The University of Rochester
                             Rochester, New York

                         Thursday, October 17, 1985

                                  ABSTRACT

     Vision systems with real time performance will have to make extensive use
     of parallel algorithms and computer architectures. One key problem such a
     system will have to solve is the management of form.  This includes shape
     recognition, navigation, segmentation and object aquisition.  These sub-
     problems potentially can be solved in parallel by a single algorithm based
     on finding rigid transformations.  The talk will focus on the underlying
     theory as well as computer experiments.

     Dana H. Ballard received his Ph.D. in 1974 from the University of
     California, Irvine.  He has written numerous papers and books on computer
     vision.  He is presently an Associate Professor of Computer Science at the
     University of Rochester.

-Steve Holland

------------------------------


Date: Fri, 4 Oct 1985  12:31 EDT
From: ELIZABETH%MIT-OZ@MIT-MC.ARPA
Subject: 9.382 -- Vision seminar

[Forwarded by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

	Recognition Algorithms for the Connection Machine

                             Anita Flynn

                          Monday, October 7
                   Eighth Floor Playroom, A. I. Lab
                               4:00 pm


  Many problems in early vision appear to be inherently local and
exploitble on parallel computer architectures.  This talk describes a
problem in late vision, that of recognizing an unknown object and
matching it to a data base model given only sparse sensory data
points.  The algorithm presently used on a sequential machine is first
explined, and then various algorithms for parallel computation are
explored.  Tradeoffs in space-time efficiency are discussed in terms
of implementation on a Connection Machine.  The parallel version is
shown to run three to four orders of magnitude faster than the
sequential one.

------------------------------

End of PARSYM Digest
********************

∂09-Oct-85  0155	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #42
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  01:55:09 PDT
Date: Tuesday, October 8, 1985 4:08AM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #42
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest           Wednesday, 9 Oct 1985      Volume 3 : Issue 42

Today's Topics:
               Implementation - Destructive Assignment,
                  & Connected Components & Cut Tests
----------------------------------------------------------------------

Date: Mon, 30 Sep 85 09:32:42 EDT
From: Paul Hudak <Hudak@YALE.ARPA>
Subject: Prolog vs Lisp and destructive assignment

The analogy that Prolog is to Logic Programming as Lisp is to
Functional Programming is a good one.  The analogy would be
even stronger if Prolog had destructive assignment, although
it's easy to argue that Prolog's Assert and Retract are forms
of destructive assignment.

However, as Peter Deutsch points out, despite our need for practical
languages, the "purists" should not give up hope that clever compiler
techniques might make pure logic or functional programming a practical
reality.  For example, there now exist fairly powerful inferencing
strategies for converting call-by-need evaluation to call-by-value,
even in the presence of higher-order functions.  This makes lazy
evaluation not nearly as expensive as it was once thought.  As for
destructive assignment, several strategies have been developed to
infer that a purely functional update is being done to an object with
a *single reference*, so that the update can be done destructively
"behind the scenes."  Papers on this topic include an early paper by
Schwartz (1978), one by Mycroft (Edinburgh thesis 1981), and a recent
one by Hudak and Bloss (POPL 85).

In the latter paper we show that if one were to translate (in the
most "obvious" way) an imperative program using destructive assign
-ment on arrays, into a purely functional program without side
effects, then it will always be the case that the updates could be
done destructively behind the scenes.  The reason is rather obvious:
in an imperative program, once a change is made to an array there is
no way to access the old version of it, and the most obvious
translation to a functional program tends to preserve this property!
Indeed, this highlights precisely the *power* (rather than
disadvantage)  of functional or logic programs in being able to refer
to, or "hang on" to, old objects.  It is a feature that can be
simulated in conventional languages only be making explicit copies
of the array in question.  (To formalize the notion of translation
we use a version of McCarthy's flowchart-schema-to-recursive-schema
translation.)

Although I haven't done it, I also believe that these results could
be generalized to logic programming.

Well, this is the good news.  The bad news is that it is not clear
how one resaons about the space-efficiency of such a program, since
no inferencing strategy can guarantee that *all* copying is avoided.
I don't have an immediate solution to this problem, but am not too
concerned because:

1) In the past we have been able to educate programmers well enough
   that they realize that a tail-recursive call is as space-efficient
   as a loop.  (Indeed, this situation can be viewed as a special
   case of updating arrays!)  Perhaps a similar degree of education
   would help here.

2) It is possible to engineer special constructs that enforce the
   single-reference property, while allowing the elegance of a purely
   functional solution.  Such constructs appear, for example, in
   certain data-flow languages.

-- Paul Hudak

------------------------------

Date: Tue, 1 Oct 85 11:13:07 -0200
From: Ehud Shapiro  <udi%wisdom.bitnet@WISCVM.ARPA>
Subject: Connected Components

In my paper, A Subset of Concurrent Prolog and its Interpreter
(ICOT TR-003), there is a Concurrent Prolog solution to the
connected components problem (which also happens to be a logic
program solution, if you delete the read-only's).  Since the
program and its description are rather short, I enclose them
here.

We associate with each node in the graph a distinct integer
number.   The name of a connected component is the smallest
number of any node  in the component.  The algorithm, for a
graph with vertices V is:

  For each node n in V do
       Set Xn to n.
      Repeat |V| times:
          Send Xn to all nodes adjacent to n,
          Receive a number from every adjacent node,
          Let Xn' be the minimum of Xn and the numbers received.
          Set Xn to Xn'.
      Set n's connected component number to Xn.

The algorithm is not very efficient.  For an n-vertex graph its
computation depth is in the order of n, and its length is in the
order of n@+(2).  This means, roughly, that given n processors,
the algorithm runs in linear time.  Better parallel algorithms
are known (Shiloach&Vishkin), but this algorithm --- and its
implementation --- are among the simplest.

The program assumes a specialized adjacency list representation
of the graph.  With each node N we associate a stream variable
Xn. The graph is represented as a list of n triples (N, Xn, As),
where N is the node number, Xn is its associated stream variable,
and As is a list of the stream variables associated with the
nodes adjacent to N.@:  Given this graph, the program computes
a list of pairs (N, C), where C is the name of the connected
component of the node N.

(1) cc(Graph, CList) :-
        cc(Graph, Graph, CList).

(2) cc(Graph, [(N, [N|Xn], As)|Gs], [(N, C)|Cs]) :-
        node(Graph, N, Xn, As, C),
        cc(Graph, Gs, Cs).
(3) cc(←, [], []).

(1) node([←|G], Xn, [Xn1|Xns], As, C) :-
        min(Xn, As?, Xn1, As1), node(G, Xn1, Xns, As1, C).
(2) node([], C, [], ←, C).

(1) min(Xn, [[B|Bs]|As], Xn1, [Bs|As1]) :-
        lt(Xn, B) | min(Xn, As?, Xn1, As1).
(2) min(Xn, [[B|Bs]|As], Xn1, [Bs|As1]) :-
        le(B, Xn) | min(B, As?, Xn1, As1).
(3) min(Xn, [], Xn, []).


Clauses (1)-(3) spawn a node process for each node in V; note
that they generates the stream of pairs (N, C) before the
component name C of each node is determined.

Each node process has four arguments.  The first is the graph
itself, which it uses to count |V| iterations.  Following are
the Xn variable explained above, the list of streams of adjacent
nodes, and the node's final component number.  The node process
iterates |V| times, using Clause (1), and on each iteration
performs the operations described above.

The @i(min) process extracts the smallest element among all the
first elements of the adjacent streams and the current node number,
and returns a list of the rest of the streams.

For example, if we invoke the program on the 7-vertex graph in
which 1 is connected to 2 and 3, 2 is connected to 4, 5 is
connected to no one, and 6 is connected to itself and to 7, we
get the following result:

| ?- solve(cc([(1, X1, [X2, X3]), (2, X2, [X1, X4]),
(3, X3, [X1]), (4, X4, [X2]

|              (5, X5, []), (6, X6, [X6, X7]), (7, X7, [X6])], Cs)).
|
Cs = [(1, 1), (2, 1), (3, 1), (4, 1), (5, 5), (6, 6), (7, 6)],
X1 = [1, 1, 1, 1, 1, 1, 1, 1],
X2 = [2, 1, 1, 1, 1, 1, 1, 1],
X3 = [3, 1, 1, 1, 1, 1, 1, 1],
X4 = [4, 2, 1, 1, 1, 1, 1, 1],
X5 = [5, 5, 5, 5, 5, 5, 5, 5],
X6 = [6, 6, 6, 6, 6, 6, 6, 6],
X7 = [7, 6, 6, 6, 6, 6, 6, 6]
}

which shows, in addition to the component numbers, also at which
cycle each node discovered its correct component name.  We see
that  after three cycles every node knew its final component name.

In the paper "Distributed Programming in Concurrent Prolog"
(Weizmann Institute TR CS84-02, by A. Shafrir and myself, a
Concurrent Prolog implementation of a much more complex
distributed algorithm for connected components is described.

-- Ehud Shapiro

------------------------------

Date: Mon, 7 Oct 85 11:51:16 BST
From: mcvax!cdsm@doc.ic.ac.uk
Subject: Results of cut tests

                         Results of CUT Tests

I've received a large number of replies both via the network and
letters and am very grateful to all those who took the trouble
to send in results. The results as they currently stand are
listed below, and the test is also repeated (I've added a cut to
the first clause of 'do' to allow for those systems which give
all results, e.g. UNSW). As you may see, the second one listed
(all Y except for 'not' test) is numerically most popular but it
does not include any of the compiled systems. Somewhat
surprisingly, the 'Edinburgh Standard' is not a standard.


Implementation                     Test 1  2  3  4  5  6

DEC-10 (Comp)                           Y  Y  Y  Y  Y  Y
Waterloo, MUprolog, Quintus(Int),       Y  Y  Y  Y  Y  N
UNH(1.3), Salford, Criss, Prolog-1,
BIM, York
Quintus (Int)                           Y  Y  Y  Y  Y  I
MProlog(1.5)                            Y  Y  Y  Y  N  N
Prolog-2 (Int, Comp)                    Y  Y  Y  N  N  Y
DEC-10 (Int), C-Prolog, Prolog-X        Y  Y  Y  N  N  N
POPLOG, LM-Prolog                       Y  Y  Y  I  I  I
Quintus (Comp)                          Y  Y  I  N  N  N
micro, sigma                            Y  N  N  Y  Y  N
UNSW                                    Y  N  N  Y  N  N
ICL                                     Y  N  N  N  N  N

where Y means did cut, N means did not cut and I means that it
was trapped as an illegal use and failed.

(There are several contradictory results for the Quintus
Compiler, presumably corresponding to different releases.)

If there are any other discriminating cases not covered above I
would be glad to hear of them. I would stress that I do not
consider there to be any "right" answers.

---------------------------------------------------------
/* Tests to distinguish various implementations of cut */
/* Chris Moss, Imperial College, June 1985 */

test1 :- do('Testing that cut is implemented). ', t1).
test2 :- do('Test if cut acts within disjunction', t2).
test3 :- do('Test if it cuts previous choices within disjunction',t3).
test4 :- do('Test if cut acts when passed as metacall', t4).
test5 :- do('Test if & cut acts within metacall', t5 ).
test6 :- do('Test if cut acts through not', t6).

do(Message,Test) :- w(Message), Test, !.
do(Message,Test) :- w('Does act').
w(X) :- write(X), nl.
t :- w('Does not act').

t1  :- (true;w('Did not cut alternatives correctly'),fail),
       !, w('Succeeds going forwards'), fail.
t1  :- w('Failed to cut goal').

t2  :- (!;w('Fails to cut disjoint alternatives')), fail.
t2  :- t.

t3  :- t3a(X),(!,fail;w('Fails to cut disjunction')).
t3  :- t.
t3a(!).
t3a(X) :- w('Did not cut alternatives'), fail.

t4  :- t3a(X), X, w('Ok going forwards'),fail.
t4  :- t.

t5  :- t5a(X), X, fail.
t5  :- t.
t5a((true,!)).

t6  :- not(not(!)), fail.
t6  :- t.

------------------------------

Date: Mon, 7 Oct 85 11:50:20 BST
From: mcvax!cdsm@doc.ic.ac.uk
Subject: Assert/Retract tests

                     Tests for Assert and Retract

Below is a set of tests to look at the effects of assert and
retract along the lines of my earlier tests for cut. I would be
very grateful if anyone with access to a Prolog not listed in
the set of results which I already have could run this and send
me the results (directly) and I will post the list at some
future date in the network. So far I haven't found two
implementations which give the same results! For some
implementations, you may need to add declarations or make other
slight changes to get the desired effects.

------------------------------------------------------------
main :-
    test1, test2, test3, test4, test5, test6, test7, test8, test9.
    /* Tests to distinguish various implementations of assert */

test1 :- do('1. Test assert on existing predicate',t1).
test2 :- do('2. Test assert on new predicate',t2).
test3 :- do('3. Does it find added 3rd clause on backtracking', t3).
test4 :- do('4. Does it find added next (last) clause on backtracking', t4).
test5 :- do('5. Test retract on single element clause',t5).
test6 :- do('6. Does retract backtrack? ', t6).
test7 :- do('7. Does retract work on invoked clause',t7).
test8 :- do('8. Does retract work on next (last) invoked clause',t8).
test9 :- do('9. Test retracts and asserts on invoked clause', t9).

do(Message,Test) :- w(Message), Test, !.
do(Message,Test) :- w('No Clause found').
w(X) :- write(X), nl.
t :- w('Added clause').
tr :- w('Failed to retract clause').

t1  :- assert((t1a :- t)), t1a.
t1a :- w('Clause 1'),fail.

t2  :- assert((t2a :- t)), t2a.

t3  :- assert((t3:-t)),fail.
t3  :- fail.

t4  :- assert((t4:-t)),fail.

t5  :- retract((t5a :- tr)), t5a.
t5a :- tr.

t6  :- retract((t6a:-X)),fail.
t6  :- t6a.
t6a :- tr.
t6a :- w('Does not backtrack').

t7  :- t7a, fail.
t7  :- fail.
t7  :- tr.
t7a :- retract((t7:-tr)),!.

t8  :- t8a, fail.
t8  :- tr.
t8a :- retract((t8:-tr)),!.

t9  :- retract((t9:-w('Second'))),assert((t9:-t)),fail.
t9  :- w('Second').
-----------------------------------------------------

Results I already have are as follows:

                        1   2 3 4 5 6 7 8 9
cprolog                 CA  A A N N N N N A
Dec-10, interpreted     CA  A A A N F N N A
Dec-10  compiled        CN  N N N N F F F S
micro prolog (CP/M-80)  CA  A A N E D F N A
muprolog                CA  A A A N N N N A
poplog                  CA  A N N N N F F S
sigmaprolog (unix)      CA  A A N E D N F A
sigmaprolog (Dec front) CA  A A N N F N F A

where letters stand for the first letter of the message printed
out by the test, and E stands for a "no such predicate" error
message. Note that in the latter case you may have to split up
the "main" procedure into several parts.

This set of tests is certainly not complete. In particular it
doesn't explore the interactions of assert and retract (except
for one test, 9), nor does it look at versions of assert and
retract which have numbered arguments. It's main purpose is
probably to ring the alarm bells about the differences which
have crept in and point to the need for reform. One might note
that retract together with garbage collection and reuse of an
"orphaned" pointer is the easiest way to crash many Prolog
systems. I haven't included such a test as the details appear to
differ for every implementation which is liable.

-- Chris Moss

------------------------------

End of PROLOG Digest
********************

∂09-Oct-85  0954	HANRAHAN@SU-SCORE.ARPA 	petty cash 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  09:53:56 PDT
Date: Wed 9 Oct 85 09:47:49-PDT
From: Katherine Hanrahan <HANRAHAN@SU-SCORE.ARPA>
Subject: petty cash
To: csd@SU-SCORE.ARPA
Stanford-Phone: (415) 497-2273
Message-ID: <12149771436.10.HANRAHAN@SU-SCORE.ARPA>

Due to an administrative delay in the accounting department there will be
no petty cash for our department until this friday, October 11th after
1:30.  Sorry for any inconvenience this may cause.  Katie
-------

∂09-Oct-85  1024	@SU-SCORE.ARPA:ejm@su-shasta.arpa 	Re:  topics for prob. theory course
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  10:21:59 PDT
Received: from su-shasta.arpa by SU-SCORE.ARPA with TCP; Wed 9 Oct 85 10:18:39-PDT
Received: by su-shasta.arpa with TCP; Wed, 9 Oct 85 10:20:51 pdt
Date: Wed, 9 Oct 85 10:20:51 pdt
From: Ed McCluskey <ejm@su-shasta.arpa>
Subject: Re:  topics for prob. theory course
To: MAYR@SU-SCORE.ARPA, faculty@SU-SCORE.ARPA

what'λλλλλwhat's the matter with stat116dλs?
ejmc

∂09-Oct-85  1113	@SU-SCORE.ARPA:ejm@su-shasta.arpa 	Re:  petty cash
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  11:13:10 PDT
Received: from su-shasta.arpa by SU-SCORE.ARPA with TCP; Wed 9 Oct 85 11:06:15-PDT
Received: by su-shasta.arpa with TCP; Wed, 9 Oct 85 11:08:38 pdt
Date: Wed, 9 Oct 85 11:08:38 pdt
From: Ed McCluskey <ejm@su-shasta.arpa>
Subject: Re:  petty cash
To: HANRAHAN@SU-SCORE.ARPA, csd@SU-SCORE.ARPA

there is a new policy that you are fined 200 megabucks for this

∂09-Oct-85  1349	DAVIES@SUMEX-AIM.ARPA 	Alliant -- Computer Systems Seminar Today!!    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  13:48:56 PDT
Date: Wed, 9 Oct 1985  13:49 PDT
Message-ID: <DAVIES.12149815420.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: Alliant -- Computer Systems Seminar Today!!
cc:   Davies@SUMEX-AIM.ARPA

Date : 10/09/1985	Event : Computer Systems Seminar
Day  : Wednesday	Person: Craig Mundie
Time : 4:15		From  : Alliant Computer Systems
Place: Terman		Title : Parallel Processing with Vector Engines and
       Auditorium		Auto-Decomposing Compilers

According to the abstract, the architecture and system software of the
new Alliant multiprocessor will be described.

	-- Byron

∂09-Oct-85  1648	admin@ucbcogsci.Berkeley.EDU 	Cognitive Science Seminar--Oct. 15 (Ron Kaplan, Xerox PARC & Stanford)
Received: from UCB-VAX.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  16:48:20 PDT
Received: by UCB-VAX.ARPA (5.28/5.11)
	id AA05747; Wed, 9 Oct 85 16:47:54 PDT
Received: by ucbcogsci (5.28/5.12)
	id AA16122; Wed, 9 Oct 85 16:48:25 PDT
Date: Wed, 9 Oct 85 16:48:25 PDT
From: admin@ucbcogsci.Berkeley.EDU (Cognitive Science Program)
Message-Id: <8510092348.AA16122@ucbcogsci>
To: cogsci-friends@ucbcogsci.Berkeley.EDU
Subject: Cognitive Science Seminar--Oct. 15 (Ron Kaplan, Xerox PARC & Stanford)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                    Cognitive Science Seminar -- IDS 237A

        TIME:             Tuesday, October 15, 11:00 - 12:30
        PLACE:            240 Bechtel Engineering Center
        DISCUSSION:       12:30 - 1:30 in 200 Building T-4

        SPEAKER:          Ronald M. Kaplan,
                          Xerox Palo Alto Research Center  and  Center
                          for  the  Study of Language and Information,
                          Stanford University

        TITLE:            ``Interactive Modularity''

        Comprehensible  scientific  explanations  for   most   complex
        natural  phenomena  are  modular  in character.  Phenomena are
        explained in terms of the operation of separate  and  indepen-
        dent  components, with relatively minor interactions.  Modular
        accounts of complex cognitive phenomena, such as language pro-
        cessing,  have  also  been proposed, with distinctions between
        phonological, syntactic, semantic, and pragmatic modules,  for
        example,  and  with  distinctions  among  various rules within
        modules.  But these modular accounts  seem  incompatible  with
        the   commonplace  observations  of  substantial  interactions
        across component boundaries: semantic and  pragmatic  factors,
        for  instance,  can  be shown to operate even before the first
        couple of phonemes in an utterance have been identified.

             In this talk I consider several  methods  of  reconciling
        modular descriptions in service of scientific explanation with
        the apparent  interactivity  of  on-line  behavior.   Run-time
        methods  utilize  interpreters that allow on-line interleaving
        of operations from different modules, perhaps including  addi-
        tional  "scheduling"  components  for  controlling  the cross-
        module flow of information.  But depending on their mathemati-
        cal properties, modular specifications may also be transformed
        by off-line, compile-time operations into  new  specifications
        that  directly  represent  all  possible cross-module interac-
        tions.  Such compilation techniques allow for run-time  elimi-
        nation  of  module  boundaries  and  of intermediate levels of
        representation.  I will illustrate these techniques with exam-
        ples  involving  certain  classes of phonological rule systems
        and structural correspondences in Lexical-Functional Grammar.
        --------------------------------------------------------------
        UPCOMING TALKS

        Oct 22:     Lotfi Zadeh, Computer Science, UCB
        Oct 29:     Mardi Horowitz, Psychiatry, UCSF
        Nov 5:      Edward Zalta, CSLI, Stanford
        Nov 12:     Robert Wilensky, Computer Science, UCB
        Nov 19:     Richard Alterman, Computer Science, UCB
        Nov 26:     Eve Clark, Linguistics, Stanford
        Dec 3:      Bernard Baars, Langley Porter, UCSF
        --------------------------------------------------------------
     

∂09-Oct-85  1656	@SU-CSLI.ARPA:admin@ucbcogsci.Berkeley.EDU 	Cognitive Science Seminar--Oct. 15 (Ron Kaplan, Xerox PARC & Stanford) 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  16:50:03 PDT
Received: from UCB-VAX.ARPA by SU-CSLI.ARPA with TCP; Wed 9 Oct 85 16:46:31-PDT
Received: by UCB-VAX.ARPA (5.28/5.11)
	id AA05747; Wed, 9 Oct 85 16:47:54 PDT
Received: by ucbcogsci (5.28/5.12)
	id AA16122; Wed, 9 Oct 85 16:48:25 PDT
Date: Wed, 9 Oct 85 16:48:25 PDT
From: admin@ucbcogsci.Berkeley.EDU (Cognitive Science Program)
Message-Id: <8510092348.AA16122@ucbcogsci>
To: cogsci-friends@ucbcogsci.Berkeley.EDU
Subject: Cognitive Science Seminar--Oct. 15 (Ron Kaplan, Xerox PARC & Stanford)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                    Cognitive Science Seminar -- IDS 237A

        TIME:             Tuesday, October 15, 11:00 - 12:30
        PLACE:            240 Bechtel Engineering Center
        DISCUSSION:       12:30 - 1:30 in 200 Building T-4

        SPEAKER:          Ronald M. Kaplan,
                          Xerox Palo Alto Research Center  and  Center
                          for  the  Study of Language and Information,
                          Stanford University

        TITLE:            ``Interactive Modularity''

        Comprehensible  scientific  explanations  for   most   complex
        natural  phenomena  are  modular  in character.  Phenomena are
        explained in terms of the operation of separate  and  indepen-
        dent  components, with relatively minor interactions.  Modular
        accounts of complex cognitive phenomena, such as language pro-
        cessing,  have  also  been proposed, with distinctions between
        phonological, syntactic, semantic, and pragmatic modules,  for
        example,  and  with  distinctions  among  various rules within
        modules.  But these modular accounts  seem  incompatible  with
        the   commonplace  observations  of  substantial  interactions
        across component boundaries: semantic and  pragmatic  factors,
        for  instance,  can  be shown to operate even before the first
        couple of phonemes in an utterance have been identified.

             In this talk I consider several  methods  of  reconciling
        modular descriptions in service of scientific explanation with
        the apparent  interactivity  of  on-line  behavior.   Run-time
        methods  utilize  interpreters that allow on-line interleaving
        of operations from different modules, perhaps including  addi-
        tional  "scheduling"  components  for  controlling  the cross-
        module flow of information.  But depending on their mathemati-
        cal properties, modular specifications may also be transformed
        by off-line, compile-time operations into  new  specifications
        that  directly  represent  all  possible cross-module interac-
        tions.  Such compilation techniques allow for run-time  elimi-
        nation  of  module  boundaries  and  of intermediate levels of
        representation.  I will illustrate these techniques with exam-
        ples  involving  certain  classes of phonological rule systems
        and structural correspondences in Lexical-Functional Grammar.
        --------------------------------------------------------------
        UPCOMING TALKS

        Oct 22:     Lotfi Zadeh, Computer Science, UCB
        Oct 29:     Mardi Horowitz, Psychiatry, UCSF
        Nov 5:      Edward Zalta, CSLI, Stanford
        Nov 12:     Robert Wilensky, Computer Science, UCB
        Nov 19:     Richard Alterman, Computer Science, UCB
        Nov 26:     Eve Clark, Linguistics, Stanford
        Dec 3:      Bernard Baars, Langley Porter, UCSF
        --------------------------------------------------------------
     

∂09-Oct-85  1703	EMMA@SU-CSLI.ARPA 	Newsletter October 10, No. 49  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  17:03:09 PDT
Date: Wed 9 Oct 85 16:51:08-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter October 10, No. 49
To: friends@SU-CSLI.ARPA
Tel:  497-3479



                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
October 10, 1985                Stanford                       Vol. 2, No. 49
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, October 10, 1985

   12 noon		TINLunch
     Ventura Hall       ``Artificial Intelligence Meets Natural Stupidity''
     Conference Room    by Drew McDermott
			Discussion led by Roland Hausser, U. of Munich
			(Abstract on page 1)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	``Ontology and Intensionality''
     Room G-19		Edward Zalta, CSLI
			Discussion led by John Perry

   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *NEXT* THURSDAY, October 17, 1985

   12 noon		TINLunch
     Ventura Hall       ``Economy of Speech Gestures''
     Conference Room    by Bjorn Lindblom (who will be present)
			Discussion led by Bill Poser
			(Abstract on page 2)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	``On the Notion of `Logophoricity' ''
     Room G-19		Peter Sells, CSLI
			(Abstract on page 2)
			
   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←
                    ABSTRACT FOR THIS WEEK'S TINLUNCH
             Artificial Intelligence Meets Natural Stupidity

   McDermott discusses three `mistakes', or rather bad habits, which are
   frequent in A.I. work.  He speaks from his own experience and cites
   several illuminating and amusing examples from the literature. In this
   TINLunch I will be discussing his thoughts on treating reference in
   A.I., which are discussed in the section entitled `unnatural
   language'.						--Roland Hausser

!
Page 2                     CSLI Newsletter                   October 10, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S TINLUNCH
                       Economy of Speech Gestures

       This paper discusses a functionalist approach to phonetics and
   phonology in which the properties of phonological systems are to be
   deduced from biological and social factors rather than from from
   axioms governing a language-particular formal system.	--Bill Poser
                              ←←←←←←←←←←←←
                    ABSTRACT FOR NEXT WEEK'S SEMINAR
                    On the Notion of `Logophoricity'

      The notion of `logophoricity' was introduced in studies of African
   languages in which a morphologically distinct `logophoric' pronoun has
   a distribution distinct from other pronouns, used with predicates of
   communication and consciousness.  More recently, this notion has been
   used in accounts of anaphora with non-clause-bounded reflexive
   pronouns, as are found in the Scandinavian languages, and Japanese.
   Such analyses propose a feature [+log] which is supposed to be
   specified on certain NPs by certain predicates.  I will present the
   beginnings of a formal construction of the notion of `logophoricity'
   using the Discourse Representation Structures framework developed by
   Hans Kamp.  I will propose that there is no such thing as
   logophoricity per se, but rather that it stems out of the interaction
   of two more primitive notions: the person with respect to whose
   consciousness (or `SELF') the report is made, and the person from
   whose point-of-view the report is made (the `PIVOT').  I will show how
   this system extends to certain facts (from Japanese) which are not
   analyzable with the simple feature [+log], and how it enables one to
   characterize cross-linguistic variation in what counts for
   `logophoricity'.				--Peter Sells
                              ←←←←←←←←←←←←
                       ENVIRONMENTS GROUP MEETING
           Monday, October 14, noon, Ventura Trailer Classroom

      David Levy (Xerox PARC and CSLI) will continue to describe his work
   on a theoretical foundation for document preparation environments.
   Specifically, he will describe in some detail the theory of marking
   itself, and its relevance to various computer systems.  We will
   discuss some points that came up in questions, such as the relation of
   ``indirect marking'' to different kinds of tools, the contrast between
   a psychological theory (how people think when they use a system) and
   an ontological account (of the basic objects, actions, and
   relationships that are available for them to work with), and the
   problems of multiple levels of representation (e.g., a macro command
   stands for a sequence of ``characters'' which in turn represent
   various ``figures'', etc.).
      See the summary of the meeting on October 7 (later in this
   newsletter) for more information.
                              ←←←←←←←←←←←←
                         LINGUISTICS COLLOQUIUM
                   ``Underspecification and Opacity''
                        Douglas Pulleyblank, USC
           Tuesday, October 15, Bldg. 200, Rm. 217, 3:15 p.m.

!
Page 3                     CSLI Newsletter                  October 10, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                              LOGIC SEMINAR
     ``Computability of Standard Processes in Analysis and Physics''
                 Marian Pour-El, University of Minnesota
                    Monday, October 14, Noon to 1:15
                        Ventura Hall Seminar Room

  Note the change of time and place.
      The regular meeting time of this seminar has been changed to
   Friday, noon.  We will meet alternate weeks beginning Friday, October
   25.							--Sol Feferman
                              ←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                 Summary of the meeting on September 26

      Farrell Ackerman gave a talk entitled ``Brackets and Branches:
   Phrasal Verbs.'' Assuming a provisional (and, perhaps, traditional)
   definition of phrasal verbs as morpholexically composed entities whose
   constitutive pieces exhibit syntactic independence, the discussion
   focused on the syntactic and lexical aspects of these Janus-like
   elements.
      From a syntactic perspective the interaction of phrasal verbs and
   rule of V(erb)-movement in, e.g., Vata (a Kru language analyzed in
   Koopman 1983) was discussed. V-movement in Vata is one particular case
   of V-movement motivated along similar lines, i.e., hypothesizing a V
   final d-structure representation, for the Germanic languages German
   and Dutch.  Evidence of similar syntactic discontinuities between
   particles (called `preverbs') and associated verb stems was given for
   the Ugric language Hungarian.  On the other hand, it was suggested
   that in this instance it is not the V but the particle which `moves.'
   After a presentation of the preverb-verb sequence possibilities in
   Hungarian discussion turned to the lexical aspects of preverb-verb
   collocations.
      From a lexical perspective the set of Hungarian preverbs can be,
   roughly, divided into two groups: prefixes and arguments.  The
   prefixes (minus a class of intriguing exceptions which were not
   discussed ) are categorially indeterminate and do not exhibit
   inflectional morphology indicating any relation of the prefix with the
   verb. Arguments, in contrast, are categorially determinable (in fact,
   are typically instantiated by and restricted to appear as a major
   lexical category) and bear inflectional morphology indicating their
   grammatical relation to the verb.  The combination of prefix + verb
   was hypothesized to be a type of verb derivation via prefixation while
   argument + verb was regarded as a type of lexical compounding.
   Evidence for the lexical nature of these phrasal verbs was taken to be
   their ability to serve as input for further derivational processes
   such as nominalization and adjectivilization.
      The assumption that phrasal verbs are lexical compositions leads to
   problems for the so-called Lexical Integrity Hypothesis, the procedure
   of Bracket Erasure in Lexical Phonology and, in general, leads to what
   have become know as `Bracketing Paradoxes.' It was proposed (following
   Simpson 1983 and Komlosy and Ackerman 1983) that there is a process of
   `bracket retention' restricted to the domain of predicate formation
   which accounts for the main difference, i.e., the syntactic
   separability of preverbs, in the behavior of preverbs in numerous
   languages.
!
Page 4                     CSLI Newsletter                  October 10, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                  SUMMARY OF ENVIRONMENTS GROUP MEETING
                           September 30, 1985

      At the first meeting of the environments group we set out the
   general directions for our discussions.  We identified some major
   dimensions along which to compare and examine environments and made an
   initial list of examples that might be presented.  This list is very
   sketchy---the random result of what happened to come up in
   conversation.  We are eager for further details and suggestions
   (either systems for general consideration, or about which specific
   people would like to talk):

   Programming environments: Interlisp, Smalltalk, Cedar, [all 3 Xerox],
      (Linton) [Berkeley/Stanford], Gandalf [CMU], Mentor [INRIA],
      ZetaLisp [Symbolics], Kee [Intellicorp], HPRL, HPLisp [last 2
      Hewlett-Packard]

   Grammar development environments: LFG [CSLI], HPSG [HP], BLT [CSLI],

   Specification environments: Aleph [CSLI], (Balzer)[ISI]

   Language development environments: MUIR [CSLI]

   Document preparation environments: (Levy) [CSLI], Notecards [Xerox]

   Data access and manipulation environments: 

   Mathematical and logical deduction environments: MACSYMA [MIT], FOL
      [Stanford]

      There is a variety of application areas not as central to CSLI
   concerns, but in which environments are built.  These include VLSI
   design, CAD/CAM, image manipulation, mail systems, etc. In addition,
   most operating systems take on the functions of an environment, either
   for use outside of applications programs or as a base within them.
   So-called ``intelligent agents'' are one attempt to provide a uniform
   environment for a particular user interacting with multiple systems.
      For each kind of environment there are specific problems dealing
   with the particular structures being worked with (programs, proofs,
   grammars, formatted documents, etc.).  There is also a framework of
   common problems having to do with the basic structure of items being
   manipulated (text, trees, databases, etc.), their representation on a
   screen or hardcopy, interfaces for operating through that
   representation, storage on one or more devices, consistency between
   aspects (e.g., source and compiled code, specifications and proofs),
   change over time (versions, releases, etc.), coordination of access
   among a group, etc.
      Our plan is to address the basic conceptual issues by looking at
   one particular environment or group of related environments in each
   session.  
!
Page 5                     CSLI Newsletter                  October 10, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                  SUMMARY OF ENVIRONMENTS GROUP MEETING
                             October 7, 1985

      David Levy gave an overview of his work on a theoretical basis for
   document preparation environments.  He demonstrated the problems with
   existing ``marking environments'' which combine conflicting approaches
   to text layout, drawing, and window placement.  The failure to
   generalize the common elements in all of these leads to greater
   complexity and to blind spots that create difficulty in maintaining,
   documenting, and using such systems.  Many of the relevant issues
   apply to older marking technologies, but the computer has two novel
   properties that demand a clear and explicit theory. First, marking is
   indirect---the linkage between human physical action and what appears
   on the screen (or paper) is mediated by linguistic or quasi-linguistic
   commands.  Second, there is a clear distinction between the surface
   presentation (what you see) and the internal representation (its
   underlying structure).  The computer, unlike earlier forms, lets you
   manipulate the underlying structure directly, with possibly complex
   and distributed consequences to the surface presentation.
      He then showed how we might begin to develop a theory of marking
   with a coherent ontological basis.  For example, we need to look at
   something as mundane as the ``carriage return'' as having distinct and
   sometimes confused aspects: it is a character (in the standard
   representation), it denotes an area of non-marked space on a page, it
   indicates a possible place to split a line in normal formatting, etc.
   By carefully delineating the concepts involved in these different
   aspects, we can produce systems that are simpler, easier to
   understand, and more amenable to generalization.
                              ←←←←←←←←←←←←
                             LICS CONFERENCE

      A new conference, LICS, (an acronym for ``Logic in Computer
   Science'') will meet in Cambridge, Mass, June 16-18, 1986.  The topics
   to be covered include abstract data types, computer theorem proving
   and verification, concurrency, constructive proofs as programs, data
   base theory, foundations of logic programming, logic-based programming
   languages, logics of programs, knowledge and belief, semantics of
   programs, software specifications, type theory, etc.  For a local copy
   of the full call for papers, contact Jon Barwise (Barwise@CSLI) or
   Joseph Goguen (Goguen@SRI-AI), members of the LICS Organizing
   Committee.
!
Page 6                     CSLI Newsletter                 October 10, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
            COMMON SENSE AND NON-MONOTONIC REASONING SEMINARS
            Organized by John McCarthy and Vladimir Lifschitz
               Computer Science Dept., Stanford University

      A series of seminars on Common Sense and Non-monotonic reasoning
   will explore the problem of formalizing commonsense knowledge and
   reasoning, with the emphasis on their non-monotonic aspects.
      It is important to be able to formalize reasoning about physical
   objects and mental attitudes, about events and actions on the basis of
   predicate logic, as it can be done with reasoning about numbers,
   figures, sets and probabilities.  Such formalizations may lead to the
   creation of AI systems which can use logic to operate with general
   facts, which can deduce consequences from what they know and what they
   are told and determine in this way what actions should be taken.
      Attempts to formalize commonsense knowledge have been so far only
   partially successful. One major difficulty is that commonsense
   reasoning often appears to be non-monotonic, in the sense that getting
   additional information may force us to retract some of the conclusions
   made before.  This is in sharp contrast to what happens in
   mathematics, where adding new axioms to a theory can only make the set
   of theorems bigger.
      Circumscription, a transformation of logical formulas proposed by
   John McCarthy, makes it possible to formalize non-monotonic reasoning
   in classical predicate logic. A circumscriptive theory involves, in
   addition to an axiom set, the description of a circumscription to be
   applied to the axioms. Our goal is to investigate how commonsense
   knowledge can be represented in the form of circumscriptive theories.
      John McCarthy will begin the seminar by discussing some of the
   problems that have arisen in using abnormality to formalize common
   sense knowledge about the effects of actions using circumscription.
   His paper Applications of Circumscription to Formalizing Common Sense
   Knowledge is available from Rutie Adler 358MJH.  This paper was given
   in the Non-monotonic Workshop, and the present version, which is to be
   published in Artificial Intelligence, is not greatly different. The
   problems in question relate to trying to use the formalism of that
   paper.
      The seminar will replace the circumscription seminar we had last
   year.  If you were on the mailing list for that seminar then you will
   be automatically included in the new mailing list. If you would like
   to be added to the mailing list (or removed from it) send a message to
   Vladimir Lifschitz (VAL@SAIL).

      The first meeting is in 252MJH on Wednesday, October 30, at 2pm.

-------

∂09-Oct-85  1705	AAAI-OFFICE@SUMEX-AIM.ARPA 	Cog Science Society Proposal    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  17:05:23 PDT
Date: Wed 9 Oct 85 17:01:08-PDT
From: AAAI <AAAI-OFFICE@SUMEX-AIM.ARPA>
Subject: Cog Science Society Proposal
To: officers: ;
cc: aaai-OFFICE@SUMEX-AIM.ARPA
Telephone: (415) 328-3123
Postal-Address: 445 Burgess Drive, Menlo Park, CA 94025
Message-ID: <12149850319.12.AAAI-OFFICE@SUMEX-AIM.ARPA>


My response to the the Cognitive Science Society Proposal is that the
office can do it with some reservations. Although the Cognitive
Science Society has approximately 700 members (which is less than 7%
of our membership of ~9,750 members) we foresee that the initial
start-up will take approximately 6 working days of our time to update
their files (only after they have rewrite their membership management
programs and convert their database files into MS-DOS DBase II or III
file format; this task will take some time and effort, which they will
be financially responsible for).  We then anticipate once a week (8
hours) maintenance of their database.  If we were to establish a
formal arrangement (as opposed to a trial run) we would be expected and
responsible for the opening of their mail and the depositing of dues
into a local Menlo Park account; we can expect to spend about 16 hours
a month on that task.  In addition, we will need to spend some time
setting up the account and cost center with our accountant.

They have also asked us to duplicate and mail 6 mailings a year; we
will have an outside mailing house perform that task and direct bill
them.  I intend to charge all our direct labor charges to the
Cognitive Science Society on a monthly basis.  They have also asked us
to send and count their election ballots; I personally think that the
Secretary-Treasurer of their organization should be responsible for
the counting of the ballots, not the AAAI.

This office's principal responsibilities are to implement the services
and programs of the AAAI.  When the conference months arrive
(March-August), we spend all our time on the conference and the
maintenance of the membership database (heaviest renewal months are
July-Sept).  It would appear that additional staffing would be
required to keep abreast of the Cognitive Science Society's needs
during this period, because our staff will be working on the
conference and membership exclusively.  The Cognitive Science Society
should not expect the best of service during this period.

During the rest of the year, I do believe we will be able maintain
their database with the caveat that the database remains relatively
small.  Our membership is growing at a rate of 200 new members per
month with an overall attrition rate of less than 8% per year; we now
have 9,750 members.  Our clerical staff is just able to maintain the
current sized database.

My recommendation is to try it for one year and see how it taxes the
staff.  However, I do strongly believe that cooperation with the Cog
Sci Society must go well beyond their current proposal.  Other forms
of scientific cooperation and collaboration (eg. Hart's suggestion for
jointly sponsored workshops) should be part and parcel of our agreement
with Cog SCi in the future.

--- Claudia

-------

∂09-Oct-85  1915	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	Call for Papers, Spring'86 CUG   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Oct 85  19:15:28 PDT
Received: from ames-vmsb.ARPA by SU-SCORE.ARPA with TCP; Wed 9 Oct 85 19:10:40-PDT
Date: 9 Oct 85 17:56:00 PST
From: welch@ames-vmsb.ARPA
Subject: Call for Papers, Spring'86 CUG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA


The next meeting of the Cray User's Group has begun seeking papers.
The theme of the next meeting is Unix.  Papers need not emphasize the
theme but can also cover other standard Cray topics such as COS, CTSS,
CFT, and day to day operations as well.  The meeting is being sponsered
by Boeing Computer Services.  All papers should come from people working
at Cray affiliated sites.  Sites not already possessing Crays but
considering purchase of a Cray may contact David Lexton at the address
below for special approval.  The keynote speaker will be a prominent member
of the Unix community currently involved with Cray.

--eugene miya
  NASA Ames Research Center
  eugene@ames-nas.ARPA
  {hplabs,ihnp4,hao,nsc,menlo70,decwrl}!ames!amelia!eugene
  Software Tools Special Interest Committee, CUG
  Session Chair on the C Programming Language, Spring'86 CUG
     {send me C related papers for direct review}

------------------------ Please tear here ---------------------------

CUG
CALL for PAPERS
SEATTLE, WA MAY 5-8 1986
THEME: UNIX

NAME:

ORGANIZATION:

ADDRESS:



TELEPHONE: (    )

TITLE OF PAPER:

TWO OR THREE SENTENCE ABSTRACT:




EQUIPMENT REQUIRED OTHER THAN 35MM SLIDE PROJECTOR OR OVERHEAD PROJECTOR:


SUGGESTED SESSION:

GENERAL SESSION		←←←		LANGUAGES		←←←
I/O			←←←		LIBRARIES		←←←
FRONT ENDS		←←←		MULTITASKING		←←←
NETWORKING		←←←		OPTIMIZATION		←←←
OPERATIONS		←←←		PERFORMANCE EVALUATION	←←←
OPERATING SYSTEMS	←←←		APPLICATIONS		←←←
GRAPHICS		←←←		DATA MANAGEMENT		←←←

OTHER: ←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

RETURN BY JANUARY 1, 1986 to:

DAVID LEXTON
University of London Computer Centre
20 Guilford Stree
London  WC1N  1DZ England
------

∂09-Oct-85  2015	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ucbernie.Berkeley.EDU    
Received: from [36.36.0.196] by SU-AI.ARPA with TCP; 9 Oct 85  20:15:20 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 9 Oct 85 20:12:14-PDT
Received: from UCB-VAX.ARPA by SU-SCORE.ARPA with TCP; Wed 9 Oct 85 20:12:06-PDT
Received: by UCB-VAX.ARPA (5.28/5.11)
	id AA01898; Wed, 9 Oct 85 14:06:41 PDT
Received: by ucbernie.ARPA (5.28/5.6)
	id AA06900; Wed, 9 Oct 85 14:05:09 PDT
Date: Wed, 9 Oct 85 14:05:09 PDT
From: karp@ucbernie.Berkeley.EDU (Richard Karp)
Message-Id: <8510092105.AA06900@ucbernie.ARPA>
To: aflb.all@su-score.arpa, allmsgs@ernie.Berkeley.EDU,
        csfac@ernie.Berkeley.EDU, theory-b@ernie.Berkeley.EDU,
        theory@ernie.Berkeley.EDU


MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley  (415)642-0143

WORKSHOP ON COMPUTATION IN ALGEBRA AND NUMBER THEORY
October 14-18, 1985
All lectures will take place in the MSRI Lecture Hall.

Monday, October 14

Hendrik Lenstra, "Factoring Integers With Elliptic Curves"   9:30

Carl Pomerance, "Integer Factoring Algorithms"  11:00

Don Coppersmith, "Solution of Sparse Systems of Linear Equations over Finite
Fields"                                          2:00

Jeffrey Lagarias, "Multidimensional Continued Fraction Algorithms"  4:00

Tuesday, October 15

Joachim von zur Gathen, "Deterministic Factorization of Polynomials"  9:00

Erich Kaltofen, "Multivariate Polynomial Factorization: from Curves to
Vandermonde Determinants"                  10:30

George Andrews, "Scratchpad and Rogers-Ramanujan Type Identities"   2:00

Andrew Odlyzko, "On the Complexity of the Riemann Zeta Function"

Wednesday, October 16

Susan Landau, "Polynomial Time Algorithms for Solvable and Simple Galois 
Groups"                                               9:00

John McKay, "The Computation of Galois Groups"      10:30

Gary Miller, "Permutation Groups and their Relation to Graph
Isomorphisms"                                        2:00

Eugene Luks, "Polynomial Time Computation in Permutation Groups"  4:00

Thursday, October 17

Barry Trager, "Integral Bases and Nonsingular Models of Curves"  9:00

Patrizia Gianni, "Ideal Theoretic Operations Using Grobner Bases"  10:30

David Bayer, "On the Geometry of Computing Syzygies"  2:00

Michael Stillman, "Castelnuovo Regularity and the Complexity of Computing
Syzygies"                                             4:00

Friday, October 18

Stephen Wolfram, Cellular Automata"  9:00

Laszlo Lovasz, "Algorithmic Problems on Algebraic Matroids"  10:30

∂10-Oct-85  0904	LIBRARY@SU-SCORE.ARPA 	Socrates: Are people having problems when they attempt to telnet to Socrates?
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Oct 85  09:04:43 PDT
Date: Thu 10 Oct 85 09:01:54-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates: Are people having problems when they attempt to telnet to Socrates?
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12150025222.27.LIBRARY@SU-SCORE.ARPA>

Someone has brought to my attention that when they telnet to Forsythe and then
try to logon to Socrates they are immediately logged off.  Are others having
the same problem?  If so, send me a message with subject header Socrates Logon
Problem, identify the computer you are attempting to telnet from and indicate
briefly exactly what is happening, or not happening.

Harry Llull
-------

∂10-Oct-85  1107	DELAGI@SUMEX-AIM.ARPA 	"Volunteers"
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 10 Oct 85  11:07:14 PDT
Date: Thu 10 Oct 85 11:07:54-PDT
From: Bruce Delagi <DELAGI@SUMEX-AIM.ARPA>
Subject: "Volunteers"
To: aap@SUMEX-AIM.ARPA
Message-ID: <12150048159.66.DELAGI@SUMEX-AIM.ARPA>


As mentioned yesterday, there's about 1.5 sessions worth of formal lecture
for the three meetings on parallel KBS hardware systems architecture.  The 
remaining 1.5 sessions will be devoted to presentations by seminar attendees.

There are five presentations needed.  The idea is to use the lecture material 
to assess some alternative machines in terms of the design choices they
represent.  For October 21st, the representative machines are [1] the HEP,
[2] the Cosmic Cube, and [3] the RPP-3.  For October 28th, two presentations 
will force a contrast between [4] dataflow machines and languages and [5]
parallelizing compilers for conventional languages.  I think this can be best
done by the presenters taking the proponent's position [respectively for
Arvind/Dennis/Gurd and Gajski/Kuck] using the arguments developed in the
papers for the 28th and the lectures and discussions on the 14th and 21st.

Presentations on the HEP, the Cosmic Cube, and the RPP-3 should be pretty
short: say 10 minutes or so.  The data flow / dusty deck debate leaders
should prepare a statement of position [10 minutes] and a rebuttal [shorter]
exposing the issues as clearly as possible for a discussion to follow.

Volunteers please...................../bruce
-------

∂10-Oct-85  1357	LIBRARY@SU-SCORE.ARPA 	[The Mailer Daemon <Mailer@SU-SCORE.ARPA>: Message of 10-Oct-85 13:49:31]    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Oct 85  13:56:21 PDT
Date: Thu 10 Oct 85 13:51:18-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: [The Mailer Daemon <Mailer@SU-SCORE.ARPA>: Message of 10-Oct-85 13:49:31]
To: : ;
Message-ID: <12150077903.27.LIBRARY@SU-SCORE.ARPA>

Date: Thu 10 Oct 85 13:49:34-PDT
From: The Mailer Daemon <Mailer@SU-SCORE.ARPA>
To: LIBRARY@SU-SCORE.ARPA
Subject: Message of 10-Oct-85 13:49:31

Message failed for the following:
facutly@SU-SCORE.ARPA.#Internet: No such mailbox
	    ------------
Date: Thu 10 Oct 85 13:49:31-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books
To: su-bboards@SU-SCORE.ARPA
cc: facutly@SU-SCORE.ARPA
Message-ID: <12150077580.27.LIBRARY@SU-SCORE.ARPA>

A Hierarchical Associative Processing System. by Heinrich J. Stuttgen.
Lecture Notes In Computer Science. no. 195  TK7895.M4S78 1985.

Automata On Infinite Words. Ecole de Printemps d'Informatique Theorique
Le Mont Dore, May 1984. ed. by M. Nivat and D. Perrin.  
QA267.E26 1984.

Applications of Artificial Intelligence II. Proceedings of SPIE. ed.
John Gilmore. Q334.A663.

User Interface Management Systems. Eurographic Serminars. ed. Gunther
Pfaff.  QA76.9.I58W67 1983.

Introducing Artificial Intelligence. by G.L. Simons. National Computing
Centre Publication.(8512444)

Programmierung Paralleler Prozesse. by Bachmann. Technische Universitat
Dresden. Weiterbildungszentrum Fur Mathematische Kybernetik Und
Rechentechnik/Informationsverarbeitung.  (8508714)

Discrete Mathematics And Applied Modern Algebra. by Henry Laufer.
QA162.L38 1984.

Topics In The Theory Of Computation. Annals of Discrete Mathamatics. No.24.
by Marek Karpinski and Jan Van Leeuwen.  QA267.I56 1983.

STATCOMP 83 Papers. A Symposium on Statistical Packages, Microcomputers and
Database Management.  (8508548)

TIMSAC-88 Part 1 and Part 2.  Computer Science Monographs. A Publication
of The Institute of Statistical Mathematics. (8517294)

H. Llull
-------
-------
-------

∂10-Oct-85  1408	CAROL@SU-CSLI.ARPA 	New office
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 10 Oct 85  14:07:54 PDT
Date: Thu 10 Oct 85 14:02:14-PDT
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: New office
To: Folks@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA

Hi-

I have just moved from F-2 to luxurious quarters in Ventura 28
(second floor behind the stairs).  So far there's no phone, but
Jamie has kindly offered to share his - 7-3170.  

-Carol
-------

∂11-Oct-85  0042	NEUMANN@SRI-CSL.ARPA 	RISKS-1.21   
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 11 Oct 85  00:42:36 PDT
Date: Thu 10 Oct 85 23:27:02-PDT
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSL.ARPA>
Subject: RISKS-1.21
Sender: NEUMANN@SRI-CSL.ARPA
To: RISKS-LIST@SRI-CSL.ARPA

RISKS-LIST: RISKS-FORUM Digest  Thursday, 10 Oct 1985  Volume 1 : Issue 21

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Public Accountability (Jim Horning, Peter Neumann)
  The Titanic Effect (JAN Lee)
  Databases, Grades, etc. (Brian Borchers, Andy Mondore, Mark Day [twice],
    Alan Wexelblat, Ross McKenrick, Randy Parker)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

From: horning@decwrl.ARPA (Jim Horning)
Date: 10 Oct 1985 1244-PDT (Thursday)
To: RISKS@SRI-CSL
Subject: Public Accountability

It has now been more than a year since the ACM Council passed its
resolution on computer systems and risks to the public. Quoting from
RISKS-1.1:

  The second part of the resolution includes a list of technical questions
  that should be answered about each computer system.  This part states that:

    While it is not possible to eliminate computer-based systems failure
    entirely, we believe that it is possible to reduce risks to the public
    to reasonable levels.  To do so, system developers must better
    recognize and address the issues of reliability.  The public has the
    right to require that systems are installed only after proper steps
    have been taken to assure reasonable levels of reliability.

    The issues and questions concerning reliability that must be addressed
    include:

    1. What risks and questions concerning reliability are involved when
       the computer system fails?

    2. What is the reasonable and practical level of reliability to
       require of the system, and does the system meet this level?
  
    3. What techniques were used to estimate and verify the level of
       reliability?

    4. Were the estimators and verifiers independent of each other
       and of those with vested interests in the system?

In the intervening year, I am not aware that the developers of ANY
computer system have made public their answers to these four questions.
Readers of this forum will surely not leap to the conclusion that no
computer system presents risks to the public worthy of discussion.

I would like to start a discussion on how we can change this.
What can we do as professionals to make it the norm for system
developers to present risk assessments to the public?

  * What can we do to make it more attractive to present a risk assessment?

    - Explicitly invite developers of particular systems to publish draft
      assessments here in the RISKS Forum, with the promise of constructive
      feedback.

    - Inaugurate a section in CACM for the publication of refined risk
      assessments of systems of great public interest or importance.

    - Publish the risk assessments without refereeing. This gives the
      developers "first shot," gets the material out more quickly, and lowers
      the barrier to publication, without limiting the opportunity for public
      discussion and debate.

    - Encourage developers to also address John McCarthy's question:
    
        5. What are the risks inherent in the best available alternative
           to the system in question?

    - Encourage the exploration of legal steps that would make
      Risk Assessments as routine as Environmental Impact Statements.
      (This could be useful even if most of the former are as pro forma and
      unenlightening as most of the latter; they provide a starting point.)

  * What can we do to make it less attractive not to present a risk assessment?

    - First, make it very clear that we, as a profession, believe that it
      is incumbent on system developers to present their risk assessments,
      and that delay or refusal amounts to malpractice of a very high order.

    - Periodically publish (and publicize) the status of all outstanding
      invitations to present risk assessments.

    - Bring cases of persistent noncompliance to the attention of ACM
      Council for appropriate action.

I present these suggestions as a springboard for discussion, not as a
definitive program for action. I welcome suggestions for improvement.

Jim H.

------------------------------

Date: Thu 10 Oct 85 13:30:24-PDT
From: Peter G. Neumann <Neumann@SRI-CSL.ARPA>
Subject: Public Accountability
To: RISKS@SRI-CSL.ARPA
cc: Horning@DECWRL.DEC.COM

Perhaps we can learn something from some steps that have been taken by the
National Computer Security Center (although for security and not
reliability).  The NCSC (formerly the DoD CSC) has established a set of
criteria for trusted computer systems, in the form of a range of
increasingly stringent requirements.  They have evaluated various systems
against those criteria.  They have also explored what kinds of applications
should have which requirements.  To date, two systems have been accorded
high rankings:  (1) the highest existing rating [A1] to the Honeywell SCOMP
-- whose kernel design and trusted subsystems have undergone formal design
proofs demonstrating that the kernel specifications satisfy a security
condition of no adverse flow, and that the trusted subsystems satisfy other
relevant properties involving privilege, integrity, and functional
correctness; (2) a somewhat lower rating [B2] to Multics, which has not been
subjected to any formal analysis, but which satisfies certain of the
hardware and software-engineering criteria and which has withstood extensive
penetration attacks.  Other systems have also been evaluated, but given
lesser ratings -- implying greater potential security risks.  (The NCSC also
publishes an Evaluated Products List.)  The literature on this subject is
extensive, but I note a few items here on the Criteria and on the SCOMP
proofs.  Other than that, you can look at the IEEE Proceedings for the
Security and Privacy conferences each April and the National Computer
Security Conferences held at NBS roughly once a year (previously called the
DoD/NBS Computer Security Conference).

I of course need to add that the work on secure and trusted computing
systems is only a very small part of that addressing the potential "risks to
the public", which of course also involve reliability in various forms,
human safety, timely responsiveness, etc.  But my point is that some steps
in the right direction are actually being taken in the security community --
although those are still only small steps with respect to the overall
problem.  Nevertheless, something can be learned from the security work
in addressing the ACM goals that Jim has reminded us of once again.

So, let us try to address some of Jim's points specifically in the future.


REFERENCES:

  CRITERIA: Department of Defense Trusted Computer System Evaluation
  Criteria, CSC-STD-001-83.  Write to Sheila Brand, National Computer
  Security Center, Ft Geo G Meade MD 20755.  (SBrand@MIT-Multics)

  SCOMP KERNEL DESIGN PROOFS:  J.M. Silverman, Reflections on the Verification
  of the Security of an Operating System, Proceedings of the 9th ACM Symposium
  on Operating Systems Principles, October 1983, pp. 143-154.

  SCOMP TRUSTED SUBSYSTEM DESIGN PROOFS: T.C.V. Benzel and D.A. Tavilla,
  Trusted Software Verification: A Case Study, Proceedings of the 1985
  Symposium on Security and Privacy, IEEE Computer Society, Oakland CA, April
  1985, pp. 14-31.

(Note: In a proof of design consistency, the proof must show that a formal
specification satisfies a set of requirements, e.g., for security or fault
tolerance.  The difference between requirements and specifications in that
case is generally that the former tend to be simply-stated global
properties, while the latter tend to be detailed sets of constraints defined
[e.g.,] functionally on state transitions or algebraically on inputs and
outputs.)

----------------------------------------------------------------------

Date:     Tue, 8 Oct 85 16:26 EST
From:     Jan Lee <janlee%vpi.csnet@CSNET-RELAY.ARPA>
To:       RISKS@sri-csl.arpa
Subject:  The Titanic Effect

THE TITANIC EFFECT  (Source unknown):

    The severity with which a system fails is directly proportional
    to the intensity of the designer's belief that it cannot.

COROLLARY:

    The quantity and quality of built-in redundancy is directly
    proportional to the degree of concern about failure.


JAN

----------------------------------------------------------------------

Date: Wed, 9 Oct 85 09:43:20 EDT
From: Brian←Borchers%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
To: risks@sri-csl.arpa
Subject: Databases, Grades, etc. 

Here at RPI, The Registrar's database, as well as the Bursar's systems are
on the same machine that we use for academic work.  We've put a lot of
effort into making MTS secure...

   [By the way, some of this issue's discussions on the low risks of putting
    grades on student computers reflect overall a benevolent naivete about
    the system security risks.  I have not tried to comment on each message
    individually, but find them intriguing.  One could turn the students
    loose to see how many flaws they can find, perhaps as a part of a
    course: give them all F's in the database, and let them try to earn 
    their A's by breaking in!  On the other hand, you might find the last
    one who breaks changes some of the A's to F's!  It is dangerous to
    announce in public that you as an administrator believe you have made 
    unauthorized alterations difficult or impossible; knowing how badly
    flawed most of the systems are, that is a very large red flag to wave.
    But then you apparently need to be shown that you are wrong.  Audit trails 
    are great too, but watch out for the penetrated superuser interface.
    (This comment only touches the tip of the iceberg.  Judging from this
    issue's contributions, I imagine the discussion might run for a while.
    Let's get down to specific cases if we can, but not just students' 
    grades -- which are only an illustrative problem.)  PGN]  

------------------------------

Date: Wed, 9 Oct 85 11:36:54 EDT
From: Andy←Mondore%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
To: RISKS@SRI-CSL.ARPA
Subject:  Databases, Grades, etc. 

Hal Murray in RISKS-1.20 asks whether any schools are "brave/stupid enough"
to have students and grades on the same computer.  Well, that is the
situation here.  Administrators and students use the same mainframe.
Administrative here generally use several extra layers of security such as
extra passwords, allowing sign-ons to certain accounts only from specific
terminals, and logging all successful and unsuccessful sign-ons to our
accounts.  So far, we in the Registrar's Office have not detected any
unauthorized sign-ons (and we have never noticed any strange changes to
files) although we occasionally detect an unsuccessful sign-on attempt from
an unauthorized location.  Generally, changing passwords frequently and
using non-words and special characters for passwords seems to take care of
the unauthorized access problem.

One thing that Hal did not mention is security problems when students and
grades are on separate computers but on the same network, perhaps to share a
special device or certain files. It seems to me that there could be almost
as many potential security problems with this configuration as with my
configuration. Does anyone have experience with this type of configuration
and if so, what problems have you had?

------------------------------

Date: Wed 9 Oct 85 09:50:12-EDT
From: Mark S. Day <MDAY@MIT-XX.ARPA>
Subject:  Databases, Grades, etc. 
To: risks@SRI-CSL.ARPA

I tend to think that stories about crackers changing their grades should be
taken with a fairly large dose of salt.

I once had the occasion to interview a vice-chancellor at my undergraduate
university, and he explained the system there in some detail.  Basically,
they handled grades the way a bank handles money -- that is to say, there
were 'audit trails' and paper copies of everything.

Discrepancies between what was in the computer and what was in the paper
records would eventually get caught, even if there was a period of time when
the wrong grades would show up on an official transcript.  Correctly faking
ALL the records so as to escape an audit would have required a great deal of
knowledge (basically, it would have to be an inside job -- and they didn't
have students working in these offices).

--Mark

------------------------------

Date: Wed 9 Oct 85 11:39:54-PDT
From: Peter G. Neumann <Neumann@SRI-CSL.ARPA>
Subject: Databases, Grades, etc.
To: MDAY@MIT-XX.ARPA
cc: Neumann@SRI-CSL.ARPA

Mark, I find myself disbelieving some of what you say.  Manual comparisons
tend to get done (if at all) when the data is entered.  Auditing tends to
get slighted, since people often tend to assume the computer is right! (The
Santa Clara inmate who changed his release date did get caught, but perhaps
only because a guard overheard him bragging about how he was going to get
out early.)  Most vice-chancellors (and many administrators) would probably
tell you they are doing things right, and better than everyone else.  That
is the head-in-the-sand view of computing -- everything is just fine.  But
audit trails and paper copies of everything are not sufficient, even if they
are diligently used.  "Discrepancies ... would eventually get caught" is
certainly hopeful, at best.  Peter

------------------------------

Date: Wed 9 Oct 85 14:51:40-EDT
From: Mark S. Day <MDAY@MIT-XX.ARPA>
Subject: Databases, Grades, etc.
To: Neumann@SRI-CSL.ARPA

Peter,

I still maintain that the scenario of "student changes grades to A's and
lives happily ever after" seems dubious.  It seems to be one of these
apocryphal stories where everyone knows about someone who's done it, but
there's little evidence floating around about it (sort of like stories about
people putting cats in microwave ovens to dry them).

Your other comments are well-taken.  It certainly requires administrators,
etc., to have a perception of the problem; that was one reason I was
impressed by my interview with the vice-chancellor.  Several years ago,
crackers were not yet in the public eye, so I was pleasantly surprised to
find out that anyone cared about the overall integrity of the grades
database and talked about the need to regularly audit it.

--Mark

------------------------------

Date: Wed, 9 Oct 85 12:40:43 cdt
From: Alan Wexelblat <wex@mcc.ARPA>
To: risks%sri-csl@mcc
Subject: Databases, Grades, etc.

The University I attended had its database on a computer that student
courses used.  However, it was (is) a very well-kept secret.  It is hard
(under the OS of this system) to find out what people *might* log on.  Also,
it seems that hacker-types don't get the work-study jobs in the registrar's
office.  I'm not sure how long they can keep it up, but it's worked well for
at least 5 years.  (I only found out because I got to be good friends with
the facilities people, and one of them happened to mention it.)

I guess this just reinforces the belief that human beings can make a risky
or less-risky depending on how they use it.  --Alan Wexelblat  WEX@MCC.ARPA

------------------------------ 

Date:    Wed, 09 Oct 85 16:12:15 EDT
From:    Ross McKenrick  <CRMCK%BROWNVM.BITNET@WISCVM.ARPA>
To:      RISKS@SRI-CSL.ARPA
Subject: Databases, Grades, etc.

Students and grades can exist on the same machine.  Attempting to partition
the students' and administrations' computing environments is a crude form of
security which eliminates the possibility of providing students online
services such as class lists and registration changes.

We take a two-fold approach to database security at Brown University:
1) prevention and 2) detection.  We minimize our window of vulnerability
by making it nearly impossible to gain unauthorized access to the
programs which update grades, etc, and by making it nearly impossible
to change a grade (even when authorized) without generating audit
records, security log entries, and notifications slips for the professor.

We are not dealing with an EFT-type environment where "smash-and-grab" might
work.  A student is at Brown for four years, and his/her transcript is
maintained by the Registrar forever.  A student could theoretically change
his/her grade for a day or two *in the computer database* (which does not
mean that the grade, which is intangible, was actually changed).  However, a
system of checks and balances, built into and outside of the computer
system, would eventually result in discovery, correction, and punishment.

Suppose I could design a security system which was 90% (OK, 80%) secure.
Then, rather than spend more time making it more secure (point of
diminishing returns), I could spend my time on a recording/ reporting system
which was 80% secure.  Now, I finish by dovetailing the two systems to make
it very unlikely that a hacker could survive both levels.  Meanwhile, the
Registrar, who is rightfully suspicious of computer security anyway, devises
manual checks and balances which adds another level of security beyond the
computer entirely.

Wouldn't it simply be easier for the hacker to forge a grade change slip
from his/her professor?  Now, considering all of this, you've got certain
expulsion and prosecution under Rhode Island State Law hanging over your head.

Why would someone who's got all of this figured out have trouble passing
courses, anyway?

A collegue of mine likens a malicious hacker in a fairly-well-secured
computer environment to a bull in a china shop.  The china is replaceable,
the bull is dead meat.

Now comes the auto-theft-prevention-device philosophy: why pick on
a fairly-well-secured environment when there are so many unsecured
environments to fool with?

Security in computer-recordkeeping is a very serious subject.  But
you must keep it in perspective with the alternatives: manual record-
keeping, locked doors and desks, etc.

------------------------------

Date: Thu, 10 Oct 85 09:34 EDT
From: Randy Parker <PARKER@MIT-REAGAN.ARPA>
Subject: Databases, Grades, etc.
To: RISKS@SRI-CSL.ARPA

Regarding the subject of the risks of computer technology (in general)
and large databases (FBI's NCIC, TRW, NSA, tenant/landlord database in
California, and the Census), there is a reporter from the New York Times
who has done quite a bit of research into it.  His name is David
Burnham, and his results a few years back are published in his book "The
Rise of the Computer State" (paper/hardback).

The book is a call to arms to the millions who don't realize the serious
threat to our freedom that this poses.  Its shortcoming, as such, is
that it is mostly a quite comprehensive series of anecdotal reports on
various errors and abuses (incorrect warrants, Nixon abuse of phone
company records, abuse by politicians, etc) of large informations
stores.  However, if you need convincing that this is a problem, you
will find it in this book.

Burnham spoke here this past Tuesday and updated his list of "horror
stories", as he puts it.  One number he gave, if my notes are right, is
that an FBI audit of their NCIC showed that ~10% of the warrants proposed
by this system are somehow based on incomplete or invalid information.
He also mentioned proposals to increase the NCIC to include a White
Collar Crime Index that would specifically contain the names of SUPPOSED
white collar criminals and their ASSOCIATES.  As Burnham pointed out, if
they can't maintain the actual criminal information properly, what are
the consequences when they start accumulating hearsay and gossip?

A serious problem is that no one has a real interest in keeping this
information very correct, except for the falsely accused citizen, and,
as Burnham pointed out, he doesn't have any way to correct it.  I also
agree quite heartily with Matt Bishop that "the greatest risk is not
from the technological end but the human end."  The technology itself is
not to blame, but we need, especially as computer professionals whose
opinions in the matter would be seriously regarded, to recognize the
growing threat to our liberties and to act accordingly.

Randy Parker
MIT AI Lab
PARKER@MIT-REAGAN

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂11-Oct-85  0804	BRESNAN@SU-CSLI.ARPA 	announcement of MSDI presentations    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Oct 85  08:04:05 PDT
Date: Fri 11 Oct 85 07:59:49-PDT
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: announcement of MSDI presentations
To: folks@SU-CSLI.ARPA
cc: bresnan@SU-CSLI.ARPA

The working group on Morphology/Syntax/Discourse Interactions will
be giving a series of open presentations of the results of its
summer research project, which have been abstracted in the CSLI Newsletter.

On Thursday, October 24, at 10:00 a.m. in the Ventura Conference
Room.  Joan Bresnan will present Bresnan and Mchombo's "Subject,
Topic, and Agreement in Chichewa".  A written version of this
work will be available on Monday, October 21, at the Ventura
Receptionist's Desk.
-------

∂11-Oct-85  1315	PATASHNIK@SU-SUSHI.ARPA 	Speaker for October 17   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 11 Oct 85  13:15:03 PDT
Date: Fri 11 Oct 85 13:04:49-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Speaker for October 17
To: aflb.all@SU-SUSHI.ARPA

We still have no speaker for October 17 (this coming week).  Please
let me know if you want give a talk then.
	--Oren
-------

∂11-Oct-85  1502	LIBRARY@SU-SCORE.ARPA 	Socrates:  Workshops to be held in the Green Library
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 11 Oct 85  15:01:49 PDT
Date: Fri 11 Oct 85 14:58:36-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates:  Workshops to be held in the Green Library
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12150352299.8.LIBRARY@SU-SCORE.ARPA>

The Libraries will be offering instruction in the use of Socrates, the online
catalog, this Fall Quarter.  Provided will be a brief overview of Socrates,
how to search in Guided mode, and most of the session will concentrate
on using Socrates in Command mode. Handson practice, using Macintosh
computers, will be available.  The workshops will be limited in size,
and advance registration is recommended. Please register by calling the
General Reference Department at 497-1811 or at the Reference Desk in 
Green Library.

Dates: Oct. 16 Wed. 3-5pm Oct. 23 Wed. 9-11 am Oct. 29 Tues. 9-11 am
       Nov. 6 Wed. 3-5pm Nov. 12 Tues. 3-5pm Nov. 18 Mon 9-11 am
       Nov. 25 Mon. 3-5 pm

Sessions will be held in the McDermott Room, Green Library

-------

∂11-Oct-85  2217	BRESNAN@SU-CSLI.ARPA 	presentation by Peter Sells 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Oct 85  22:17:50 PDT
Date: Fri 11 Oct 85 22:15:10-PDT
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: presentation by Peter Sells
To: folks@SU-CSLI.ARPA

On October 17 Peter Sells will be giving a talk entitled "On the Notion
of `Logophoricity'" at CSLI.  This presentation is also part of the
Morphology/Syntax/Discourse Interactions series of presentations I
announced previously, and a written version of the paper will be
available upon request from the author by Monday October 21.
-------

∂12-Oct-85  0800	PATASHNIK@SU-SUSHI.ARPA 	AFLB Abstract  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 12 Oct 85  08:00:48 PDT
Date: Sat 12 Oct 85 07:58:25-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: AFLB Abstract
To: aflb.all@SU-SUSHI.ARPA

We have a speaker for the upcoming AFLB.  Here's the abstract:


17-Oct-85  :  Richard Beigel (Stanford)

		Sorting n objects with a k-sorter.

A k-sorter is an instruction (or black box or what-you-will) that
sorts k objects in constant time.  An interesting question is ``How
quickly can you sort n objects with a k-sorter?''  We will demonstrate
asymptotic bounds of nlogn/klogk and 4nlogn/klogk.  We will also show
how to sort in nlogn/(k-1) + O(1); this is of interest when k is small.
Finally we will discuss the particular case of sorting with a 3-sorter.

***** Time and place: October 17, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352 (Bldg 460).

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

For more information about future and past AFLB meetings and topics
are in the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂14-Oct-85  0755	@SUMEX-AIM.ARPA:HAUNGA@SU-SCORE.ARPA 	Title & Abstract for the Siglunch Friday 18th. 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 14 Oct 85  07:55:46 PDT
Received: from SU-SCORE.ARPA by SUMEX-AIM.ARPA with TCP; Mon 14 Oct 85 07:55:48-PDT
Date: Mon 14 Oct 85 07:52:23-PDT
From: Ana Haunga <HAUNGA@SU-SCORE.ARPA>
Subject: Title & Abstract for the Siglunch Friday 18th.
To: SIGLUNCH@SUMEX-AIM.ARPA
cc: haunga@SU-SCORE.ARPA
Message-ID: <12151061140.11.HAUNGA@SU-SCORE.ARPA>


Siglunch will be held at the Chemistry Gazebo at 12:05-1:00 p.m.

-------

Title: Probabilistic Interpretations for MYCIN's Certainty Factors

Abstract:

I will show that the original definition of certainty factors (CF's)
is inconsistent with the "defining desiderata" of the CF combination
functions.  I will then show that if this inconsistency is removed by
redefining CF's in terms of the desiderata then CF's have
probabilistic interpretations.  In other words, I will show that
certainty factors are nothing more than transformed probabilistic
quantities.  The construction of the interpretation provides insights
into the assumptions made when propagating CF's through an inference
net.  For example, it can be shown that all evidence which bears
directly on a hypothesis must be conditionally independent on the
hypothesis and its negation.  After presenting the interpretations,
I will discuss several ramifications of the correspondence between
CF's and probabilities.
-------
-------

∂14-Oct-85  0851	RICHARDSON@SU-SCORE.ARPA 	Tuesday CSD Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 14 Oct 85  08:51:11 PDT
Date: Mon 14 Oct 85 08:44:08-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Tuesday CSD Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12151070563.14.RICHARDSON@SU-SCORE.ARPA>

Tomorrow's lunch speaker and topic will be: Tom Wasow on "An Interdepartmental
Program in Symbolic Systems". MJH 146 at 12:15.

-------

∂14-Oct-85  0907	@SUMEX-AIM.ARPA:HAUNGA@SU-SCORE.ARPA 	SPEAKER for Friday 18th SIGLUNCH is David Heckerman.
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 14 Oct 85  09:07:41 PDT
Received: from SU-SCORE.ARPA by SUMEX-AIM.ARPA with TCP; Mon 14 Oct 85 09:05:42-PDT
Date: Mon 14 Oct 85 09:02:13-PDT
From: Ana Haunga <HAUNGA@SU-SCORE.ARPA>
Subject: SPEAKER for Friday 18th SIGLUNCH is David Heckerman.
To: SIGLUNCH@SUMEX-AIM.ARPA
Message-ID: <12151073854.11.HAUNGA@SU-SCORE.ARPA>


Sorry for not mentioning it in the previous
message.

--a
-------

∂14-Oct-85  1231	CAROL@SU-CSLI.ARPA 	Phone number   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Oct 85  12:31:04 PDT
Date: Mon 14 Oct 85 12:27:21-PDT
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Phone number
To: Folks@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA

Hi-

I have a new phone number here in Ventura 28:  321-2136.  Don't 
forget to dial 9.

-Carol
-------

∂14-Oct-85  1557	ACUFF@SUMEX-AIM.ARPA 	Explorer Update   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 14 Oct 85  15:57:50 PDT
Date: Mon 14 Oct 85 15:56:25-PDT
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorer Update
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12151149256.15.ACUFF@SUMEX-AIM.ARPA>

Folks -
   Good News:  We now have 10 MB (at least for a while) of memory in
our first Explorer.

   Bad News:  No new software or machines until next week (but both
should arrive then).

	-- Rich
-------

∂14-Oct-85  1849	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #22  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 14 Oct 85  18:49:27 PDT
Date: 14 Oct 85 1820-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #22
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Tuesday, 15 Oct 1985      Volume 1 : Issue 22

Today's Topics:

            Seminar: Randomized Routing on Fat-Trees (MIT)
     Seminar: Connectionist Parallel Distributed Processing (UCB)
      A reading list for "Connectionism and Parallel Dist. Proc."

----------------------------------------------------------------------

Date: 10/08/85 11:37:44
From: LISA at MIT-MC.ARPA
Re:   R.L. Greenberg, "Randomized Routing on Fat-Trees"

[Forwarded by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

                                    SEMINAR

                       DATE: Thursday, October  10, 1985
                       TIME:  Lecture 2:00 p.m.
                       PLACE: NE43-512A

                       "RANDOMIZED ROUTING ON FAT-TREES"

                              Ronald L. Greenberg
                                    M.I.T.
                        Laboratory for Computer Science

                                   ABSTRACT

     Fat-trees are a class of routing networks for hardware-efficient
parallel computation.  This talk will present a randomized algorithm
for routing messages on a fat-tree.  The qualiity of the algorithm is
measured in terms of the load factor of a set of messages to be
routed, which is a lower bound on the time required to deliver the
messages.  We show that if a set of messages has load factor lambda =
omega (lg n lg lg n) on a fat-tree with n processors, the number of
delivery cycles (routing attempts) that the algorithm requires is
O(lambda) with probability 1 - O(1/n).  The best previous bound was
O(lambda lg n) for the off-line problem where switch settings can be
determined in advance.  In a VLSI-like model where hardware cost is
equated with physical volume, we use the routing algorithm to
demonstrate that fat-trees are universal routing networks in the sense
that any routing network can be efficiently simulated by a fat-tree of
comparable hardware cost.  This is joint work with Charles Leiserson.

------------------------------

Date: Wed, 25 Sep 85 14:09:11 PDT
From: chertok%ucbcogsci at Berkeley.EDU (Paula Chertok)
To:   AIList
Re:   Seminar - Connectionist Parallel Distributed Processing (UCB)

[Forwarded from AIList by Steven A. Swernofsky <SASW@MIT-MC.ARPA>]

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                    Cognitive Science Seminar -- IDS 237A

        TIME:                Tuesday, October 1, 11:00 - 12:30
        PLACE:               240 Bechtel Engineering Center
        (followed by)
        DISCUSSION:          12:30 - 1:30 in 200 Building T-4

        SPEAKER:          David  Rumelhart,  Institute  for  Cognitive
                          Science, UCSD

        TITLE:            ``Parallel Distributed Processing:  Explora-
                          tions in the Microstructure of Cognition''

        Parallel Distributed Processing (PDP) is the name which I  and
        my  colleagues  at  San  Diego  have  given  to  the  class of
        neurally-inspired models of cognition we have  been  studying.
        We  have  applied  this  class  of "connectionist" models to a
        variety of  domains  including  perception,  memory,  language
        acquisition  and motor control.  I will briefly present a gen-
        eral framework for the class of PDP  models,  show  how  these
        models  can  be applied in the case of acquisiton of verb mor-
        phology, and show how such  macrostructural  concepts  as  the
        schema  can be seen as emerging from the microstructure of PDP
        models.  Implications of the PDP perspective  for  our  under-
        standing of cognitive processes will be discussed.

------------------------------

Date: Fri, 11 Oct 85 15:01:57 est
From: "Marek W. Lugowski" <marek%indiana.csnet@CSNET-RELAY.ARPA>
Subject: A reading list for a "Connectionism and Parallel Dist. Proc."
Subject: course

[Long message, last in this digest. -- BD]

I posted the following to AIList@SRI-AI recently.  PGS%OZ@MC suggests
that it might be of interest here.  As I do not subscribe to PARSYM,
any replies ought to be made to MAREK@INDIANA.CSNET (or MAREK%OZ@MC,
same difference).

			-- Marek

P.s.  The proposed course is to be taught at Indiana University's CS
Dept. sometime within a year.  It's meant for advanced graduate
students.
                           ----------------
A list of sources for teaching a graduate AI course "Connectionism and
Parallel Distributed Processing" at Indiana University's Computer Science
Department.  Compiled by Marek W. Lugowski (marek@indiana.csnet) and Pete
Sandon (sandon@wisc-ai.arpa), Summer 1985.


Ackley, D. H., "Learning Evaluation Functions in Stochastic Parallel
     Networks," CMU Thesis Proposal, December 1984.

Amari, S-I., "Neural Theory of Association and Concept-Formation," Biological
     Cybernetics, vol. 26, pp. 175-185, 1977.

Ballard, D. H., "Parameter Networks: Toward a Theory of Low-Level Vision,"
      IJCAI, vol. 7, pp. 1068-1078, 1981.

Ballard, D. H. and D. Sabbah, "On Shapes," IJCAI, vol. 7, pp. 607-612, 1981.

Ballard, D. H., G. E. Hinton, and T. J. Sejnowski, "Parallel Visual
     Computation," Nature, vol. 103, pp. 21-26, November 1983.

Ballard, D. H., "Cortical Connections: Structure and Function," University of
     Rochester Technical Report #133, July 1984.

Block, H. D., "A Review of "Perceptrons: An Introduction to Computational
     Geometry"," Information and Control, vol. 17, pp. 501-522, 1970.

Bobrowski, L., "Rules of Forming Receptive Fields of Formal Neurons During
     Unsupervised Learning Processes," Biological Cybernetics, vol. 43,
     pp. 23-28, 1982.

Bongard, M., "Pattern Recognition", Hayden Book Company (Spartan Books), 1970.

Brown, C. M., C. S. Ellis, J. A. Feldman, T. J. LeBlanc, and G. L. Peterson,
     "Research with the Butterfly Multicomputer," Rochester Research Review,
      pp. 3-23, 1984.

Christman, D. P., "Programming the Connection Machine," MIT EECS Department
     Masters Thesis, 1984.

Csernai, L. P. and J. Zimanyi, "Mathematical Model for the Self-Organization
     of Neural Networks," Biological Cybernetics, vol. 34, pp. 43-48, 1979.

Fahlman, S. E., NETL, a System for Representing and Using Real Knowledge ,
     MIT Press, Cambridge, Massachusetts, 1979.

Fahlman, S. E., G. E. Hinton, and T. J. Sejnowski, "Massively Parallel
     Architectures for AI: NETL, Thistle and Boltzmann Machines," Proceedings
     of the National Conference on Artificial Intelligence, 1983.

Feldman, J. A., "A Distributed Information Processing Model of Visual Memory,"
     University of Rochester Technical Report #52, December 1979.

Feldman, J. A., "Dynamic Connections in Neural Networks," Biological
     Cybernetics, vol. 46, pp. 27-39, 1982.

Feldman, J. A. and D. H. Ballard, "Computing with Connections," in Human and
     Machine Vision, ed. J. Beck, B. Hope and A. Rosenfeld (eds), Academic
     Press, New York, 1983.

Feldman, J. A. and L. Shastri, "Evidential Inference in Activation Networks,"
     Rochester Research Review, pp. 24-29, 1984.

Feldman, J. A., "Connectionist Models and Their Applications: Introduction,"
     Special Issue of Cognitive Science, vol. 9, p. 1, 1985.

Fry, G., ed., Nanotechnology Notebook, an unpublished collection of published
     and unpublished material on molecular computing.  Contact either the
     editor (cfry@mit-prep.arpa) or Scott Jones (saj@mit-prep.arpa) at MIT
     Artificial Intelligence Laboratory for distribution and/or bibliography
     information.  Contains material by Eric Drexler, Richard Feynman, Kevin
     Ulmer and others.

Fukushima, K., "Cognitron: A Self-organizing Multilayered Neural Network,"
     Biological Cybernetics, vol. 20, pp. 121-136, 1975.

Fukushima, K., "Neocognitron: A Self-organizing Neural Network Model for
     a Mechanism of Pattern Recognition Unaffected by Shift in Position,"
     Biological Cybernetics, vol. 36, pp. 193-202, 1980.

Hebb, D. O., The Organization of Behavior, Wiley, New York, 1949.

Hewitt, C., "Viewing Control Structures as Patterns of Passing Messages",
     Artificial Intelligence: An MIT Perspective, Winston and Brown, Editors,
     MIT Press, Cambridge, Massachusetts, 1979.

Hewitt, C. and P. de Jong, "Open Systems", MIT Artificial Laboratory Memo #691,
     December 1982.

Hewitt, C. and H. Lieberman, "Design Issues in Parallel Architectures for
    Artificial Intelligence", MIT Artificial Intelligence Lab Memo #750,
    November 1983.

Hewitt, C., an article on asynchronous parallel systems not simulable
     by Turing Machines or Omega-order logic, BYTE, April 1985.

Hillis, D. W., "New Computer Architectures and Their Relationship to Physics
     or Why Computer Science Is No Good",  International Journal of
     Theoretical Physics, vol. 21, pp. 255-262, 1982.

Hinton, G. E., "Relaxation and its Role in Vision," University of Edinburgh
     Ph.D. Dissertation, 1977.

Hinton, G. E. and J. A. Anderson, Parallel Models of Associative Memory,
     Lawrence Erlbaum Associates, Hillsdale, New Jersey, 1981.

Hinton, G. E. and T. J. Sejnowski, "Optimal Perceptual Inference," Proceedings
     of the IEEE Computer Society Conference on CV and PR, pp. 448-453,
     June 1983.

Hinton, G. E., T. J. Sejnowski, and D. H. Ackley, "Boltzmann Machines:
     Constraint Satisfaction Networks that Learn," CMU Department of Computer
     Science Technical Report No. 84-119, May 1984.

Hinton, G. E., "Distributed Representations", CMU Department of Computer
     Science Technical Report No. 84-157, October 1984.

Hirai, Y., "A New Hypothesis for Synaptic Modification: An Interactive Process
     between Postsynaptic Competition and Presynaptic Regulation," Biological
     Cybernetics, vol. 36, pp. 41-50, 1980.

Hirai, Y., "A Learning Network Resolving Multiple Match in Associative
     Memory," IJCPR, vol. 6, pp. 1049-1052, 1982.

Hofstadter, D. R., "Who Shoves Whom Inside the Careenium?, or, What is the
     Meaning of the Word 'I'?", Indiana University Computer Science Department
     Technical Report #130, Bloomington, Indiana, 1982.

Hofstadter, D, "The Architecture of Jumbo", Proceedings of the International
     Machine Learning Workshop, Monticello, Illinois, June 1983.

Hofstadter, D. R., "The Copycat Project", MIT Artificial Intelligence Lab
     Memo #755, Cambridge, Massachusetts, January 1984.

Hofstadter, D. R., Metamagical Themas, Basic Books, New York, 1985.

Hopfield, J.J., "Neural Networks and Physical Systems with Emergent Collective
     Computational Abilities," Proceedings of the National Academy of
     Sciences, vol. 79, pp. 2554-2558, 1982.

Jefferson, D. E., "Virtual Time", UCLA Department of Computer Science
     Technical Report No. 83-213, Los Angeles, California, May 1983.

Jefferson, D. E. and H. Sowizral, "Fast Concurrent Simulation Using the Time
     Warp Mechanim, Part 1: Local Control", Rand Corporation Technical Report,
     Santa Monica, California, June 1983.

Jefferson, D. E. and H. Sowizral, "Fast Concurrent Simulation Using the Time
     Warp Mechanim, Part 2: Global Control", Rand Corporation Technical Report,
     Santa Monica, California, August 1983.

John, E. R., "Switchboard versus Statistical Theories of Learning and Memory,"
     Science, vol. 177, pp. 850-864, September 1972.

Kandel, E. R., "Small Systems of Neurons," Scientific American, vol. 241,
     pp. 67-76, September 1979.

Kanerva, P., Self-Propagating Search: A Unified Theory of Memory, Center
     for Study of Language and Information (CSLI) Report No. 84-7 (Stanford
     Department of Philosophy Ph.D. Dissertation), Stanford, California, 1984.
     To appear as a book by Bradford Books (MIT Press).

Kanerva, P., "Parallel Structures in Human and Computer Memory", Cognitiva 85,
     Paris, June 1985.

Kirkpatrick, S., and C. D. Gelatt, Jr. and M. P. Vecchi, "Optimization by
     Simulated Annealing", Science, vol. 220, no. 4598, 13 May 1983.

Kohonen, T., "Self-Organized Formation of Topologically Correct Feature Maps,"
     Biological Cybernetics, vol. 43, pp. 59-69, 1982.

Kohonen, T., "Analysis of a Simple Self-Organizing Process," Biological
     Cybernetics, vol. 44, pp. 135-140, 1982.

Kohonen, T., "Clustering, Taxonomy, and Topological Maps of Patterns," IJCPR,
     vol. 6, pp. 114-128, 1982.

Kohonen, T., Self-Organization and Associative Memory, 2nd edition,
     Springer-Verlag, New York, 1984.

Lieberman, H., "A Preview of Act1", MIT Artificial Intelligence Lab Memo #626.

Malsburg, C. von der, "Self-Organization of Orientation Sensitive Cells in the
     Striate  Cortex," Kybernetik, vol. 14, pp. 85-100, 1973.

McClelland, J. L., "Putting Knowledge in its Place: A Scheme for Programming
     Parallel  Processing Structures on the Fly," Cognitive Science, vol. 9,
     pp. 113-146, 1985.

McClelland, J. L. and D. E. Rumelhart, "Distributed Memory and the
     Representation of General and Specific Information," Journal of
     Experimental Psychology: General, vol. 114, pp. 159-188, 1985.

McClelland, J. L. and D. E. Rumelhart, Editors, Parallel Distributed
     Processing:  Explorations in the Microstructure of Cognition, Volume 1,
     Bradford books (MIT Press), Cambridge, Massachusetts, 1985 (in press).

McCullough, W. S., Embodiments of Mind, MIT Press, Cambridge, Massachusetts,
     1965.

Minsky, M. and S. Papert, Perceptrons: An Introduction to Computational
     Geometry, MIT Press, Cambridge, Massachusetts, 1969.

Minsky, M., "Plain Talk about Neurodevelopmental Epitemology", MIT Artificial
     Intelligence Lab Memo #430, June 1977.

Minsky, M., "K-Lines: A Theory of Memory", MIT Artificial Intelligence Lab
     Memo #516, June 1979.

Minsky, M., "Nature Abhors an Empty Vaccum", MIT Artificial Intelligence Lab
     Memo #647, August 8, 1981.

Mozer, M. C., "The Perception of Multiple Objects: A Parallel, Distributed
     Processing Approach", UCSD Thesis Proposal, Institute for Cognitive
     Science, UCSD, La Jolla, California, August 1984.

Nass, M. M. and L. N. Cooper, "A Theory for the Development of Feature
     Detecting Cells in Visual  Cortex," Biological Cybernetics, vol. 19,
     pp. 1-18, 1975.

Norman, "Categorization of Action Slips", Psychological Review, vol. 88, no. 1,
     1981.

Palm, G., "On Associative Memory," Biological Cybernetics, vol. 36, pp. 19-31,
     1980.

Plaut, D. C., "Visual Recognition of Simple Objects by a Connection Network,"
     University of Rochester Technical Report #143, August 1984.

Rosenblatt, F., Principles of Neurodynamics, Spartan Books, New York, 1962.

Rumelhart, D. E. and D. Zipser, "Feature Discovery by Competitive Learning,"
     Cognitive Science, vol. 9, pp. 75-112, 1985.

Sabbah, D., "Design of a Highly Parallel Visual Recognition System," IJCAI,
     vol. 7, pp. 722-727, 1981.

Sabbah, D., "Computing with Connections in Visual Recognition of Origami
     Objects," Cognitive Science, vol. 9, pp. 25-50, 1985.

Smolensky, P., "Schema Selection and Stochastic Inference in Modular
     Environments", Proceedings of the AAAI-83, Washington, 1983.

Theriault, D. G., "Issues in the Design and Implementation of Act2", MIT
     Artificial Intelligence Lab Technical Report #728, June 1983.

Touretzky, D. S. and G. E. Hinton, "Symbols Among the Neurons: Details of
     a Connectionist Inference  Architecture," IJCAI, vol. 9, pp. 238-243,
     1985.

Uhr, L., "Recognition Cones and Some Test Results," in Computer Vision
     Systems, ed. A. R. Hanson and E. M. Riseman, Academic Press, New York,
     1978.

Uhr, L. and R. Douglass, "A Parallel-Serial Recognition Cone System for
     Perception: Some Test Results," Pattern Recognition, vol. 11, pp. 29-39,
     1979.

Van Lehn, K., "A Critique of the Connectionist Hypothesis that Recognition
     Uses Templates, not Rules," Proceedings of the Sixth Annual Conference
     of the Cognitive Science Society, 1984.

Willshaw, D. J., "A Simple Network Capable of Inductive Generalization,"
     Proceedings of the Royal Society of London, vol. 182, pp. 233-247, 1972.
                           ----------------
A list of contacts known to us that have taught or are interested in
teaching a similar course/seminar:

J. Barnden, Indiana U. Computer Science
G. Hinton, CMU Computer Science
D. Hofstadter, U. of Michigan Psychology
J. McClelland, CMU Psychology
D. Rumelhart, UCSD Psychology
P. Smolensky, U. of Colorado Computer Science
D. Touretzky, CMU Computer Science

------------------------------

End of PARSYM Digest
********************

∂15-Oct-85  0820	DAVIES@SUMEX-AIM.ARPA 	Meeting at 9:30 Wednesday  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 15 Oct 85  08:20:22 PDT
Date: Tue, 15 Oct 1985  08:21 PDT
Message-ID: <DAVIES.12151328649.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: Meeting at 9:30 Wednesday
cc:   Davies@SUMEX-AIM.ARPA

Harold Brown will report on last week's DARPA Strategic Computing
meeting.

∂15-Oct-85  0829	RICHARDSON@SU-SCORE.ARPA 	Tuesday Lunch Series    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 15 Oct 85  08:29:29 PDT
Date: Tue 15 Oct 85 08:27:29-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12151329676.15.RICHARDSON@SU-SCORE.ARPA>

Don't forget -- lunch today in MJH 146 at 12:15!
-------

∂15-Oct-85  0932	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 15 Oct 85  09:32:32 PDT
Received: by su-aimvax.arpa with Sendmail; Tue, 15 Oct 85 09:29:22 pdt
Date: Tue, 15 Oct 85 09:29:22 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

We'll meet at 11AM tomorrow.
There is, as usual, no agenda.
I'll try to suggest some open problems, and perhaps others
can suggest their favorites as well.

By the way: the following was returned to me;
perhaps someone wants to look at it:
S Kunifuji and H. Yokuta, "Plorog and relational databases for
fifth generation computer systems."

∂15-Oct-85  1103	RICHARDSON@SU-SCORE.ARPA 	Tuesday CSD Lunch Series
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 15 Oct 85  11:03:35 PDT
Date: Tue 15 Oct 85 11:01:16-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Tuesday CSD Lunch Series
To: faculty@SU-SCORE.ARPA
Message-ID: <12151357670.15.RICHARDSON@SU-SCORE.ARPA>

Nils will not be available for the CSD Tuesday Lunch Series on November 5
and on November 19. Is there anyone out there that has a favorite lunch
topic and is willing to lead a discussion on same for either of these dates?

Anne
-------

∂22-Oct-85  1545	BARWISE@SU-CSLI.ARPA 	Logic Lunch  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  15:42:25 PDT
Date: Sun 20 Oct 85 14:32:39-PDT
From: Jon Barwise <BARWISE@SU-CSLI.ARPA>
Subject: Logic Lunch
To: Folks@SU-CSLI.ARPA

Reminder: Logic lunch moves to the philosophy lounge tomorrow, Monday,
at noon.
-------

∂22-Oct-85  1828	HANRAHAN@SU-SCORE.ARPA 	petty cash 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  17:22:04 PDT
Date: Tue 22 Oct 85 10:37:01-PDT
From: Katherine Hanrahan <HANRAHAN@SU-SCORE.ARPA>
Subject: petty cash
To: csd@SU-SCORE.ARPA
Stanford-Phone: (415) 497-2273
Message-ID: <12153188264.32.HANRAHAN@SU-SCORE.ARPA>

Sorry for the inconvenience (once again) but there will be no petty cash
until friday afternoon.   Katie
-------

∂22-Oct-85  1859	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar on Logic and the Foundations of Mathematics    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  18:57:35 PDT
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 22 Oct 85 18:53:00-PDT
Date: 22 Oct 85  1722 PDT
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Seminar on Logic and the Foundations of Mathematics   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    



Speaker: Dag Westerstahl, CSLI

Title: Branching generalized quantifiers

Time: Friday, Oct. 25, Noon to 1:15 

Place: Room 383N (faculty lounge, 3d floor, Math. Bldg.).

                                     S. Feferman


NB.  This will be our regular time from now on, every other week.

∂22-Oct-85  1908	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #43
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  15:48:00 PDT
Date: Monday, October 21, 1985 6:53PM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #43
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Tuesday, 22 Oct 1985      Volume 3 : Issue 43

Today's Topics:
                 LP Philosophy - Hewitt's Challenge,
               Implementation - Destructive Assignment,
                      Announcement - Third ICLP,
                        Puzzles - Games & Fun
----------------------------------------------------------------------

Date: Thu, 17 Oct 85 16:46:21 bst
From: William Clocksin <wfc%computer-lab.cambridge.ac.uk@ucl-cs>
Subject: Lisp in Prolog

One of Carl Hewitt's recent comments concerns writing a Lisp
system in Prolog.  It has subsequently been argued by
contributors that Hewitt's comments are irrelevant to Prolog
as a foundation for AI (whatever "foundation" means anyway).
I agree with them, but setting that aside for a moment, it
cannot have escaped the attention of most people that writing
a compiler (in Prolog) to COMPILE Lisp into some target (say LAP)
would be as easy as pie.  For those of us who have written
compilers in both Lisp and Prolog, I daresay it would be a lot
easier in Prolog.  I might further suggest that the compiler
test is more useful than the interpreter test, and that the
interpreter test has no especial theoretical advantage, since
the compiler can be seen as a meaning-preserving transformation.
(Ahem. Which in fact it isn't for Lisp if you're not careful
with bindings).

------------------------------

Date: 9 Oct 85 12:56 PDT
From: Kahn.pa@Xerox.ARPA
Subject: Destructive assignment

First I would like to comment on Paul Hudak discussion on mutable
arrays and elaborate on an earlier point by Fernando.

A couple years ago I implemented mutable arrays in LM-Prolog.  The
idea is that an update predicate have arguments for both the old and
new version of the array being changed.   The semantics of the
predicate was given as a straight-forward but expensive simulation
of arrays using lists and updates using copying.  The implementation
was, of course, free to represent the versions of the array in any
way that maintained the semantics of the predicate.

The most common representation was a perfectly ordinary array for
the most recent version and for the old version a data structure
representing the old values for component that have changed and
pointer to the newest version.  The details of all this can be
found in the paper by Lars-Henrik Eriksson and Manny Rayner in the
proceedings of the Second International Logic Programming Conference
held in Uppsala Sweden 1984. As Paul points out that when the old
version is not referenced after the update that a compiler can
optimize even that overhead away.

Concurrent Prolog has an even more direct way of programming as if
one had destructive assignment by relying on tail recursion
optimization. But both schemes and the one that Fernando talked about
earlier have one deficiency/ackwardness.  Namely, a "pointer" to a
mutable data structure does not see any changes since they continue to
point the old version. As Fernando points out to get around this one
needs to start at the top of any system of data structures.  In
Concurrent Prolog one avoids this problem in a different ackward
manner -- every time in a conventional program with side effects that
one "passes out" pointers one needs to explictly call "merge" to split
the references.

------------------------------

Date: 9 Oct 85 12:56 PDT
From: Kahn.pa@Xerox.ARPA
Subject: Prolog vs Lisp

Apropos the long debate started by Carl Hewitt:

As co-author of LM-Prolog, the best performing Prolog
implemented in Lisp,  I thought some performance numbers
might be useful.   Naive reverse runs at 10K Lips on a
CADR Lisp Machine with special micro-code support,
primarily for unification and trailing.  Without any
micro-code support it runs 2 to 3 times slower.  While
there are much faster Prologs available, I would argue
that LM-Prolog is commerically viable without the
micro-code.  There have been sales of the 3600 version
of LM-Prolog despite the fact that it is not supported
by micro-code.

But part of Fernando's counter to Carl was that a serious
Prolog implementation needs sub-primitives that Lisp does
not provide.  It is true that LM-Prolog even without micro
-code relies on Zeta-Lisp's sub-primitives to manipulate
pointers and create invisible pointers.  This makes
de-referencing variables very fast.   While I don't have
any figures I don't think that is so important, at least
for the naive reverse benchmark.  In other words I believe
a pure Common Lisp implementation of Prolog on say a 3600
would run 3 or 4 times slower than Symbolics Prolog (which
is fully micro-coded).  Depending upon how important a
factor of 3 or 4 is, one evaluates differently Carl's claim
that Lisp is good for implementing Prolog (and not vice
versa).

I think part of this whole debate is confused with the larger
debate of single paradigm vs multi-paradigm languages.  My
feeling is that while a single paradigm system is elegant that
too often it doesn't fit the problem well and ackward cliches
are used.  For example, it is widely believed that for some
kinds of problems object-oriented programming is most
appropriate because it encasulates state and behavior so well.

Concurrent Prolog advocates in such situations program objects
in a complex cliche of tail recursive predicates where one
argument is a stream of messages.  No serious object-oriented
language requires that each method list all the instance variables
in the head and their new values again at the end of each "method"
(the tail recursive call).  I am not happy with the argument that
goes -- well some problems are best programmed with Lisp, others
with Prolog, others with SmallTalk, and still others with OPS5.
Any significantly large problem is going to have sub-problems
that are best handled by different paradigms.

The debate should not be Lisp vs. Prolog but how can we combine
Lisp and Prolog (and Smalltalk and ...) in a coherent well
-integrated fashion. Its not easy.  LM-Prolog was one attempt at
doing this, as well as ICOT's ESP,  Prolog-KR and LogLisp.  I
tried to integrate Prolog with Loops.  None of these integrations
are perfect but I think this is the direction to go for BUILDING
TOOLS for BUILDING REAL APPLICATIONS.  The CommonLoops effort at
Xerox represents to me the best effort to date to build a tight
integration of two paradigms (object and procedure-oriented).

In contrast, to what I just said I think the single paradigm
approach can be a great research strategy.  Much of the Logic
Programming community is caught up in the game of finding out
how far can one go with logic programming.  Can one write
simulators, text editors, graphics, operating systems, embedded
languages, and so on in Prolog or a language like it? It is
rightfully considered cheating to "escape to Lisp" or jump into
some object-oriented sub-system.  Their purpose is to explore
the paradigm itself -- its uses, its limitations, to stretch it
and pull it in new directions not to build real applications.
When building real applications the question is not can this or
that be done in Prolog, we all know that everything can be written
in Prolog,  but what language can give the best support for building
the application in the most fitting way.

------------------------------

Date: Fri, 18 Oct 85 17:35:28 -0200
From: Ehud Shapiro  <udi%wisdom.bitnet@WISCVM>
Subject: Third ICLP - a reminder

In case you have forgot, or in case you didn't read the Digest
over the summer, the deadline for papers to the Third International
Conference on Logic Programming is December 1st, 1985.

I enclose again the call for papers, since the previos message
suffered from several minor mistakes.

Regards,

-- Ehud Shapiro

P.S. If you have been a referee for the previous Conference or
for the previous Sumposia, please expect to receive a few papers
for refereeing for the winter vacation...  We will not ask
explicit permission to send them, but just send them to you and
hope for the best (i.e. for your cooperation...)


                           CALL FOR PAPERS

         Third International Conference on Logic Programming

        Imperial College of Science and Technology, London, UK

                           July 14-18, 1986


In cooperation with:

           Association for Computing Machinery
           British Computer Society
           Japan Society for Software Science and Technology

The conference will consider all aspects of logic programming,
including, but not limited to:

        Theory and foundations
        Architectures and Implementations
        Methodology
        Programming Languages and Environments
        Applications
        Relations to other computation models, programming
        languages, and programming methodologies.

Of special interest are papers related to parallel processing,
papers discussing novel applications and applications that address
the unique character of logic programming, and papers which
constitute a contribution to computer science at large.

Papers can be submitted under two categories, short --
up to 2000 words,  and long -- up to 6000 words.  Submissions will
be considered on the basis of appropriateness, clarity, originality,
significance, and overall quality.

Authors should send eight copies of their manuscript, plus an
extra copy of the abstract, to:

        Ehud Shapiro
        ICLP Program Chairman
        The Weizmann Institute of Science
        Rehovot 76100, Israel.

Deadline for submission of papers is December 1, 1985.
Authors will be notified of acceptance or rejection
by February 28, 1986.
Camera ready copies are due April 1st, 1986.


        Keith Clark
        General Chairman
        Imperial College of Science and Technology
        180 Queen's Gate
        London SW7 2BZ, United Kingdom

Local Arrangements and Exhibition Chairman

        Richard Ennals
        Imperial College of Science and Technology
        180 Queen's Gate
        London SW7 2BZ, United Kingdom

Program Committee
        Michel van Caneghem, University of Marseille, France
        Keith Clark, Imperial College, UK
        Veronica Dahl, Simon Fraser University, Canada
        Maarten van Emden, University of Waterloo, Canada
        Kazuhiro Fuchi, ICOT, Japan
        Koichi Furukawa, ICOT, Japan
        Ake Hansson, Uppsala University, Sweden
        Kenneth M. Kahn,  Xerox PARC, USA
        Peter Koves, Logicware Inc., Canada
        Giorgio Levi, University of Pisa, Italy
        John Lloyd, University of Melbourne, Australia
        Frank G. McCabe, Imperial College, UK
        Jack Minker, Maryland University, USA
        Fernando Pereira, SRI International, USA
        Antonio Porto, University of Lisbon, Portugal
        Ehud Shapiro, Chairman, Weizmann Institute, Israel

------------------------------

Date: Fri, 11-Oct-85 06:11:56 PDT
From: Joe Armstrong  <mcvax!enea!erix!joe@Seismo>
Subject: An exciting game in PROLOG

# unpacked will be owned by you and have default permissions.)
#
# This archive contains:
# READ←ME game.pl

echo x - READ←ME
cat > "READ←ME" << '//E*O*F READ←ME//'

A THRILLING GAME IN PROLOG WITH HIGH SPEED HIGH RESOLUTION VT100
GRAPHICS.

This program is donated to the public domain and may be used,
modified re-distributed, or simply throw away as seen fit by the
user.

With this program I hope to fill a 'gap' in the market - I have
often seen requests for "really badly written banal games written
in FORTRAN" - since PROLOG has been described as the FORTRAN of
logic programming I hope this program fills that gap.

I make no apologies for the code, though I suspect that purests
may object to the "(cut,...,fail);true" constructs that are
liberally  sprinkled throughout the program.

Plans are underway to write yet more thrilling, exciting,
spectacular, addictive, brilliant, wonderful, enchanting games in
SASL, HOPE, PARLOG, ML, etc.

-- Joe Armstrong.

//E*O*F READ←ME//

echo x - game.pl
cat > "game.pl" << '//E*O*F game.pl//'
/*------------------------------------------------------------*/
/* to run this program: enter prolog and type ['game.pl'],go. */
/*------------------------------------------------------------*/

go :-
        noscroll,
        system('stty cbreak',←),
        start,!,
        play.

play :-
        putat(24,1,"type space to spin the chamber"),
        get0(←),
        putat(24,1,"                              "),
        random(6,N),
        move(24,60),put("you spun "),write(N),
        possibly←die(N),
        play.

possibly←die(6) :- die.
possibly←die(N).

putat(X,Y,Text) :- move(X,Y),put(Text).

start  :-
        cls,print←thing←at(2,5,gun),print←thing←at(1,60,head).

die :-
        put(7),
        print←thing←at(11,20,bang),
        un←print←thing←at(11,20,bang),
        move←thing←at(2,39,30),
        print←thing←at(2,69,bullit),
        print←thing←at(11,20,aagh),
        un←print←thing←at(11,20,aagh),
        un←print←thing←at(1,60,head),
        cls,
        system('echo "your game got played" | mail joe@erix.UUCP',←),
        putat(24,1,"like to dice (sic) with death again (y/n):"),
        get0(X),
        possibly←doit←again(X).

possibly←doit←again(121) :- start,!,play.
possibly←doit←again(110) :- abort.

move←thing←at(X,Y,N) :-
    (for(I,1,N),
    Y1 is Y + I -1,
    print←thing←at(X,Y1,bullit),
    un←print←thing←at(X,Y1,bullit),
    fail);true.


un←print←thing←at(X,Y,Z) :- un←print←thing←at←1(X,Y,Z);true.

un←print←thing←at←1(X1,Y,Z) :-
    F =.. [Z,N,L],!,
    call(F),
    X is X1 + N - 1,
    move(X,Y),
    undraw(L),nl,fail.

undraw(L) :-
    (length(L,M),!,for(I,1,M),put(32),fail);true.

print←thing←at(X,Y,Z) :- print←thing←at←1(X,Y,Z);true.

print←thing←at←1(X1,Y,Z) :-
    F =.. [Z,N,L],!,
    call(F),
    X is X1 + N - 1,
    move(X,Y),
    put(L),nl,fail.

for(I,I,Upper).
for(I,Lower,Upper) :-
    Lower < Upper,
    Next is Lower + 1,
    for(I,Next,Upper).

move(X,Y) :- printf("%c%c%d%c%d%c",[27,91,X,59,Y,72]).

scroll(X,Y) :- printf("%c%c%d%c%d%c",[27,91,X,59,Y,114]),move(24,1).

noscroll :- scroll(1,24).

cls :- printf("%c%c%c%c",[27,91,50,74]).

seed(13).

random(R,N) :-
        retract(seed(S)),
        N is (S mod R) +1,
        NewSeed is (125*S+1) mod 4096,
        asserta(seed(NewSeed)),!.

gun(1, "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx").
gun(2, "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx").
gun(3, "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx").
gun(4, "xxxxxxxxxxxxxxxxxxxxxxxxx").
gun(5, "xxxxxxxxxxxxx x    x").
gun(6, "xxxxxxxxxxxxx  x  x").
gun(7, "xxxxxxxxxxxxx  x x").
gun(8, "xxxxxxxxxxxxxxxxx").
gun(9, "xxxxxxxxxxx").
gun(10,"xxxxxxxxxxx").
gun(11,"xxxxxxx").
gun(12,"xxxxxxx").
gun(13,"xxxxxxx").
gun(14,"xxxxxxx").
gun(15,"xxxxxxx").
gun(16,"xxxxxxx").

head(1, "        xxxxxxxxx").
head(2, "      xxxxxxxxxxxx").
head(3, "     xx          xxx").
head(4, "    x         xxxxxxx").
head(5, "   x  xxx      xxxxxx").
head(6, "   x xxxxx  x    xxx").
head(7, "    x xxxx    xx  xx").
head(8, "    x  xx       x  x").
head(9, "    x  xx        x x").
head(10,"   x  x          x x").
head(11,"  x               x").
head(12," x               x").
head(13,"x               x").
head(14,"x               x").
head(15,"xxxxx           x").
head(16," xx              x").
head(17," xxx              x").
head(18,"  xx   xxx        x").
head(19,"   xxxx   x      x").
head(20,"          x      x").
head(21,"          x      x").

bullit(1,"==>").
bullit(2,"===>").
bullit(3,"==>").

bang(1, "xxxx     xx    x    x   xxx  ").
bang(2, "x   x   x  x   xx   x  x   x ").
bang(3, "x  x   x    x  x x  x  x   x ").
bang(4, "xxx    xxxxxx  x x  x  x     ").
bang(5, "x  x   x    x  x  x x  x  xxx").
bang(6, "x   x  x    x  x   xx   x  x ").
bang(7, "xxxx   x    x  x    x    xx  ").

aagh(1, "  xx     xx     xxx   x    x").
aagh(2, " x  x   x  x   x   x  x    x").
aagh(3, "x    x x    x  x   x  x    x").
aagh(4, "xxxxxx xxxxxx  x      xxxxxx").
aagh(5, "x    x x    x  x  xxx x    x").
aagh(6, "x    x x    x  x  x   x    x").
aagh(7, "x    x x    x   xx    x    x").

//E*O*F game.pl//

exit 0

------------------------------

End of PROLOG Digest
********************

∂22-Oct-85  1911	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  18:24:36 PDT
Date: Tue 22 Oct 85 10:26:57-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12153186430.20.RICHARDSON@SU-SCORE.ARPA>



Lunch today --- MJH 146 at 12:15

-------

∂22-Oct-85  1915	ullman@su-aimvax.arpa 	meeting this week
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  15:52:35 PDT
Received: by su-aimvax.arpa with Sendmail; Mon, 21 Oct 85 10:52:35 pdt
Date: Mon, 21 Oct 85 10:52:35 pdt
From: Jeff Ullman <ullman@diablo>
Subject: meeting this week
To: nail@diablo

This week we actually have a speaker.
Richard Treitel is  going to speak about conjuct ordering.

The meeting will begin at TEN AM, in 301 MJH,
as both Richard and I have to leave before noon.

PS: You all have interesting things to speak about;
I know.  So a volunteer or too would be appreciated.
				---Jeff

∂22-Oct-85  1916	SELLS@SU-CSLI.ARPA 	Hinrichs dissertation    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  16:00:45 PDT
Date: Mon 21 Oct 85 15:15:21-PDT
From: Peter Sells <Sells@SU-CSLI.ARPA>
Subject: Hinrichs dissertation
To: folks@SU-CSLI.ARPA

If anybody's interested, I have received a copy of Erhard Hinrich's
dissertation "A Compositional Semantics for Aktionsarten and NP Reference in
English", and will willingly lend it for viewing or copying.

Peter
-------

∂22-Oct-85  1919	BRESNAN@SU-CSLI.ARPA 	"Topic, Pronoun, and Agreement in Chichewa"
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  15:53:43 PDT
Date: Mon 21 Oct 85 11:26:19-PDT
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: "Topic, Pronoun, and Agreement in Chichewa"
To: folks@SU-CSLI.ARPA

This is the title of the presentation to be given this Thursday
(Oct. 24) at 10:00 at the Ventura Conference Room at CSLI by Joan
Bresnan.  A written version of this work by Bresnan and Mchombo
is now available at the receptionist's desk at CSLI.
-------

∂22-Oct-85  1919	LIBRARY@SU-SCORE.ARPA 	Math/CS Library:  Electronic Services--If New To Stanford Please Read   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  18:40:34 PDT
Date: Tue 22 Oct 85 17:16:44-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library:  Electronic Services--If New To Stanford Please Read
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12153261031.9.LIBRARY@SU-SCORE.ARPA>

The Math/CS Library offers some services through electronic messaging.  Our
address is Library@SCORE.  We announce new technical reports and new books
on the bulletin boards.  When we announce new material they are usually
on display in the library for a week to two weeks.  If you would like to
be placed on a waiting list to see the material after it is off of display,
send us a message and include the item you would like to see (accession
number or call number, author, and title) and your name, electronic address,
and physical address.

You may also renew books electronically.  If you have a reference question,
send that to us and we will get back to you with the answer or advice on
how to approach the problem. 

When using the Math/CS Library electronically, please attempt to send
us as much information as you have.  Please be as precise and concise
as you can.  Always include your name, physical address, and electronic
address.  If you make a request that doesn't require an answer back,
don't expact a response back from us such as ok etc.  If there is
some problem with your request (such as you want to renew a book and
someone else is waiting for it) we will get back to you.  When
we have electronic mailing addresses we will recall materials from
you with an electronic message.    If you find a citation in Socrates
that would probably be in the Math/CS Library but you are referred
to a HELP STATUS screen, you could send the record to me and we 
will make sure we place you on a list to be notified when received.

I hope you will find some of these services helpful to you while at
Stanford.

Harry Llull
-------

∂22-Oct-85  1921	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.Berkeley.EDU  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  16:04:38 PDT
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Sat 19 Oct 85 16:06:40-PDT
Received: from UCB-VAX by SU-SCORE.ARPA with TCP; Sat 19 Oct 85 16:06:16-PDT
Received: by UCB-VAX (5.28/5.13)
	id AA06342; Sat, 19 Oct 85 15:58:29 PDT
Received: by ernie (5.28/5.13)
	id AA02466; Sat, 19 Oct 85 15:57:59 PDT
Date: Sat, 19 Oct 85 15:57:59 PDT
From: karp@ernie.Berkeley.EDU (Richard Karp)
Message-Id: <8510192257.AA02466@ernie>
To: Theory-b@ernie.Berkeley.EDU, aflb.all@su-score.arpa,
        allmsgs@ernie.Berkeley.EDU, csfac@ernie.Berkeley.EDU,
        theory@ernie.Berkeley.EDU


SEMINAR ANNOUNCEMENTS
MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley  (415)-548-6843

Tuesday, October 29    2:00   MSRI Lecture Hall
Steven Fortune  "Sweepline Algorithms for Voronoi Diagrams"

Tuesday, October 29    4:00    MSRI Lecture Hall
Ketan Mulmuley   "Computing the Rank of a Matrix Over an Arbitrary Field is 
in NC(2)

∂22-Oct-85  1923	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #24  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  16:11:01 PDT
Date: 21 Oct 85 2121-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #24
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Tuesday, 22 Oct 1985      Volume 1 : Issue 24

Today's Topics:

        Second PARSYM Survey -- Data Abstractions (continued)

----------------------------------------------------------------------

Date: Sun 20 Oct 85 21:28:15-EDT
From: vijay <Vijay.Saraswat@C.CS.CMU.EDU>
Subject: PARSYM Digest second survey.

Let me take this opportunity to respond to the PARSYM Survey on data
abstractions. I cannot resist, particularly after Bert Halstead's post
on MULTILISP (PARSYM Digest V1 #20). The remarks I will make are
generally applicable to the spectrum of concurrent logic programming
languages (Concurrent Prolog, Guarded Horn Clauses, Parlog,
Delta-Prolog), but I will make them in the context of the language
CP[!,|,&,;] on which I have been working.  (I assume familiarity with
the basic notions of logic programming.)

    1.  Are you using or have you developed data structures or
	abstractions specialized for parallel computing?  Please
	describe.

Most CLP languages are based on the notion of pattern-matching via
unification.  In these languages the (logical) variable X stands for
some fixed but possibly unknown object.  As computation proceeds,
several processes can cooperate to construct a value for this object.
By the very nature of unification, (almost) ALL operations in these
languages can happen even if the variable does not have a binding.
(Remember that as far as operation on data is concerned, there is ONLY
ONE primitive in these languages -modulo details- and that is
unification: this primitive is used for parameter passing/"value
return", "assignment", data-structure construction/selection etc.)

Most CLP languages, however, provide for mechanisms that allow the
user to restrict the bidirectionality of unification: CP[!,|,&] does
it by the notion of the wait-annotation (!). This annotation can
decorate occurrences of terms (variables/constants/structured terms)
ONLY in the head of a clause. Whenever a goal is unified against the
head of a clause, then !-unification succeeds iff all !-decorated
terms in the head of the clause unify against non-variables; otherwise
!-unification may suspend or fail. (!-unification fails iff (normal)
unification fails.)

For example, given the clause "+(A,0!,A)." !-unification of
--the goal "+(X,Y,2)" will SUSPEND,
--the goal "+(X,0,Y)" will SUCCEED with X unifying against Y,
--the goal "+(X,2,Y)" will FAIL.


The "future" construct that Halstead describes seems to be capturing
precisely the behaviour of the logical variable.  The logic
programming paradigm already has (the equivalent of) that construct
built-in.  On the other hand, CLP languages must specify extra-logic
control annotations to allow a programmer to specify (in the
terminology of Halstead) "strict operators".  For instance here is the
definition of "strict car":
          car([A | Rest]!, A).
Note:
-- "the value returned by car" -A- may itself be unknown (i.e. a
"future")... if it is desired to wait for A to be instantiated as
well, one needs the axiom "car([A! | Rest]!, A!)." instead.
-- if the ! was removed from the axiom, the resultant axiom would have
allowed the CREATION of a list whose car was A, rather than waiting
for the list to be instantiated.

To give another example, here is the definition of a "strict plus",
which assumes the presence of an is/2 arithmetic predicate.

+(X!,Y!,Z):- Z is X+Y.
+(X!,Y,Z!):- Y is Z-X.
+(X,Y!,Z!):- X is Z-Y.
+(X,0!,X).
+(0!,X,X).

    2.  In particular, do you use specialized data abstractions to
	represent configurations of multiple processes?  Please
	describe.

Again, the use of the logical variable allows the programmer to create
very versatile configurations of processes.  In fact the basic
programming technique that Shapiro's Concurrent Prolog supports well
is the creation of process structures which exactly mimic the
data-structures in a more conventional sequential algorithm.
Specifically, with other collaborators, he has published literally
dozens of algorithms which use the pipeline structure that Halstead
mentions, along with other such exotic structures as "pyramidal
computers", H-trees, aleph trees, quad trees etc.

I have personally been exploring using unification to achieve
algorithms with "surprising" complexity: e.g. Danny Dolev et al's
"Lord of the Rings" problem (computing the maxima in a ring of
processes) can be solved in O(n) unifications (as contrasted with
O(n log n) messages in his algorithm)... the trick is that unification
allows you to dynamically reduce the size of the ring at practically
no cost.

 CP[!,|] (which is identical to Concurrent Prolog but for the '!' which
is simpler, and cleaner than Concurrent Prolog's '?') also naturally
supports data-flow programming in general and the soft-systolic
paradigm in particular.

 CP[!,&,;], which has AND-concurrency and OR-sequentiality (;), is also
the natural framework for the computational paradigm of constraints.
This paradigm is characterised by collections of processes with
provisions for `distributed backtracking'.  Unlike Guy Steele's
system, however, it is also possible to describe algebraic
transformations of the constraint network in a CP[!,&,;]-like
language.

    3.  Do you use specialized data structures, a la FRONS, for
	representing non-deterministic collections of data?  Please
	describe.

With Larry Rudolph I have been exploring the use of ACId
data-structures, i.e. the free algebra built from a binary constructor
f/2 which is associative, commutative and idempotent for nil (i.e.
f(nil,nil)=nil).  As is well known these equations can be built into
the unifier engine.  One fallout is that two terms may now have more
than one most general unifiers. This allows a very powerful
`non-determinstic' way to provide random (`associative') access to the
multi-set data-structure.  In fact, in conjunction with
AND-concurrency and the wait (!) control annotation this allows the
facile simulation of shared-memory (PRAM-like) architectures in the
CLP framework.

This data-structure is quite powerful when used in conjunction with
the don't know commit [&] in CP[!,|,&] which allows the pursuit of
disjoint disjunctive paths by processes.

    4.  What are some useful references on the topic of data abstraction
	in parallel computing (including your own)?

More information on the language  CP[!,|,&] can be obtained from my
paper on `Partial correctness semantics for CP[!,|,&]' which is to be
presented at the Fifth Conference on Foundations of Software
Tech. and Theo. Comp. Sci., New Delhi Dec 1985 and by to-be-out-soon
tech rep on `An operational semantics for CP[!,|,&]'. The primary reference
on Concurrent Prolog remains the paper `A subset of Concurrent Prolog
and its interpreter', Weizmann Instt. tech rep, 1983.  An
algorithm-oriented reference is `Systolic programming: a pardigm for
parallel processing' by Shapiro in FGCS 84. A paper on
Parlog is to appear soon in TOPLAS.  None of these papers explicitly
discuss the topic of data-abstraction.

Vijay Saraswat.

------------------------------

End of PARSYM Digest
********************

∂22-Oct-85  2055	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  18:50:11 PDT
Date: Mon 21 Oct 85 08:16:10-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12152900478.19.RICHARDSON@SU-SCORE.ARPA>

Tomorrow, Oct. 22, the Tuesday Lunch series will continue with Dean Eustis
on the subject of "The Role of Personal Computers in Courses and Labs" in
MJH 146 at 12:15.

-------

∂22-Oct-85  2056	RICHARDSON@SU-SCORE.ARPA 	[The Mailer Daemon <Mailer@SU-SCORE.ARPA>: Message of 17-Oct-85 10:40:56] 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  18:53:08 PDT
Date: Mon 21 Oct 85 08:52:45-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: [The Mailer Daemon <Mailer@SU-SCORE.ARPA>: Message of 17-Oct-85 10:40:56]
To: Tob@SU-AI.ARPA, rwf@SU-AI.ARPA, dek@SU-AI.ARPA, zm@SU-AI.ARPA,
    jmc-lists@SU-AI.ARPA, tw@SU-AI.ARPA
Message-ID: <12152907140.19.RICHARDSON@SU-SCORE.ARPA>

Date: Sun 20 Oct 85 10:56:22-PDT
From: The Mailer Daemon <Mailer@SU-SCORE.ARPA>
To: RICHARDSON@SU-SCORE.ARPA
Subject: Message of 17-Oct-85 10:40:56

Message undeliverable and dequeued after 3 days:
Tob@SU-AI.ARPA.#Internet: Cannot connect to host
rwf@SU-AI.ARPA.#Internet: Cannot connect to host
dek@SU-AI.ARPA.#Internet: Cannot connect to host
zm@SU-AI.ARPA.#Internet: Cannot connect to host
jmc-lists@SU-AI.ARPA.#Internet: Cannot connect to host
tw@SU-AI.ARPA.#Internet: Cannot connect to host
	    ------------
Date: Thu 17 Oct 85 10:40:55-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Sr. Faculty Meeting in November
To: tenured@SU-SCORE.ARPA
Message-ID: <12151878255.11.RICHARDSON@SU-SCORE.ARPA>


Please let me know if you have any agenda item proposals for the November
Sr. Faculty Meeting.

Thanks,
Anne
-------
-------
-------

∂22-Oct-85  2058	@SU-SCORE.ARPA:WINOGRAD@SU-CSLI.ARPA 	Postponement of PhD program meeting  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Oct 85  18:57:44 PDT
Received: from SU-CSLI.ARPA ([36.9.0.46].#Internet) by SU-SCORE.ARPA with TCP; Mon 21 Oct 85 11:39:31-PDT
Date: Mon 21 Oct 85 11:35:41-PDT
From: Terry Winograd <WINOGRAD@SU-CSLI.ARPA>
Subject: Postponement of PhD program meeting
To: su-bboards@SU-CSLI.ARPA, phd@SU-SCORE.ARPA, faculty@SU-SCORE.ARPA,
    phd-program@SU-SCORE.ARPA
cc: cheadle@SU-SCORE.ARPA

Many of the theory students an faculty will be going to the FOCS meeting
tomorrow and would like to participate.   So it will be at the same time next
week (same place unless another announcment is made) i.e., 2:15 on Tuesday.
--t
-------

∂23-Oct-85  0609	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  06:09:00 PDT
Date: Wed 23 Oct 85 06:10:16-PDT
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: AFLB abstracts
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12153401849.9.PATASHNIK@SU-SUSHI.ARPA>

Here are the abstracts for the next two AFLBs.
		*************************************
	
24-Oct-85  :  Victor Pan (SUNY Albany)

  Improved processor bounds for algebraic and combinatorial problems in RNC

Our two main results improve the processor bounds of two important problems:

Problem 1: Computing the exact inverse and determinant of an n by n matrix
           whose entries are L-bit integers, L=n↑O(1).
Problem 2: Finding a perfect matching in a general graph.

The improved solutions maintain the best running time (O(log↑2 n),
O(log↑3 n), resp.) for the two problems.  A solution to Problem 1 is
used in a number of parallel algorithms for algebraic problems as well
as solving Problem 2.  A solution for Problem 2 is used in parallel
algorithms for several combinatorial problems.  Consequently, the new
algorithms lead to improved solutions to several algebraic and
combinatorial problems.

***** Time and place: October 24, 12:30 pm in MJ352 (Bldg. 460) ******

31-Oct-85  :  Steven Fortune (ATT Bell Labs)

	    Sweepline Algorithms for Voronoi Diagrams

We present a simple, sweepline algorithm for computing Voronoi
diagrams.  The algorithm is applicable to Voronoi diagrams arising
from a set of points, a set of line segments, and a set of weighted
points.  In all cases it achieves an O(n log n) worst-case time bound,
and a linear space bound.  The algorithm is easy to implement, since
it avoids the merge step required by divide-and-conquer algorithms.
An application to a version of the piano mover's problem will also be
discussed.

***** Time and place: October 31, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

More information about future and past AFLB meetings and topics are in
the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂23-Oct-85  0749	@SU-CSLI.ARPA:RPERRAULT@SRI-AI.ARPA 	Talk by Bill Rounds    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  07:49:50 PDT
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Wed 23 Oct 85 07:45:02-PDT
Date: Wed 23 Oct 85 07:47:54-PDT
From: Ray Perrault <RPERRAULT@SRI-AI.ARPA>
Subject: Talk by Bill Rounds
To: friends@SU-CSLI.ARPA, aic-staff@SRI-AI.ARPA
cc: rperrault@SRI-AI.ARPA

                    SEMINAR ANNOUNCEMENT REMINDER

                LOGIC AND LANGUAGE: CHARACTERIZING THE
                     COMPLEXITY OF LOGIC GRAMMARS

                          William C. Rounds
                        University of Michigan

                  4 p.m., Wednesday October 23, 1985
          SRI International, Conference Room EJ228 (Bldg. E)


    Modern artificial intelligence has seen the introduction of logic
as a tool for describing the syntax and semantics of natural language
grammars. In this talk I introduce two new notations for expressing
grammars, called CLFP and ILFP. These notations extend the first order
theory of concatenation and integer arithmetic with a Least Fixed
Point operator to accommodate recursive definitions of predicates. The
notations can be thought of as variants of definite clause grammars.
They are extremely easy to write and to understand. I prove that a
language is definable in CLFP if and only if it is recognizable by a
Turing machine in exponential time, and definable in ILFP if and only
if it is recognizable in polynomial time. As an application, I show
how to express head grammars in ILFP, thereby proving that head
languages are recognizable in polynomial time in a particularly easy
way.


-------

∂23-Oct-85  1048	CONTRERAS@SU-SCORE.ARPA 	Class Lists    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  10:47:57 PDT
Date: Wed 23 Oct 85 10:38:35-PDT
From: Tina Contreras <CONTRERAS@SU-SCORE.ARPA>
Subject: Class Lists
To: Instructors@SU-SCORE.ARPA
Message-ID: <12153450693.24.CONTRERAS@SU-SCORE.ARPA>


Your class lists are available here at the reception desk in MJH.

Tina
-------

∂23-Oct-85  1103	MODICA@SU-SCORE.ARPA 	Class Lists - Correction    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  11:03:18 PDT
Date: Wed 23 Oct 85 10:47:34-PDT
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: Class Lists - Correction
To: instructors@SU-SCORE.ARPA
Message-ID: <12153452330.46.MODICA@SU-SCORE.ARPA>

Please disregard previous message about Class Lists. Class Lists will be
available in MJH Rm. 251 today.
                               Gina
-------

∂23-Oct-85  1133	CONTRERAS@SU-SCORE.ARPA 	Mailboxes 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  11:33:52 PDT
Date: Wed 23 Oct 85 11:27:52-PDT
From: Tina Contreras <CONTRERAS@SU-SCORE.ARPA>
Subject: Mailboxes
To: Faculty@SU-SCORE.ARPA
cc: Staff@SU-SCORE.ARPA
Message-ID: <12153459665.24.CONTRERAS@SU-SCORE.ARPA>



Faculty and Staff mailboxes are going to be rearranged in alphabetical order.

Thank-You 


Tina Contreras
Receptionist
-------

∂23-Oct-85  1303	YAMARONE@SU-CSLI.ARPA 	R/B + swiss Available Now at Front Desk   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  13:03:11 PDT
Date: Wed 23 Oct 85 12:57:51-PDT
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: R/B + swiss Available Now at Front Desk
To: folks@SU-CSLI.ARPA


Hungry yet?

3.25 will buy you a scrumptious sandwich ....a roast beef + swiss
on a french roll with all the acutraments!!

Come on now...don't hesitate, or it'll be too late.

-------

∂23-Oct-85  1432	ullman@su-aimvax.arpa 	financial detail 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  14:31:26 PDT
Received: by su-aimvax.arpa with Sendmail; Wed, 23 Oct 85 14:29:27 pdt
Date: Wed, 23 Oct 85 14:29:27 pdt
From: Jeff Ullman <ullman@diablo>
Subject: financial detail
To: nail@diablo

I may have been misconstrued at the meeting today.
Concerning Dave Smith's conjecture that the optimal ordering
of two sets of independent conjuncts is a merger of the
optimal orderings of the two sets, what I said was that
I would bet $1 that it was false (in particular because the
presence of a second set of conjuncts turns the problem
of optimally ordering the first set into finding a weighted
optimal ordering).
I did NOT say I would pay $1 to anyone solving the problem.
If you wish to take the bet, you have to be willing to
lose $1 if anyone comes up with a counterexample.
				---Jeff

∂23-Oct-85  1611	admin@cogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 29, 1985    
Received: from UCBVAX.BERKELEY.EDU by SU-AI.ARPA with TCP; 23 Oct 85  16:11:10 PDT
Received: by UCB-VAX (5.29/5.13)
	id AA20261; Wed, 23 Oct 85 13:51:07 PDT
Received: by cogsci (5.29/5.13)
	id AA07435; Wed, 23 Oct 85 13:52:25 PDT
Date: Wed, 23 Oct 85 13:52:25 PDT
From: admin@cogsci.Berkeley.EDU (Cognitive Science Program)
Message-Id: <8510232052.AA07435@cogsci>
To: allmsgs@cogsci.Berkeley.EDU, cogsci-friends@cogsci.Berkeley.EDU
Subject: UCB Cognitive Science Seminar--Oct. 29, 1985
Cc: admin@cogsci.Berkeley.EDU

                   BERKELEY COGNITIVE SCIENCE PROGRAM
                              Fall 1985
                  Cognitive Science Seminar - IDS 237A
                   Tuesday, October 29, 11:00 - 12:30
                     240 Bechtel Engineering Center
               Discussion: 12:30 - 1:30 in 200 Building T-4

                          ``Person Schemata''

                         Mardi J. Horowitz M.D.
                    Professor of Psychiatry, U.C.S.F.

          The speaker directs the recently formed Program  on  Cons-
     cious  and  Unconscious  Processes of the John and Catherine T.
     MacArthur Foundation.  Research on person schemata  is  one  of
     the core agendas of this program.
          After a brief description of the program,  the  discussion
     will  focus  on  clinical  phenomena  as segmented by different
     states of mind in a single individual.  By examining the confi-
     guration  in  each state of mind as it occurs over time, it may
     be possible to infer what the self schemata and role  relation-
     ship  models  are  that  organize thoughts, feelings and action
     into observed patterns.  The theory that forms  the  basis  for
     such  inferences  includes  the  postulate  that  each person's
     overall  self  organization  may  include  a  partially  nested
     hierarchy  of multiple self-concepts.  A frequent set of states
     of mind in pathological grief reactions will provide a concrete
     illustration  of  phenomena, methods of inference, and a theory
     of person schemata.
  ---------------------------------------------------------------------
     UPCOMING TALKS
        November  5: Edward Zalta, CSLI, Stanford
        November 12: Robert Wilensky, Computer Science, UCB
        November 19: Richard Alterman, Computer Science, UCB
        November 26: Eve Clark, Linguistics, Stanford
        December  3: Bernard Baars, Langley Porter, UCSF
   ---------------------------------------------------------------------
     ELSEWHERE ON CAMPUS
        John Dalbey, SESAME student, will present ``The Totally Effort-
        less  Problem  Solver''  at  the  SESAME  Colloquium on Monday,
        October 28, 4:00pm, 2515 Tolman Hall.

        Tom Wickens, U.C.L.A., will speak on ``Response Interactions in
        Visual  Detections''  at  the  Cognitive Psychology Colloquium,
        Friday, November 1, 4:00pm, Beach Room, 3105 Tolman Hall.

∂23-Oct-85  1633	@SU-CSLI.ARPA:admin@cogsci.Berkeley.EDU 	UCB Cognitive Science Seminar--Oct. 29, 1985
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  16:31:57 PDT
Received: from UCB-VAX by SU-CSLI.ARPA with TCP; Wed 23 Oct 85 16:30:24-PDT
Received: by UCB-VAX (5.29/5.13)
	id AA20261; Wed, 23 Oct 85 13:51:07 PDT
Received: by cogsci (5.29/5.13)
	id AA07435; Wed, 23 Oct 85 13:52:25 PDT
Date: Wed, 23 Oct 85 13:52:25 PDT
From: admin@cogsci.Berkeley.EDU (Cognitive Science Program)
Message-Id: <8510232052.AA07435@cogsci>
To: allmsgs@cogsci.Berkeley.EDU, cogsci-friends@cogsci.Berkeley.EDU
Subject: UCB Cognitive Science Seminar--Oct. 29, 1985
Cc: admin@cogsci.Berkeley.EDU

                   BERKELEY COGNITIVE SCIENCE PROGRAM
                              Fall 1985
                  Cognitive Science Seminar - IDS 237A
                   Tuesday, October 29, 11:00 - 12:30
                     240 Bechtel Engineering Center
               Discussion: 12:30 - 1:30 in 200 Building T-4

                          ``Person Schemata''

                         Mardi J. Horowitz M.D.
                    Professor of Psychiatry, U.C.S.F.

          The speaker directs the recently formed Program  on  Cons-
     cious  and  Unconscious  Processes of the John and Catherine T.
     MacArthur Foundation.  Research on person schemata  is  one  of
     the core agendas of this program.
          After a brief description of the program,  the  discussion
     will  focus  on  clinical  phenomena  as segmented by different
     states of mind in a single individual.  By examining the confi-
     guration  in  each state of mind as it occurs over time, it may
     be possible to infer what the self schemata and role  relation-
     ship  models  are  that  organize thoughts, feelings and action
     into observed patterns.  The theory that forms  the  basis  for
     such  inferences  includes  the  postulate  that  each person's
     overall  self  organization  may  include  a  partially  nested
     hierarchy  of multiple self-concepts.  A frequent set of states
     of mind in pathological grief reactions will provide a concrete
     illustration  of  phenomena, methods of inference, and a theory
     of person schemata.
  ---------------------------------------------------------------------
     UPCOMING TALKS
        November  5: Edward Zalta, CSLI, Stanford
        November 12: Robert Wilensky, Computer Science, UCB
        November 19: Richard Alterman, Computer Science, UCB
        November 26: Eve Clark, Linguistics, Stanford
        December  3: Bernard Baars, Langley Porter, UCSF
   ---------------------------------------------------------------------
     ELSEWHERE ON CAMPUS
        John Dalbey, SESAME student, will present ``The Totally Effort-
        less  Problem  Solver''  at  the  SESAME  Colloquium on Monday,
        October 28, 4:00pm, 2515 Tolman Hall.

        Tom Wickens, U.C.L.A., will speak on ``Response Interactions in
        Visual  Detections''  at  the  Cognitive Psychology Colloquium,
        Friday, November 1, 4:00pm, Beach Room, 3105 Tolman Hall.

∂23-Oct-85  1733	EMMA@SU-CSLI.ARPA 	Newsletter October 24, No. 51  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 23 Oct 85  17:33:25 PDT
Date: Wed 23 Oct 85 16:56:02-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter October 24, No. 51
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
October 24, 1985                Stanford                       Vol. 2, No. 51
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, October 24, 1985

   12 noon		TINLunch
     Ventura Hall       ``A Problem for Actualism About Possible Worlds''
     Conference Room    by Alan McMichael
			Discussion led by Edward Zalta
			
   2:15 p.m.		CSLI Seminar
     Redwood Hall	Discourse, Intention, and Action
     Room G-19		Two talks given by Phil Cohen and Amichai Kronfeld
			
   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←
          CSLI ACTIVITIES FOR *NEXT* THURSDAY, October 31, 1985

   12 noon		TINLunch
     Ventura Hall       The Formation of Adjectival Passives
     Conference Room    by B. Levin and M. Rappaport
			Discussion led by Mark Gawron
			(Abstract on page 2)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	Foundations of Document Preparation
     Room G-19		David Levy, CSLI and Xerox PARC
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Redwood Hall	The Structure of Social Facts
     Room G-19		Prof. John Searle, Dept. of Philosophy, UC Berkeley

                              ←←←←←←←←←←←←
                               CORRECTION

      The coordinator for the Situation Theory and Situation Semantics
   (STASS) project is Jon Barwise not David Israel as stated in last
   week's newsletter.

!
Page 2                     CSLI Newsletter                   October 24, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                    ABSTRACT OF NEXT WEEK'S TINLUNCH
                  The Formation of Adjectival Passives
                        B. Levin and M. Rappaport

      This is Working Paper #2 in the MIT Lexicon Project, and though it
   discusses some rather specific issues having to do with one (putative)
   lexical rule of adjectival passive formation, it is an interesting
   example of the lexicon at work in a GB-style theory, worked out in
   unusual detail.  It assumes no knowledge of the Lexicon Group's work
   and only a minimal knowledge of GB.
      Since Wasow 1977 it has been standard among generative grammarians
   to assume two separate passivization rules, one for verbal passives,
   another for adjectival passives.  Levin and Rappaport argue against
   the claim in Wasow 1980 that the second of these rules has a thematic
   condition and propose an analysis of their own in which many of the
   standardly-cited facts about adjectival passives fall out simply from
   stipulating which arguments of a lexical item must be realized, and
   assuming that such lexical facts are in the default case preserved in
   the output of lexical rules.  We thus have another case in which
   thematic roles appear NOT to play the part they were claimed to play
   in a specific morphological or syntactic process.  Paradoxically,
   although the paper is set in a framework which assumes specific
   thematic roles, it presents an important negative result and casts
   further doubt on the hypothesis that thematic roles play a significant
   part in mediating the relation between syntax and lexical semantics.
		   					--Mark Gawron
                              ←←←←←←←←←←←←
                        NEXT WEEK'S CSLI SEMINAR
                   Foundations of Document Preparation

      Document preparation, by which I mean the use of the computer to
   prepare graphical presentations of verbal and pictorial information on
   screens and on paper, is inherently a linguistic activity.  This
   statement is true in two senses: Documents, first of all, are
   linguistic artifacts.  But in addition, the use of the computer as a
   marking tool is inherently linguistic: we *describe* to the computer
   the documents we wish to create.
      Current document preparation tools (the likes of TeX, Tedit, Emacs,
   Scribe, etc.) are highly inadequate and unnecessarily restrictive.
   This is because, I would claim, their designers have failed to take
   explicit account of the linguistic nature of document preparation:
   these tools have been built in advance of a theory of their subject
   matter.  In this talk, I will present an overview of research aimed at
   developing a ``theory of marking'' to serve as the foundation for the
   design of such tools.  I will set forth the broad outlines of the
   theory---one that lies at the intersection of a theory of production,
   a theory of representation, and a theory of marks---and will
   demonstrate that the issues of representation, reference, and action
   with which the Center is concerned are central to this research.  The
   bulk of the talk will be devoted to illustrating the search for
   founding concepts in the theory of marks---concepts such as figure,
   ground, region, and blueprint.  Such concepts are just as essential to
   a future linguistics of written forms as to a foundation for document
   preparation.					--David Levy
!
Page 3                     CSLI Newsletter                  October 24, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                       ENVIRONMENTS GROUP MEETING
             The Rational Programming Environment - Summary
                         Wolfgang Polak, Kestrel
               October 28, 1985, Ventura Trailer Classroom

      In 1981 Rational commenced work on an Ada oriented software
   development system.  The goal was to create a commercial system
   providing Lisp-style interactiveness and environment features for Ada.
   The project encompassed a language oriented machine architecture,
   specialized hardware, an integrated language based operating system
   and programming environment, and project management support tools.
      The original design used Ada's packages to create a hierarchy of
   nested structures corresponding to conventional directory systems.
   Permanent storage was provided by implementing persistent data objects
   in the language. Programs and data are simply declarations within the
   hierarchy of packages. Programs are only stored in internal
   representation; semantic consistency (according to language semantics)
   is maintained across the whole system. This organization allows
   powerful program manipulation and query tools to be implemented
   easily.
      While very uniform, the use of packages as directories with the
   associated semantic complexities proved cumbersome.  In later versions
   the directory structure was simplified and no longer subject to the
   exact language rules.
      The system is built around a powerful action mechanism. Any number
   of directory/object manipulations can be associated with an action.
   The action can later be committed, in which case all operations take
   effect, or the action can be abandoned, in which case all operations
   are undone.
      The user interacts with the system via a multi-window editor. Each
   window is of a particular type (e.g. text, program, status, etc.). The
   system includes a general structure oriented editor which combines
   structure operations with arbitrary text manipulation. Editor commands
   are uniform across all windows; only the effect of structure
   operations depends on the type of window.
      Fast incremental compilation facilitates both interactive program
   development and command execution.
                               ----------
                          PIXELS AND PREDICATES
                    ``Visual Programming Languages --
                From Visual Assembler to Rocky's Boots''
              Warren Robinett, with an assist by Scott Kim
             CSLI trailers, Wednesday, October 30, 1:00 p.m.

      A general view of the visual programming language problem is
   presented, anchored by two concrete examples.
      The first example is a visual assembly language, where patterns of
   pixels are interpreted as low-level instructions which manipulate
   patterns of pixels (and wherein one of the PnP themes is exemplified:
   a very primitive `predicate made from pixels').
      The second example is Rocky's Boots, a high-level visual
   programming language based on the building circuits metaphor
   (construed in some circles as an educational game).
!
Page 4                     CSLI Newsletter                  October 24, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                  SUMMARY OF ENVIRONMENTS GROUP MEETING
                            October 21, 1985

      Terry Winograd described research on an environment for use by
   people who are developing and modifying languages, and need to be able
   to produce and manipulate texts in those languages during this
   evolutionary phase.  It is based on a uniform way of treating grammars
   (based on a hierarchical phylum/operator structure with attributes),
   so that structure editing, structured storage and other facilities
   that are based on the language structure can be easily created and
   developed.
      He raised a number of issues that come up in trying to make the
   environment general (for at least a broad class of existing and
   envisioned languages), display-oriented (allowing dynamic changes of
   structure and view), incremental (dealing well with continual small
   updates), and distributed (multiple users cooperating in a
   heterogeneous not-totally-reliable networked environment).
      The current system is fragmentary and has not been integrated or
   written up.  Future talks by others in the group working on it will
   address some of the more specific technical issues.
                               ----------
                          CSLI SEMINAR SUMMARY
                       Ontology and Intensionality
                  Summary of CSLI Seminar on October 10

      In this seminar, I outlined two recent developments in the theory
   of abstract objects---one concerning ontology (the theory of times)
   and one concerning intensionality (a solution to the Morning
   Star/Evening Star puzzle).  Moments of time were identified as
   abstract objects, and truth at a time was defined in terms of the
   encoding relation.  Such definitions yielded the following non-trivial
   consequences: times are maximal and consistent with respect to the
   propositions true at them; there is a unique present time; a
   proposition is always true iff it is true at all times, every
   tense-theoretic consequence of a proposition true at a time is also
   true at that time.  In the second half of the seminar, we demonstrated
   that once one uses structured entities as the denotations of
   sentences, modal and tense contexts are not, in and of themselves,
   intensional.  Intensionality arises when definite descriptions appear
   in such contexts, and by assigning definite descriptions a second
   ``intensional'' reading, on which they denote the abstract object
   which encodes the properties they imply, we get a solution to the
   substitutivity puzzles which preserves our intuitions about the
   logical form of the sentences involved.             --Edward N. Zalta
!
Page 5                     CSLI Newsletter                  October 24, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                         SITUATED ENGINE COMPANY

      The STASS project has initiated a working group on the relation
   between situation theory and computation.  The aim is two fold: learn
   what needs to be added to situation theory to enable it to give
   adequate accounts of various computational agents, and to learn how we
   might be able to use computers in doing situation theory.  These two
   aims cause us to distinguish between sigma-machines and tau-machines.
   Sigma-machines are machines that are the subject matter for a
   situation-theoretic analysis.  Tau-machines are machines built to help
   do situation theory.
      In the long run, we expect that sigma and tau machines will merged,
   that our theory machines will also be our subject matter machines.
   For now, though, we are operating on two fronts simultaneously.  A
   simple robot, Gullible, has been designed and implemented by Brian
   Smith, Mike Dixon and Tayloe Stansbury.  It moves around on a grid,
   meeting people, picking up information (and misinformation) and
   answering certain questions about other people's locations based on
   what it has experienced on its travels.  This is to serve as our first
   sigma-machine.  Four groups have been formed to come up with
   semantical analysis of this robot using situation theory.
      On the other front, Jon Barwise has been lecturing about situation
   theory and its logic, to give a feeling for the basic theory, raising
   questions about what it might be reasonable to ask a computer to do,
   and coming up with some vague ideas about how one might get it to do
   it.
      The group meets every Tuesday at Xerox PARC, at 2 p.m., for about
   two hours.					--Jon Barwise
                                ---------
                        POSTDOCTORAL FELLOWSHIPS

      The Center for the Study of Language and Information (CSLI) at
   Stanford University is currently accepting applications for a small
   number of one year postdoctoral fellowships commencing September 1,
   1986.  The awards are intended for people who have received their
   Ph.D. degrees since June 1983.
      Postdoctoral fellows will participate in an integrated program of
   basic research on situated language---language as used by agents
   situated in the world to exchange, store, and process information,
   including both natural and computer languages.

      For more information about CSLI's research programs and details of
   postdoctoral fellowship appointments, write to:

        Dr. Elizabeth Macken, Assistant Director
   	Center for the Study of Language and Information
   	Ventura Hall
   	Stanford University
   	Stanford, California 94305

   APPLICATION DEADLINE: FEBRUARY 15, 1986

-------

∂24-Oct-85  0110	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #25  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  01:10:26 PDT
Date: 24 Oct 85 0108-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #25
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Thursday, 24 Oct 1985      Volume 1 : Issue 25

Today's Topics:

                 Parallel Logic Programming at Maryland


----------------------------------------------------------------------

Date: Wed, 23 Oct 85 15:04:10 EDT
From: Jack Minker <minker@mimsy.umd.edu>
Subject: Parallel Logic Programming at Maryland

            AI and Database Research Laboratory
                           at the
                   University of Maryland
                   Jack Minker - Director


     The AI and Database Research Laboratory at the  Univer-
sity  of  Maryland  consists of Jack Minker, the director of
the laboratory, and a number of faculty and students engaged
in  diverse  activities.   One  of the primary activities of
this group involves research into parallel problem  solving,
with emphasis on distributed logic programming.

     The AIDBRL members are developing a  logic  programming
system  for  the ZMOB multiprocessor, termed PRISM (parallel
inference system).  The work started in earnest in  approxi-
mately 1982.  PRISM has been implemented and is being tested
on a simulated system.  The underlying  ideas  behind  PRISM
appeared  in  [Eisinger  et.  al., 1982] and [Kasif et. al.,
1983].

     PRISM consists of several parts:  a user host interface
that  exists  on  the  host  machine; a set of Z80A machines
(moblets) designated as problem solvers  (PSMs);  a  set  of
machines  designated  as Extensional Database Machines (EDB)
that  store  ground  atomic  formulae  (relational  database
tables);  and,  a  set of machines designated as Intensional
Database Machines (IDB) that store procedures  (the  general
rules  in  the  system).   The  system  can  also optionally
include Constraint Machines (CM) that use user-supplied con-
straints to prune unsatisfiable paths in the proof tree.

     The system supports both AND and OR  parallelism.   The
user  can  specify control in terms of the sequence of atoms
to be executed in a set of problems to be solved.  Atoms can
be  executed in parallel, sequentially, or as specified by a
partial ordering.  Similarly procedures can be specified  as
being executed sequentially, in parallel, or as specified by
a partial order.  The PSM has  been  written  in  a  modular
fashion  to permit alternative control structure programs to
be incorporated in the system.  Alternative node and literal
selection algorithms may be incorporated as part of the con-
trol structure.  The  user  may  specify  the  configuration
(i.e., the number of moblets required as a minimum) in which
a problem is to be run.  If additional  moblets  are  avail-
able, the PRISM will automatically take advantage of them.

     A large number of problems  are  currently  being  pro-
grammed  in  PRISM and experiments will be run with these to
determine the effectiveness of PRISM as  a  problem  solving
system.

     The major research directions in  the  laboratory  over
the coming year will be devoted to the following areas:

    (1)  Experimentation Using PRISM
    (2)  Control Structure Investigations
    (3)  Expert systems and PRISM
    (4)  Parallel problem solving and Architecture Issues


     If you would like further information on PRISM,  please
contact  MINKER@MIMSY.UMD.EDU  or MADHUR@MIMSY.UMD.EDU    We
would also be very interested in hearing from people who may
have problems we could run on PRISM.

References:

1.   Eisinger, N., Kasif, S., and Minker,  J.,  "Logic  Pro-
     gramming:  A  Parallel Approach", in Proceedings of the
     First International Logic Programming Conference,  Mar-
     seilles, France, 1982.

2.   Kasif, S., Kohli, M., and Minker, J., "PRISM - A Paral-
     lel Inference System for Problem Solving", in IJCAI-83,
     Karlsruhe, Germany, 1983.

------------------------------

End of PARSYM Digest
********************

∂24-Oct-85  0819	EMMA@SU-CSLI.ARPA 	Today's CSLI seminar 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  08:19:37 PDT
Date: Thu 24 Oct 85 08:20:04-PDT
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Today's CSLI seminar
To: friends@SU-CSLI.ARPA
Tel:  497-3479


  The titles of the two talks to be given by Phil Cohen and Ami Kronfeld
in today's 2:15 seminar are

Speech Acts and Rationality 
    Phil Cohen

The Referential/Attributive Distinction
   Ami Kronfeld


  As usual, the abstracts can be found in last week's newsletter.

-------

∂24-Oct-85  0847	AAAI-OFFICE@SUMEX-AIM.ARPA 	New Tutorial Chair    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  08:47:03 PDT
Date: Thu 24 Oct 85 08:45:05-PDT
From: AAAI <AAAI-OFFICE@SUMEX-AIM.ARPA>
Subject: New Tutorial Chair
To: officers: ;
cc: aaai-OFFICE@SUMEX-AIM.ARPA
Telephone: (415) 328-3123
Postal-Address: 445 Burgess Drive, Menlo Park, CA 94025
Message-ID: <12153692175.16.AAAI-OFFICE@SUMEX-AIM.ARPA>


Mark Stefik has agreed to be next year's Tutorial Chair.  The
Executive Council, by affirmative majority vote, needs to approve
his appointment.  So we will need your vote as soon as possible
to commence our preparations for next year.

Regards,
Claudia


_
-------

∂24-Oct-85  1436	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  14:35:56 PDT
Date: Thu 24 Oct 85 14:36:39-PDT
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12153756177.24.LIBRARY@SU-SCORE.ARPA>

Parallel MIMD Computation: The HEP supercomputer and Its Applications.
ed. by J. S. Kowalik.  MIT Press.  QA76.8.D436P37 1985

Tutorial On Software Maintenance. by Girish Parikh and Nicholas Zveginitzov.
IEEE Computer Society.  QA76.9.S65T87 1983.

Workshop On Distributed Computing. University of Newcastle Upon Tyne.
Computing Lab. S. K. Shrivastava.  (8508666)

Real-Time Clinical Computing by Ian Perry. R853.D37P46 1984

Principles of Database Design. Volume 1 Logical Organizations. edited by
S. Bing Yao.  QA76.9D3p73 1985 V. 1.

IEEE Standard Microcomputer System Bus. IEEE Std. 796-1983.  
TK7895.B87158 1983.

Data Communicatons Software Design. by Malcolm Lane.  TK5105.L357 1985.

Directory Of Statistical Microcomputer Software. 1985 Edition. by 
Woodward, Elliott, and Gray.   (8512454)

Stoyan, Herbert.  Maschinen-unabhangige Code-Erzeugung als Semantikerhaltende
Beweisbare Programmtransformation.  (8512466)

Computability With Pascal by Mallozzi and DeLillo. QA9.59.M34 1984.

A Design For Microprocessor-Based Systems. by Peels  (8513505)

Data Structures With Ada. by Michael Feldman. Reston Publishing Co.
QA76.73.A35F45 1985.

Advanced UCSD Pascal Programming Techniques. by Willner and Demchak.
QA76.73.U25W55 1985.

The 3-D Animated Apple by Phil Cohen.  T385.C543 1983.

Harry Llull
-------

∂24-Oct-85  1514	ullman@su-aimvax.arpa 	Dave Smith's conjecture    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  15:14:02 PDT
Received: by su-aimvax.arpa with Sendmail; Thu, 24 Oct 85 15:10:36 pdt
Date: Thu, 24 Oct 85 15:10:36 pdt
From: Jeff Ullman <ullman@diablo>
Subject: Dave Smith's conjecture
To: de2smith@sumex, genesereth@sumex, nail@diablo

Perhaps I should formulate my view of Dave Smith's conjecture
more carefully, to avoid having to pay out that dollar
needlessly.
My understanding of the model is that we wish to compute the
join of several database relations over a universal set of
attributes (i.e., the variables of the conjuncts).
We are given a cost function C(R1,...,Rk; R), that gives
the cost of joining the relation R with the already-computed
join of R1,...,Rk.
The problem is to minimize the sum
	m
     SIGMA	C(S1,...,S{i-1}; Si)
       i=1
over all permutations S1,...,Sm of the m given relations.

First, note that the input size in general is exponential in m,
making it unlikely that the problem is NP-complete, despite the
reference to my book in support of that claim.
However, there are restricted models, where the cost function C
can be encoded in some succinct way.
The simplest of these is one where there is a value n(R)
associated with each relation R, and C(R1,...,Rk; R) is
just the product of n(R1)*...*n(Rk)*n(R).
This case is too simple, and it is easy to pick an optimal
order.  It is also easy to show Smith's conjecture in this case.

Perhaps an intermediate-difficulty assumption is that
C(R1,...,Rk; R) is the product of
1. C(R1,...,R{k-1}; Rk), i.e., the cost of the previous step, and
2. A constant n(R,f) that depends only on R and the number of
	attributes of R that do not appear in R1,...,Rk (i.e.,
	the number of free variables when we reach R).
Under this assumption, the input data is proportional to the
product of the number of relations and the maximum number of
attributes in a relation, which is polynomial in what we
would intuitively regard as the "size" of the input, i.e.,
the length of the relation schemes written out.

Now, at least, we have some hope of showing NP-completeness for
the optimality problem.
I also think we can attack Smith's conjecture: if we have
two sets of relation R1,...,Rm and T1,...,Tn, that do not
share any attributes, is their optimal order always an
interleaving of the optimal orders for the two sets?

I'll stick with my belief that it is not, at least for the
general cost model, probably for the intermediate-difficulty
model.  In fact, the result is trivially true if we allow
C to be arbitrary, because once we have selected an R and a T
in our join, we are exercising parts of the C function that
were not exercised in finding the optimal orders for the R's or T's
separately, and ANYTHING can happen.
However, let's be fair, and assume that the cost
C(S1,...,Sk; S), where S is among the R's, is the product of
1. C(S1',...,Sp'; S), where S1',...,Sp' is the subsequence of S1,...,Sk
that are among the R's, and
2. A constant depending only on the subsequence of T's among S1,...,Sk.

Even with this assumption, we could suppose that the optimal
sequence for joining three R's has costs of 1, 50, 50, while
there is another sequence that has costs 1, 10, 100.
In isolation, we may prefer the former.
However, it also could be that when we interleave these with T's,
the constants (2-above) associated with the T's for a particular
interleaving are 1, 300, 200.
Then the first sequence has a cost of 25,001 and the second has 23,001.
Of course, things are not so simple, because we have to figure
in the cost of the T's, and we also have to recognize that this
may not be the optimal interleaving, all things considered.
Nonetheless, I think this point can be developed into a
concrete counterexample.
				---Jeff

∂24-Oct-85  1602	BETSY@SU-CSLI.ARPA 	Visitor Policy 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  16:02:24 PDT
Date: Thu 24 Oct 85 15:57:35-PDT
From: Betsy Macken <BETSY@SU-CSLI.ARPA>
Subject: Visitor Policy
To: folks@SU-CSLI.ARPA

We've been getting some questions that made me think that it might be
useful to send this statement about our current visitor policy again:


                  VISITORS POLICY FOR 1985-86


Non-casual visitors to CSLI fall roughly into three groups:

(1)  "Collaborators"; those whose visit is actively sought by a CSLI 
     researcher(s), for purposes of collaboration in some CSLI
     project, or at any rate to work closely with such researcher(s).
     Often travel money and other support is supplied from initiators
     or area funds.

(2)  "Sojourners"; those who are here mostly of their own volition,
     finding it convenient to spend a sabbatical or other leave in
     the midst of the resources offered here.  Support is limited to
     computer accounts and possibly shared office space.

(3)  Special cases.


Visitors bring a great deal to CSLI, and independently of that we have
some obligation to share our good fortune.  However, there are many
hidden and not so hidden costs.  To keep things in balance, I intend
to adopt the following policies, mostly modest revisions of the
policies of the past.

Collaborators will be most welcome, of course.  But:

a)   We ask that the details of an impending visit, the duration, the
     facilities needed, etc., be made available before the visit is
     too impending.  There are sudden opportunities, of course, but
     in general a couple of weeks warning, via the forms available
     from Bach-Hong, Elsie, or Jackie, seems reasonable.  Of course,
     the less the needs of the visitor, the less this matters.

b)   We will allocate a limited amount of space at Ventura for
     Collaborators; if this space is taken for the period in question,
     by the time the details reach us, we may be unable to provide
     any.

c)   Please discuss the details of any proposed long-term visits
     (much more than a month, say) with Betsy well in advance.


As to Sojourners, I think we would all agree that we are better off
with a few that we can treat pretty well, rather than a lot we don't
have time, space, or xerox paper for.  So, we will let Sojourner
applications for Visiting Scholar status during a given academic year
(June to June) accumulate (except for special cases) until March 15 of
the previous year.  Decisions will then be made on the basis of
projected available space and the credentials and projects of the
visitors.  Thus Visiting Scholar status at CSLI will be somewhat more
meaningful, and somewhat less easy to obtain than the same status in
departments at Stanford.  We will be nice to everyone, but only
committed to contributing to the space and computational needs of our
own Visiting Scholars.  If you receive a request from someone wanting
to be a Sojourner, send it to Ingrid to be added to our applications
file.

Special cases.  E.g., people whom we more actively encourage to come,
perhaps even providing some salary support or travel money.  These
will be treated as special cases.


-------

∂24-Oct-85  1724	NILSSON@SU-SCORE.ARPA 	[Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  17:24:29 PDT
Date: Thu 24 Oct 85 17:25:17-PDT
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]
To: faculty@SU-SCORE.ARPA
Message-ID: <12153786875.16.NILSSON@SU-SCORE.ARPA>


What do we think about this?  -Nils
                ---------------

Mail-From: TAJNAI created at 24-Oct-85 12:46:50
Date: Thu 24 Oct 85 12:46:50-PDT
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: Discussion on limiting size of Forum
To: nilsson@SU-SCORE.ARPA, jlh@SU-SONOMA.ARPA
Message-ID: <12153736184.13.TAJNAI@SU-SCORE.ARPA>


The following is a message that Terry and I drafted.  You may want to
bring this issue to the attention of CSD/CSL faculty.
......................................................................

Two years ago the Forum Committee recommended that we limit membership
to 50 companies.  The CSD/CSL Faculty disagreed and preferred to  keep
membership open.

Membership is now  72.  The  Forum Committee is  again discussing  the
possibility of limiting membership.  Faculty time is the resource that
seems to be the limiting factor.

Statistics for the past 4 years are as follows:

Number of Companies that belonged as of the annual meeting:

1981/82		29
1982/83		40
1983/84		55
1984/85		70

The Forum Committee is interested in your comments.

-------
-------

∂24-Oct-85  2129	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #26  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  21:29:23 PDT
Date: 24 Oct 85 2121-PDT
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #26
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Friday, 25 Oct 1985       Volume 1 : Issue 26

Today's Topics:

         Second PARSYM Survey: Data Abstractions (continued!)

----------------------------------------------------------------------

Subject: Second PARSYM Survey: Data Abstractions
Date: Thu, 24 Oct 85 15:08:30 EDT
From: Paul Hudak <Hudak@YALE.ARPA>

Project: Para-Functional Programming: A Paradigm for Programming
         Multiprocessor Systems

People: Paul Hudak, Lauren Smith  (hudak@yale, smith-lauren@yale)

Place: Yale Department of Computer Science


To respond to this survey requires a little background material:

The main thesis of our work is that the parallelism in a purely
functional program is *implicit*, not needing any special notation
as is the case, for example, in imperative languages modified for
parallel computing.  On the other hand, functional languages (or any
other languages, for that matter) do not generally provide ways to
*map* programs to particular multiprocessor topologies.  Para-functional
programming provides this mapping ability via *annotations* to the
source program.  One nice thing about this approach is that the
annotations can be made in such a way that the correctness of the
original program is preserved -- one can reason about the mapping
issues separately from issues of functional behavior.

There are two simple ideas:
1) the expression "exp $on pid" says evaluate exp on processor pid.
2) there is a dynamic variable $self that evaluates to the "currently
   executing processor."
As described below, these annotations allow one to map data structures
to particular multiprocessor topologies.  When used with recursion,
a surprisingly few number of annotations are generally needed.

    1.  Are you using or have you developed data structures or
        abstractions specialized for parallel computing?  Please
        describe.

Partly because we wish to solve numerical as well as symbolic problems,
and partly because we want high degrees of parallelism, we have a
data structure called a "parallel array."  The expression "mkv(n,f)"
creates a vector of length n whose i'th component has value f(i).
If I want to distribute that vector among a set of processors, I can
do so by mapping the body of f as I see fit.  For example, by using:
    f(i) = exp $on i
I am essentially placing the ith element of the vector on processor i.
Simple generalizations of this notion allow mapping arrays to meshes,
trees to hypercubes, etc.

    2.  In particular, do you use specialized data abstractions to
        represent configurations of multiple processes?  Please
        describe.

As implicated above, our goal is generality, and thus we can tailor
most data structures to arbitrary topologies -- all it assumes is
some initial labeling (we use integers) of the processors.  For example,
suppose I number a ring of n processors by 0,1,...,n-1.  Then I can
create a function "map" that applies a function to each element of a
list, mapping the result sequentially to the processors around the ring:

    map(f,l) = if null(l)
               then ()
               else cons( f(car(l)),
                          map(f,cdr(l)) $on right($self) )
    right(p) = (p+1) mod n

This, of course, depends on the semantics of cons.  If it is a
"parallel" cons then everything is OK.  If it is a "lazy" cons then
we provide an additional annotaion, #, that basically says "do this
eagerly."  So the last two lines of map might say:

           ... else cons( #f(car(l)),
                          #map(f,cdr(l)) $on right($self) )

# does *not* make the function strict; it just allows the computation
to proceed in parallel, and thus admits non-terminating *sub*-computations.

    3.  Do you use specialized data structures, a la FRONS, for
        representing non-deterministic collections of data?  Please
        describe.

No.  However the notion of non-determinism is not incompatible with
the ideas presented above, and the "standard" ways with which it is
dealt with in functional languages (for example FRONS or non-
deterministic merge) would seem to work just fine.

    4.  What are some useful references on the topic of data abstraction
        in parallel computing (including your own)?

For para-functional programming:

    Hudak, P. and Smith, L., "Para-functional Programming: A Paradigm
    for Programming Multiprocessor Systems," to appear in Proceedings
    of ACM Symposium on Principles of Programming Languages, January 1986.

    An earlier version of the above paper is available as a tech report
    from Yale: YALEU/DCS/TR-390, January 1985.

Papers by Warren Burton and Udi Shapiro are also relevant; they're
referenced in the above.

        5.  Other comments welcome.

The paper also talks about some obvious generalizations of the basic
ideas, such as control of operating system resources.  It also gives
a formal denotational semantics of the mapping behavior, using a notion
of "execution trees," which seems to be useful as a step toward building
interpreters and tools for debugging.

------------------------------

End of PARSYM Digest
********************

∂24-Oct-85  2231	BRESNAN@SU-CSLI.ARPA 	"Case-Assignment by Nominals in Japanese"  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 24 Oct 85  22:31:03 PDT
Date: Thu 24 Oct 85 22:28:17-PDT
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: "Case-Assignment by Nominals in Japanese"
To: folks@SU-CSLI.ARPA

On Thursday, October 31, at 10:00 a.m. in the Ventura Hall
Conference Room, Masayo Iida will present her work abstracted
below.  Written copies of the paper will be available on Monday
at the Ventural Hall Receptionist's desk.

Abstract:

Case-Assignment by Nominals in Japanese

In this paper I will discuss certain peculiar properties of a class of
Japanese deverbal nominals, which show verb-like properties in certain
environments: specifically, they assign verbal case and can be modified by
adverbs (`verbal case' includes nominative, accusative and dative, i.e.,
cases normally assigned by a verb).  These case-assignment phenomena pose a
problem for current syntactic theories, which assume that verbs alone
assign such cases, while nouns do not.  Now I have observed that a deverbal
nominal assigns verbal case only when it is concatenated with a suffix
bearing temporal information, which might be encoded with the feature
[+aspect].  The nominal assigns case when the following two conditions are
satisfied: (i) the nominal has a predicate-argument structure, and (ii) it
is concatenated with a suffix which bears an aspectual feature.  I will
propose that (syntactic) category membership is not sufficient for
determining properties of case-assignment, adverb distribution, etc., and
suggest that the factors (i) and (ii) are perhaps more relevant.

--Masayo Iida
-------

∂25-Oct-85  0030	DAVIES@SUMEX-AIM.ARPA 	Wednesday meeting -- 9 am  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  00:30:51 PDT
Date: Fri, 25 Oct 1985  00:32 PDT
Message-ID: <DAVIES.12153864695.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: Wednesday meeting -- 9 am

I will talk about CAREL (for CARE-Lisp), a distributed-memory
multiprocessor Lisp.  CAREL is designed to exploit the capabilities of
the CARE architecture and to accurately reflect the costs of
communicating data and code during execution of programs.

	-- Byron

∂25-Oct-85  0935	BARWISE@SU-CSLI.ARPA 	Logic Seminar Cancelled today    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  09:34:53 PDT
Date: Fri 25 Oct 85 08:01:44-PDT
From: Jon Barwise <BARWISE@SU-CSLI.ARPA>
Subject: Logic Seminar Cancelled today
To: BBoard@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA

Dag Westersthal's talk in the logic seminar, scheduled for today, is
postponed, due to Dag's sudden illness.  
-------

∂25-Oct-85  0938	AAAI-OFFICE@SUMEX-AIM.ARPA 	Mark's confirmation   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  09:37:55 PDT
Date: Fri 25 Oct 85 09:00:51-PDT
From: AAAI <AAAI-OFFICE@SUMEX-AIM.ARPA>
Subject: Mark's confirmation
To: officers: ;
cc: aaai-OFFICE@SUMEX-AIM.ARPA
Telephone: (415) 328-3123
Postal-Address: 445 Burgess Drive, Menlo Park, CA 94025
Message-ID: <12153957189.47.AAAI-OFFICE@SUMEX-AIM.ARPA>


Yesterday, we received a majority vote confirming Mark's appointment.

- Claudia

_
-------

∂25-Oct-85  1018	INGRID@SU-CSLI.ARPA 	House for Rent
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  10:13:51 PDT
Date: Fri 25 Oct 85 10:09:48-PDT
From: Ingrid Deiwiks - 497-3084 <INGRID@SU-CSLI.ARPA>
Subject: House for Rent
To: Folks@SU-CSLI.ARPA

Charming one bedroom house in College Terrace for rent.  Tastefully 
decorated with antiques, hardwood floors, oriental carpets, washer/dryer,
garbage disposal, dishwasher, low maintenance yard, patio.  $1,000 per
month plus deposit.  Term of rent: two months to one year, negotiable.
Please call 857-0676 after 1 pm.

PLEASE DO NOT REPLY TO THIS MESSAGE, BUT CALL THE ABOVE NUMBER.
-------

∂25-Oct-85  1032	INGRID@SU-CSLI.ARPA 	House for Rent
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  10:20:09 PDT
Date: Fri 25 Oct 85 10:16:36-PDT
From: Ingrid Deiwiks - 497-3084 <INGRID@SU-CSLI.ARPA>
Subject: House for Rent
To: Folks@SU-CSLI.ARPA

Three and a half bedroom house with two baths for rent (to family
only).  Completely new kitchen with all the gadgets.  Garden.  Located
in Palo Alto, in walking distance from the main library and two parks.
Very close to schools.  $1,300 per month.  Term of rent: two months,
beginning November 11.
Please call 322-0187.

PLEASE DO NOT REPLY TO THIS MESSAGE, BUT CALL THE ABOVE NUMBER.
-------

∂25-Oct-85  1036	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	todays logic seminar is cancelled  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  10:32:10 PDT
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Fri 25 Oct 85 10:30:30-PDT
Date: 25 Oct 85  1013 PDT
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: todays logic seminar is cancelled 
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    


The seminar on logic and the foundations of mathematics to be
given by Dag Westerstahl today (Friday) at noon has been canceled
due to illness of the speaker.  It will be rescheduled. 


∂25-Oct-85  1036	RICHARDSON@SU-SCORE.ARPA 	November Sr. Faculty Meeting 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  10:35:24 PDT
Date: Fri 25 Oct 85 10:37:22-PDT
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: November Sr. Faculty Meeting
To: tenured@SU-SCORE.ARPA
Message-ID: <12153974759.22.RICHARDSON@SU-SCORE.ARPA>

The sr. faculty meeting ordinarily scheduled for the first week of November
will be postponed until the first week of December since there are no
pressing matters to discuss.
-------

∂25-Oct-85  1106	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  11:05:31 PDT
Received: from SUMEX-AIM.ARPA by SU-SCORE.ARPA with TCP; Fri 25 Oct 85 11:06:45-PDT
Date: Fri 25 Oct 85 11:06:37-PDT
From: Edward Feigenbaum <FEIGENBAUM@SUMEX-AIM.ARPA>
Subject: Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]
To: NILSSON@SU-SCORE.ARPA, faculty@SU-SCORE.ARPA
In-Reply-To: <12153786875.16.NILSSON@SU-SCORE.ARPA>
Message-ID: <12153980085.49.FEIGENBAUM@SUMEX-AIM.ARPA>

The following comment is intended to be practical, not cynical.

We have a waning of interest in industrial-liason "supply". We have seen a
gradual waxing of industrial-liason "demand". These are not now at
a comfortable equilibrium. They can be brought into equilibrium by the
normal method used for adjusting supply and demand: appropriate price
setting. 

I suggest that the Forum price be raised 20% effective as soon as we can
reasonably do it, and 10% per year thereafter until the number of
members is "comfortable" for our faculty to support.

Ed
-------

∂25-Oct-85  1112	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  11:12:52 PDT
Received: from SUMEX-AIM.ARPA by SU-SCORE.ARPA with TCP; Fri 25 Oct 85 11:13:49-PDT
Date: Fri 25 Oct 85 11:13:41-PDT
From: Edward Feigenbaum <FEIGENBAUM@SUMEX-AIM.ARPA>
Subject: Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]
To: NILSSON@SU-SCORE.ARPA, faculty@SU-SCORE.ARPA
In-Reply-To: <12153980085.49.FEIGENBAUM@SUMEX-AIM.ARPA>
Message-ID: <12153981372.49.FEIGENBAUM@SUMEX-AIM.ARPA>

P.s. I believe that either the Chemistry for Biochemistry industrial-
affiliates program is at $25,000 per year!........Ed
-------

∂25-Oct-85  1118	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  11:18:25 PDT
Received: from SUMEX-AIM.ARPA by SU-SCORE.ARPA with TCP; Fri 25 Oct 85 11:14:26-PDT
Date: Fri 25 Oct 85 11:14:19-PDT
From: Edward Feigenbaum <FEIGENBAUM@SUMEX-AIM.ARPA>
Subject: Re: [Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>: Discussion on limiting size of Forum]
To: NILSSON@SU-SCORE.ARPA, faculty@SU-SCORE.ARPA
In-Reply-To: <12153980085.49.FEIGENBAUM@SUMEX-AIM.ARPA>
Message-ID: <12153981486.49.FEIGENBAUM@SUMEX-AIM.ARPA>

oops, typo in the above. I meant "Chemistry OR Biochemistry..."
-------

∂25-Oct-85  1304	INGRID@SU-CSLI.ARPA 	Internal Faculty Fellowships 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  13:04:21 PDT
Date: Fri 25 Oct 85 13:01:16-PDT
From: Ingrid Deiwiks - 497-3084 <INGRID@SU-CSLI.ARPA>
Subject: Internal Faculty Fellowships
To: Folks@SU-CSLI.ARPA


                   STANFORD HUMANITIES CENTER

      Announcement of Internal Faculty Fellowships: 1986-87

The Stanford Humanities Center will award as many as twelve Faculty
Fellowships for the academic year 1986-87.  Of these, up to six will
be awarded to members of the Stanford faculty.  (A similar number will
also be awarded to applicants from elsewhere, in two categories: 
(a) fellowships for already well-established and usually tenured
scholars; (b) fellowships for junior, usually untenured, scholars who
teach at colleges or universities which do not have major graduate
schools or do not have doctoral programs in their own departments.)

Application forms and further information are available from
Ingrid@su-csli, or the Stanford Humanities Center, Mariposa House,
Stanford University, Stanford, CA 94305.  Applications must be
completed by January 6, 1986.  Awards will be announced not later than
mid-March 1986.
-------

∂25-Oct-85  1433	ullman@su-aimvax.arpa 	papers received  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  14:33:33 PDT
Received: by su-aimvax.arpa with Sendmail; Fri, 25 Oct 85 14:30:07 pdt
Date: Fri, 25 Oct 85 14:30:07 pdt
From: Jeff Ullman <ullman@diablo>
Subject: papers received
To: nail@diablo

1. Parallel Evaluation of Recursive Rule Queries by Cosmodakis and
	Kanellakis.  This is the paper on sYryPS.
2. A State Transition Model for Distributed Query Processing by
	LaFortune and Wong.

∂25-Oct-85  1559	MODICA@SU-SCORE.ARPA 	Class Lists  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Oct 85  15:57:43 PDT
Date: Fri 25 Oct 85 15:58:22-PDT
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: Class Lists
To: instructors@SU-SCORE.ARPA
Message-ID: <12154033196.35.MODICA@SU-SCORE.ARPA>

Class Lists are still here and eagerly waiting to be picked up.
                            See you soon,

                           Gina (MJH room 251)
-------

∂26-Oct-85  1019	BRESNAN@SU-CSLI.ARPA 	"Case-Assignment by Nominals in Japanese"  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 26 Oct 85  10:19:15 PDT
Date: Sat 26 Oct 85 10:17:41-PDT
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: "Case-Assignment by Nominals in Japanese"
To: folks@SU-CSLI.ARPA

Because of difficulties with the Xerox machine, Masayo Iida's
paper may not be available for distribution on Monday.  Please
wait for further notice.
-------

∂27-Oct-85  1418	NILSSON@SU-SCORE.ARPA 	Student names    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Oct 85  14:18:29 PST
Date: Sun 27 Oct 85 14:19:58-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Student names
To: faculty@SU-SCORE.ARPA
Message-ID: <12154550492.21.NILSSON@SU-SCORE.ARPA>

Apparently our CSD has had the policy of not releasing to 
industrial inquirers names of our students who are expected to
graduate.  (Of course, any company who joins the forum gets lots
of info about our students--and about many other things as well.)
If the policy is merely to protect the Forum, it sounds to me
like it's a bit extreme.  We can hardly expect a company to pay $12K
per year just because they would like to get a list of names.

Maybe the policy was put in place to protect students from being pestered
by employers.  Anyway, we can't seem to find mention of an official 
policy--perhaps it's a very old one.

What do people think.  I lean toward being more lenient about this, unless
I get an outpouring of sentiment against it.  I'm particularly interested
in student comments.  -Nils
-------

∂27-Oct-85  1447	NILSSON@SU-SCORE.ARPA 	Committees  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Oct 85  14:47:09 PST
Date: Sun 27 Oct 85 14:48:17-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Committees
To: faculty@SU-SCORE.ARPA
Message-ID: <12154555647.21.NILSSON@SU-SCORE.ARPA>

There have been some minor reshufflings of our committees.
I believe everyone involved in the "shuffle" knows about it,
but for the record, here's the latest chart showing who is on
what committees.  -Nils


                        1985-1986 CSD Committees

Key:  x = new member                      F = fall quarter member 
      X = member who was on               W = winter quarter member
          this committee last year        S = spring quarter member        
      c = chairman              
      C = chairman who was on this committee last year

                   | A | C | C | C | F | F |1st| F |Ind |Lib|MS| MS |Math|PhD |
                   | d | o | o | u | a | e | Yr| o |Prof|Pub|AI|Prog|Sci |Prog|
                   | m | m | l | r | c | l |Adv| r |    |   |  |    |    |    |
                   | i | p | l | r | i | l |   | u |    |   |  |    |    |    |
                   | s |   | o | i | l | o |   | m |    |   |  |    |    |    |
                   |   |   | q | c |   | w |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Binford, T.        |   |   |   |   | X |   |   |   |    |   |X | x  |    |    |
-------------------------------------------------------------------------------
Bosack, L.         |   | x |   |   | X |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Buchanan, B.       |   |   |   |   |   |   |   |   |    | C |C |    |    |    |
-------------------------------------------------------------------------------
Cheriton, D.       | X |   |   |   | X |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Clancey, W.        |   |   |   |   |   |   |   |   |    |   |X |    |  x |    |
-------------------------------------------------------------------------------
Earnest, L.        |   |   |   |   | C |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Feigenbaum, E.     | x |   | F |   |   |   |   |   |    |   |X |    |    |    |
-------------------------------------------------------------------------------
Floyd, R.	   |   | x |   |   |   |   |   |   |    |   |  | X  |    |    |
-------------------------------------------------------------------------------
Genesereth, M.     |   | x |   | X | x |   |   |   |    |   |X |    |    |    |
-------------------------------------------------------------------------------
Golub, G.	   |   | X |   |   |   |   |   |   |    |   |  |    |  X |    |
-------------------------------------------------------------------------------
Grosz, B.          | X |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Guibas, L.	   |   |   |   |   |   |   |   |   |    |   |  | x  |    |  x |
[1/2 LWOS '85-'86] |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Hennessy, J.       |   |   |   |   |   |   |   |   |    |   |  | x  |    |    |
-------------------------------------------------------------------------------
Herriot, J.        |   |   |   |   |   |   |   |   |    |   |  |    | X  |    |
-------------------------------------------------------------------------------
Katevenis, M.	   |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
[LWOS '85-'86      |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
'86-'87]           |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Knuth, D.	   |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
[Sabbat. '85-'86]  |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Lantz, K.	   |   | x | S | x |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Manna, Z.	   |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
[LWOS '85-'86]     |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Mayr, E.	   |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
[Sabbat. Fall '85  |   |   | W | x |   |   |   |   |    |   |  |    | C  |    |
Spring '86]        |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
McCarthy, J.	   |   |   |   |   | X |   |   |   |  C |   |  |    |    |  x |
-------------------------------------------------------------------------------
!
                   | A | C | C | C | F | F |1st| F |Ind |Lib|MS| MS |Math|PhD |
                   | d | o | o | u | a | e | Yr| o |Prof|Pub|AI|Prog|Sci |Prog|
                   | m | m | l | r | c | l |Adv| r |    |   |  |    |    |    |
                   | i | p | l | r | i | l |   | u |    |   |  |    |    |    |
                   | s |   | o | i | l | o |   | m |    |   |  |    |    |    |
                   |   |   | q | c |   | w |   |   |    |   |  |    |    |    |

McCluskey, C. 	   | x |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Miller, W.	   |   |   |   |   |   |   |   | C |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Nilsson, N.        |   |   | F |   |   | x | x |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Oliger, J.         |   |   |   |   |   |   |   |   |    |   |  |  C |    |    |
-------------------------------------------------------------------------------
Papadimitriou, C.  | X |   |   |   |   |   |   |   |    |   |  |    |    |    |
[Sabbat. Spr. '86] |   |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Pratt, V.	   |   | C |   |   | x |   |   |   |    |   |  |    |    |  x |
-------------------------------------------------------------------------------
Reges, S.          |   |   |   | x |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Reid, B.           |   |   |   | x |   |   |   |   |    |   |  |    |    | x  |
-------------------------------------------------------------------------------
Rindfleisch, T.    |   |   |   |   | x |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Roberts, Eric (DEC)| x |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Rosenbloom,  P.    |   |   |   |   |   |   |   |   |    |   | X|    |    |  x |
-------------------------------------------------------------------------------
Rosenschein, S.    |   | x |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Tajnai, C.         |   |   |   |   |   | C |   | X |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Ullman, J.	   |   |   |   | C |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Wiederhold, G.     |   |   |   |   |   |   |   |   |    |   |  |  X |  x |    |
-------------------------------------------------------------------------------
Winograd, T.       |   |   |   |   |   |   |   | X |    |   |  |    |    |  c |
-------------------------------------------------------------------------------
Yao, A.		   | c |   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
CSL Volunteers     |   |   | W |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
Industr. Vols.     |  x|   |   |   |   |   |   |   |    |   |  |    |    |    |
-------------------------------------------------------------------------------
No. of Students:   | 2 | 6 | 1 | 3 | 3 | 1 |   | 2 |    | 1 | 2| 2  |    |  2 |
-------------------------------------------------------------------------------

!
Student Committee Members:

;; Key:	[] means that the committee assignment was made prior to 9/85
;;	{} indicates a slot that is not ordinarily on the committee

Admissions:
(1) Mike Dixon (MDIXON@SUSHI)
(2) Ramsey Haddad (HADDAD@SUSHI)
(3) Haym Hirsh (HIRSH@SUMEX)

Comprehensive Exam:
[AA] Allan Van Gelder (AVG@DIABLO)
[AI] Stuart Russell (RUSSELL@SUMEX)
(HW) & (SW) Jeff Naughton & Miriam Blatt
[MTC] Mike Dixon (MDIXON@SUSHI)
(NA) Ray Tuminaro

Colloquium (juice and cookies):
(1) Steve Vavasis (VAVASIS@SUSHI)

Curriculum:
[1] Anil Gangolli (GANGOLLI@SUSHI)
[2] Kathleen Kells (KELLS@SCORE)
{[3]} Devika Subramanian (SUBRAMANIAN@SUMEX)

Facilities:
(1) Joel Bion (BION@SCORE)
(2) Steve Tjiang (TJIANG@SCORE)
(3) Bruce Hitson [EE] (HITSON@SCORE)

Fellowships:
(1) William Lees (LEES@SUSHI)

Forum:
(1) Karen Pieper (PIEPER@SUSHI)
[2] Karen Scholz (SCHOLZ@SUSHI)

Library/Publications:
(1) Alejandro Quiroz (QUIROZ@SUSHI)

MSAI:
[1] Don Henager (HENAGER@SUMEX)
[2] Haym Hirsh (HIRSH@SUMEX)
(3) Naomi Rodolitz (RODOLITZ@SUMEX)

MS Program:
(1) Dave Fogelsong (FOGELSONG@SUSHI)
(2) James Fu (FU@SUSHI)

Evaluation of PhD program requirements:
(1) Eric Berglund (BERGLUND@PESCADERO)
(2) Peter Karp (KARP@SUMEX)
{3} Karen Scholz (SCHOLZ@SUSHI)
-------

∂27-Oct-85  1849	LANSKY@SRI-AI.ARPA 	PLANLUNCH REMINDER - Pat Hayes, Oct. 28 
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 27 Oct 85  18:49:18 PST
Date: Sun 27 Oct 85 18:48:47-PST
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH REMINDER - Pat Hayes, Oct. 28
To: planlunch.dis: ;

		          POSSIBLE HISTORIES

			      Pat Hayes
	         Schlumberger Palo Alto Research, AI Lab
				
	 	    11:00 AM, MONDAY, October 28
       SRI International, Building E, Room EJ228 (new conference room)


A history is a connected piece of space/time with 'natural' boundaries.  
Using these as a basic ontology for talking about events, processes, etc. 
has some advantages over some other frameworks, and doesn't have some of
the disadvantages which are sometimes attributed to it.  
However, it does have one major problem, which is the difficulty of talking
about alternative possible futures, to allow planning to be done.  
In this talk I discuss a new way of using histories which looks like 
it can overcome this problem.


-------

∂28-Oct-85  0143	@SU-SCORE.ARPA:manfred%ucsc.csnet@CSNET-RELAY.ARPA 	techreport request!    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 28 Oct 85  01:43:12 PST
Received: from CSNET-RELAY.ARPA by SU-SCORE.ARPA with TCP; Mon 28 Oct 85 01:41:34-PST
Received: from ucsc by csnet-relay.csnet id a003596; 28 Oct 85 4:41 EST
Received: by vger.UCSC (4.12/4.7)
	id AA01757; Sun, 27 Oct 85 17:55:29 pst
Date: Sun, 27 Oct 85 17:55:29 pst
From: manfred <@CSNET-RELAY.ARPA,@ucsc.CSNET (Manfred Warmuth):manfred@ucsc.CSNET>
To: CSD@su-score.arpa
Subject: techreport request!
Cc: dhaussle <@csnet-relay.arpa:dhaussle@udenver.CSNET>

I am trying to find a Ph.D. thesis done in CS or Math.
It probably appeared as a tech report.

Title: "Combinatorial entropy and uniform limit laws" 
	by J.M. Steele, 1975.
	
I would greatly appreciate it if you could  mail a copy to each of the
following addresses:

     Prof. Manfred Warmuth
     Dep. of Comp. Sc.
     265 Applied Sciences
     Santa Cruz, Ca 95064


     Prof. David Haussler
     Dep. of Math. and Comp. Sc.
     Univ. of Denver
     Denver, Co 80208


			      Thank you very much!
			      Manfred Warmuth

∂28-Oct-85  0834	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch Series
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 28 Oct 85  08:33:59 PST
Date: Mon 28 Oct 85 08:24:21-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch Series
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12154747899.15.RICHARDSON@SU-SCORE.ARPA>

The topic and speaker for the CSD Lunch on Oct. 29 at 12:15 in MJH 146 are
as follows: John Perry on "What's New at CSLI".

-------

∂28-Oct-85  0836	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar in Logic and the Foundations of Mathematics    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 28 Oct 85  08:36:46 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 28 Oct 85 08:32:57-PST
Date: 28 Oct 85  0825 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Seminar in Logic and the Foundations of Mathematics   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    



Speaker: John Etchemendy, Philosophy Dept. Stanford

Title: "Truth, the liar, and circular propositions"

Time:  Friday, Nov.1, 12:00 noon.

Place:  383N (Math. Dept. Faculty Lounge)

                                S. Feferman


∂28-Oct-85  1533	IIDA@SU-CSLI.ARPA 	Talk this Thursday by Masayo Iida   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 28 Oct 85  15:33:17 PST
Date: Mon 28 Oct 85 15:27:43-PST
From: Masayo Iida <IIDA@SU-CSLI.ARPA>
Subject: Talk this Thursday by Masayo Iida
To: folks@SU-CSLI.ARPA

This is just a reminder that Masayo Iida will be presenting her paper `Case
Assignment by Nominals in Japanese' on Thursday Oct.31st at 10am in the
Ventura Conference room.  This is part of the series of presentations by
the Morphology/Syntax/Discourse Interactions group.  Copies of the paper
will be available on Tuesday 29th, assuming the Kodak machine to be
functioning. 

--Masayo Iida
-------

∂28-Oct-85  1624	@SU-SCORE.ARPA:pratt@su-navajo.arpa 	Student names
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 28 Oct 85  16:24:19 PST
Received: from su-navajo.arpa by SU-SCORE.ARPA with TCP; Mon 28 Oct 85 16:25:40-PST
Received: by su-navajo.arpa with Sendmail; Mon, 28 Oct 85 16:25:30 pst
Date: Mon, 28 Oct 85 16:25:30 pst
From: Vaughan Pratt <pratt@su-navajo.arpa>
Subject: Student names
To: faculty@score

Non-forum members who wish either to advertise specific jobs or to do
on-campus recruiting do have the option of contacting the Career
Planning and Placement Center.  In view of this, I am inclined to
regard provision of names as an additional service beyond this standard
one that we make available as a benefit of Forum membership.
-v

∂28-Oct-85  2233	@SU-SCORE.ARPA:HADDAD@SU-SUSHI.ARPA 	Re: Student names 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 28 Oct 85  22:33:33 PST
Received: from SU-SUSHI.ARPA by SU-SCORE.ARPA with TCP; Mon 28 Oct 85 22:15:51-PST
Date: Mon 28 Oct 85 22:15:24-PST
From: Student Bureaucrats <HADDAD@SU-SUSHI.ARPA>
Subject: Re: Student names
To: NILSSON@SU-SCORE.ARPA
cc: bureaucrat@SU-SUSHI.ARPA, faculty@SU-SCORE.ARPA
Reply-To: bureaucrat@sushi
In-Reply-To: <12154550492.21.NILSSON@SU-SCORE.ARPA>
Message-ID: <12154899187.38.HADDAD@SU-SUSHI.ARPA>

	This list of names of graduating students essentially
constitutes a "mailing list".  Thus, we can immediately deduce a few
things about it:

	1) Some students would delight in having their name passed
around to potentially interested people ("tearing up junk mail only
takes a few seconds, and I will get a whole bunch of interesting
offers that I never would have otherwise heard of").

	2) Other students would feel violated and morally offended at
receiving the junk mail generated by the mailing list.

	3) The list is worth money.


I have no strong feelings about the final decision, but would like to
make a few other comments:

	1) While, personally, I think that the people who would feel
that they were being "violated" are quite silly, I would be derelict
in my duties as a student representative if I didn't recommend that:
if some such list is made available to all comers, then students
should have the option of having their names left off the "mailing
list".

	2) While the list is worth money, it is not clear that we
should necessarily start a meter running for every possible action
that we can get a dime for.

	3) Also, if it is decided that knowledge is power, and power
is money ("what's time, again?" -- sorry, you had to have seen the
movie) and that the list should only be formally given out to Forum
members, I would very strongly object if things went to the extreme
where the knowledge of who was expected to graduate became treated as
"proprietary" or "classified" knowledge, which no one was allowed to
communicate to anyone but Forum members.  The Forum should be based on
"mutually benefitial" arrangements between the CSD and the companies.
It is very unbenefitial to the students if traditional "networking",
as a way of becoming aware of currently available jobs, were to dry
up.

Ramsey.
-------

∂29-Oct-85  0614	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  06:13:55 PST
Date: Tue 29 Oct 85 06:15:13-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: AFLB abstracts
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12154986536.11.PATASHNIK@SU-SUSHI.ARPA>

Besides AFLB, there's another talk this Thursday that might interest us:

	Susan Landau, Wesleyan University and MSRI
	Primes, Codes, and the National Security Agency---
		How Number Theory Affects National Security
	4:15PM Thursday, 31 October, in Terman, M33

Her lecture will be an expository talk on public key cryptosystems, RSA,
and the political controversy surrounding them.  It will be non-technical
and aimed at an interdisciplinary audience in the sciences.

	********* These are the next two regular AFLB's: *********

31-Oct-85  :  Steven Fortune (ATT Bell Labs)

	    Sweepline Algorithms for Voronoi Diagrams

We present a simple, sweepline algorithm for computing Voronoi
diagrams.  The algorithm is applicable to Voronoi diagrams arising
from a set of points, a set of line segments, and a set of weighted
points.  In all cases it achieves an O(n log n) worst-case time bound,
and a linear space bound.  The algorithm is easy to implement, since
it avoids the merge step required by divide-and-conquer algorithms.
An application to a version of the piano mover's problem will also be
discussed.

***** Time and place: October 31, 12:30 pm in MJ352 (Bldg. 460) ******

7-Nov-85  :  Alejandro A. Sch\"affer (Stanford)

		Shortest Prefix Strings Containing All
		k-Element Permutations as Subsequences

What is the length of the shortest string consisting of elements of
{1,...,n} that contains as subsequences all permutations of any
$k$-element subset? Many authors have considered the special case
where k=n. In fact, this special case was posed by D.E. Knuth
(attributed to R. M. Karp) as problem 36 in the technical report
STAN-CS-72-292. This problem arose in analyzing shortest path
algorithms and flowgraph reduction algorithms.

I shall instead consider an incremental variation on the problem first
proposed by P.J. Koutas and T.C. Hu (Discrete Math. 11(1975) pp.
125-132).  For a fixed value of n they ask for a string on the
alphabet {1,...,n} such that for all values of k <= n, the prefix
containing all permutations of any k-element subset as subsequences is
as short as possible.  The problem can also be viewed as follows.  For
k = 1 one needs n distinct digits to find each of the n possible
permutations. In going from k to k + 1, one starts with a string
containing all k-element permutations as subsequences, and one adds as
few digits as possible to the end of the string so that the new string
contains all k+1-element permutations.  The catch is that for all
values of k most shortest prefixes for k cannot be extended minimally
to prefixes for k+1.

Until now only very weak results were known about this version with
the prefix restriction posed by Koutas and Hu.  I shall describe an
*exact* solution using only *elementary* techniques.

I intend to make occasional use of the side blackboard in MJH352.

***** Time and place: November 7, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

More information about future and past AFLB meetings and topics are in
the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂29-Oct-85  0907	MODICA@SU-SCORE.ARPA 	Class Lists  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  09:07:21 PST
Date: Tue 29 Oct 85 09:04:38-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: Class Lists
To: Instructors@SU-SCORE.ARPA
Message-ID: <12155017377.34.MODICA@SU-SCORE.ARPA>

I'll be happy to send class lists through ID mail to any of you who wish
to express an interest in receiving them. Just let me know who and where
you are.
                                Gina
-------

∂29-Oct-85  1011	TREITEL@SUMEX-AIM.ARPA 	Conjunct ordering    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  10:11:26 PST
Received: from SUMEX-AIM.ARPA by su-aimvax.arpa with Sendmail; Tue, 29 Oct 85 10:08:49 pst
Date: Tue 29 Oct 85 10:08:10-PST
From: Richard Treitel <TREITEL@SUMEX-AIM.ARPA>
Subject: Conjunct ordering
To: nail@SU-AIMVAX.ARPA
Cc: treitel@SUMEX-AIM.ARPA
Message-Id: <12155028943.38.TREITEL@SUMEX-AIM.ARPA>

Following on the last meeting at which we discussed conjunct ordering, I
decided to at least try to back up Dave Smith's contention that the problem as
he thinks of it is NP-complete  (I ended up agreeing with JDU that his book
does not contain such a result).   So here's a proof.   I expect to come to
tomorrow's meeting (10-30) to expound it.   I have not made any attempt to win
$1 for the interleaving conjecture.



Optimal Conjunct Ordering is NP-complete		Oct 85
							Richard Treitel


A version of the problem of optimally ordering the conjuncts in a conjunctive
query is proved NP-complete.


The input to the problem is:

1. a set of terms, each being a predicate symbol followed by a list of symbols
which are either constants or variables.  No function symbols.   Usual
constraints as to types and arities apply between terms having any symbol in
common.

2. for each predicate symbol P, a number N(P), the number of facts stored in
the database which begin with that symbol.

3. for each distinct data type such that the query contains any variable of
that type, a number, the number of values in the domain of the type.   For
infinite domains this is redundant because we will never try to solve a term
with such a variable free in it.   For a variable V we write S(V) to denote the
number associated with the type of the variable.


The output is an ordered list of the same terms, which minimises the cost
function.

The cost of a list T1, ... Tn of terms is
the sum for i= 1 to n of  C(T1, ... Ti), which in turn is defined by

C(T1) = N(P1)	where Pj is the predicate symbol of Tj, and
C(T1, ... Ti+1) = C(T1, ... Ti) * N(Pi+1) / B(Ti+1)

where B(Tj) is the product of S(V) over all V occurring both in Tj and in some
Tk with k < j.   These are the variables that will already be bound by the time
Tj is reached, and so their bindings will reduce the number of answers to Tj.


Now refer to page 201 of "Computers and Intractability" by Garey&Johnson, for
problem GT44, "Minimum Cut Linear Arrangement".   This problem, which is
NP-complete, is stated as follows:  given a graph G with n vertices, and an
integer K, find whether there is an ordering of the vertices (with the ordinal
number of a vertex v denoted by f(v)) such that, for every i with 1 < i < n,
the number of edges (u,v) of the graph such that f(u) <= i < f(v) is at most K.

The "1 < i < n" appears to be a typo for "1 <= i < n";  there is a puzzling
asymmetry in the problem as stated.



Suppose we are given an instance of this problem.   Reduce it to a conjunct
ordering problem as follows:

Without loss of generality, there is no edge connecting a node to itself.   For
each edge in G, create a variable, and for each vertex, a predicate symbol.
The term corresponding to each vertex consists just of its predicate symbol and
the variables of the edges that are incident to it.  Consider the query which
is the conjunction of these terms.   If a vertex has degree d, let the number
associated with its predicate symbol be n↑d, where n is the number of vertices
in the graph.  For all variables V let S(V) be n↑2.   I shall prove that there
exists an ordering of the query with cost less than n↑(K+1) iff there exists a
linear arrangement of the kind demanded.


Lemma:  for a given ordering of the vertices in the graph (terms in the query),
the number of edges (u,v) in the graph such that f(u) <= i < f(v) (call this
number Cut(i)) is precisely the number of variables which appear just once in
the first i terms of the ordering.

Proof:  immediate from the observation that every variable appears exactly
twice in the query.


Now C(T1, ... Ti) is just the product of N(Pj) for j from 1 to i divided by the
product of S(V) over all variables V appearing twice in (T1 ... Ti).  Let W(i)
be the number of such variables.   By the definition of N(Pi) given above, we
see that the product of N(Pj) for j from 1 to i is just n to the power (2*W(i)
+ Cut(i)).   Therefore C(T1, ... Ti) is n↑Cut(i) for all i from 1 to n-1.
C(T1, ... Tn) is 1.

Thus if the vertices of the graph are arranged so that Cut(i) <= K for all
relevant i, the cost of the corresponding ordering of the terms of the query
will be at most

1 + (n-1) * n↑K

which is less than n↑(K+1), whereas if there is some i such that Cut(i) > K,
then the cost will be at least n↑(K+1).
-------

∂29-Oct-85  1129	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  11:29:23 PST
Date: Tue 29 Oct 85 11:31:12-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12155044059.25.RICHARDSON@SU-SCORE.ARPA>



Lunch today in MJH 146 at 12:15.

-------

∂29-Oct-85  1219	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  12:19:04 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 29 Oct 85 12:12:03 pst
Date: Tue, 29 Oct 85 12:12:03 pst
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

We'll continue our discussion of conjunctive queries.
Richard Treitel will present a proof of NP-completeness of
a formulation of this problem.

Let's meet at 11AM in 301MJH.

∂29-Oct-85  1537	CAROL@SU-CSLI.ARPA 	Demo of Dlion gloss package   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  15:37:00 PST
Date: Tue 29 Oct 85 15:31:57-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Demo of Dlion gloss package
To: Folks@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA

**************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**N
			    23-OCT-85

Hi-

I am planning a series of informal demoes of dandelion 
packages.  Short and specific, preferably held in the late 
afternoon.  We can also use these times to discuss other
concerns.

The first demo is of particular interest to linguists.  
Michael Wescoat has written a gloss program to enable you 
to format examples of sentences in one language which require 
word by word glosses, plus more fluent renderings, in 
another. This means true beautification of papers involving 
translations.  Michael will demo this program on 12 Nov. 
at 4 p.m. in Trailer E.

Further subjects for this series could include:  Notecards, 
Loops, Sketch (advanced features and troubleshooting), Topics 
and troubles in TEdit, Augmentations to TEdit (such as 
TEditkeyplus, interface with emacs, indexing) ...

I would appreciate hearing from you all about which
of these demoes you would be likely to attend, as well
as requests and suggestions. 

I've addressed this to "Folks" - If you wish to receive mail
about dandelions, please add yourself to the dlion-users
mailing list. (Send a note to "Requests".)

 
			    -Carol     (321-2136,  Ventura 28)

EWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION
**************************************************************



-------

∂29-Oct-85  1623	EMMA@SU-CSLI.ARPA 	Halloween  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  16:23:32 PST
Date: Tue 29 Oct 85 16:17:04-PST
From: Guy Fawkes
Subject: Halloween
Sender: EMMA@SU-CSLI.ARPA
To: folks@SU-CSLI.ARPA, bboard@SU-CSLI.ARPA, consultants@SU-CSLI.ARPA
Reply-To: bboard
Tel:  497-3479

                                                                              
                                                            ←←                
             $                                             /..\               
            $      HALLOWEEN IS THIS THURSDAY             |    \          
            $                                             |     \             
         ########                                         /      \            
      ###        ###        DRESS UP AND JOIN THE SPIRIT OF         
     ##   ↑   ↑   ##               THE DAY.                                   
     ##           ##      ←←                                                  
      ###  `'`' ###      |..\     TOM WILL ALSO PROVIDE A SPECIAL TEA         
        #########       /    \                                                
                       /      \                                               
                      /       |     Lasciate ogni speranza, voi ch'entrate    
                                                                              
                                                                              
                                                                              
-------

∂29-Oct-85  2334	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	FST&TCS 5th Conference    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 29 Oct 85  23:34:02 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Tue 29 Oct 85 23:33:55-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Tue 29 Oct 85 23:29:08-PST
Received: by rsch.wisc.edu; Tue, 29 Oct 85 23:30:33 CST
Received: from crys.wisc.edu by rsch.wisc.edu; Tue, 29 Oct 85 15:45:12 CST
Message-Id: <8510292144.AA28617@crys.wisc.edu>
Received: from CS.COLUMBIA.EDU by crys.wisc.edu; Tue, 29 Oct 85 15:44:50 CST
Date: Tue 29 Oct 85 16:45:16-EST
From: Susan A Maser <MASER@CS.COLUMBIA.EDU>
Subject: FST&TCS 5th Conference
To: theory@CRYS.WISC.EDU
Status: R
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 29 Oct 85 23:30:17 CST (Tue)
Resent-From: Udi Manber <udi@rsch.wisc.edu>


MESSAGE FROM S.N. MAHESHWARI, IIT DELHI:

                           Fifth Conference

                  FOUNDATIONS OF SOFTWARE TECHNOLOGY
                                 AND
                     THEORETICAL COMPUTER SCIENCE

                         16-18 December 1985
             INDIA INTERNATIONAL CENTRE, NEW DELHI, INDIA

PROGRAMME

Monday 16 December

0800-0900   Registration

0900-0915   Inauguration

0915-1015   Keynote Address

            The Mathematics of Programming C.A.R. Hoare (Oxford University)

1015-1045   SESSION 1:  Distributed and Concurrent Programming

            Concurrent Programming Using Actors: Exploiting Large-Scale
            Parallelism
            G. Agha, C. Hewitt (Massachusetts Institute of Technology)

            A New Class of High Level Programs for Distributed Computing
            Systems
            S. Ramesh, S.L. Mehndiratta (Indian Institute of Technology,
            Bombay)

	    A Class of Termination Detection Algorithms for Distributed
	    Computations
	    D. Kumar (University of Texas)

	    New Protocols for the Election of a Leader in a Ring
	    A. Marchetti-Spaccamela (Universita' di Roma)

1245-1345   LUNCH

1345-1500   SESSION 2:  Programming

	    Program Simplification via Symbolic Interpretation
	    C. Ghezzi, D. Mandrioli, A. Tecchio (Politecnico di Milano)

	    PROLOG-Based Inductive Theorem Proving
	    J. Hsiang, M. Srivas (State University of New York at 
            Stony Brook)

	    On the Calling Behaviour of Procedures*
	    D. Armbruster (Universitat Stuttgart)

1500-1530   COFFEE

1530-1700   SESSION 3:  Algorithms

	    Approximation Algorithms for Planar Matching
	    S.M. Venkatesan (University of Minnesota)

	    Geometric Optimization and the Polynomial Hierarchy
	    C. Bajaj (Purdue University)

	    Deriving Object Octree from Images
	    J. Veenstra, N. Ahuja (University of Illinois)



Tuesday 17 December

0900-1000   Invited Talk

	    Deduction with Relation Matching
	    Z. Manna (Stanford University)
	    R. Waldinger (SRI International)

1000-1100   SESSION 4:  Specifications

	    Recursively Defined Domains and Their Inductive Principles
	    F.A. Jenson (Aalborg University Center)
	    K.G. Larson (University of Edinburgh)

	    Large Database Specifications from Small Views
	    S. Khosla, T.S.E. Maibaum, M. Sadler (Imperial College)

1100-1130   COFFEE

1130-1315   SESSION 5:  Theory

	    A Decision Method for Temporal Logic Based on Resolution
	    G. Venkatesh (Tata Institute of Fundamental Research)

	    A Generalization of the Parikh Vector for Finite and
	    Infinite Words
	    R. Siromoney, V. Rajkumar Dare (Madras Christian College)

	    The Implication Problem for Functional and Multivalued
	    Dependencies: An Algebraic Approach
	    V.S. Lakshmanan, C.E. Veni Madhavan (Indian Institute of 
	    Science)

	    A Simple Characterization of Database Serializability*
	    K. Vidyasankar (Memorial University of Newfoundland)

1315-1415   LUNCH

1415-1515   Invited Talk

	    Who Needs to Verify Programs if You Can Test Them
	    A. Chandra (IBM Thomas J. Watson Research Centre)

1515-1545   COFFEE

1545-1715   SESSION 6:  Semantics and Proof Techniques

	    Partial Correctness Semantics for CP[V,l,&]
	    V.A. Saraswat (Carnegie-Mellon University)
 
	    A Proof Technique for Rely/Guarantee Properties
	    E.W. Stark (State University of New York at Stony Brook)

	    A Complete Proof System for SCCS with Modal Assertions
	    G. Winskel (University of Cambridge)



Wednesday 18 December

0900-1000   Invited Talk

	    Demand-Driven Evaluation on Dataflow Machine
	    Arvind (Massachusetts Institute of Technology)

1000-1100   SESSION 7:  VLSI

	    Design and Implementation of a Procedural VLSI
	    Layout System
	    J.M. Mata (Universidade Federal de Minas Gerais)
	    G. Vijayan (Georgia Institute of Technology)

	    VLSI Systems for Matrix Multiplication
	    K.H. Cheng, S. Sahni (University of Minnesota)

1100-1130   COFFEE

1130-1330   SESSION 8:  Parallel Algorithms

	    Parallel Algorithms for Solving Certain Classes of 
	    Linear Recurrences
	    S. Lakshmivarahan, S.K. Dhall (University of Oklahoma)

            O(1) Parallel Time Incremental Graph Algorithms
	    D.D. Sherlekar, S. Pawagi, I.V. Ramakrishnan (University
	    of Maryland)

	    NC Algorithms for Comparability Graphs, Interval Graphs,
	    and Testing for Unique Perfect Matching
	    D. Kozen (Cornell University)
	    U.V. Vazirani (University of California, Berkeley)
	    V.V. Vazirani (Cornell University)

	    Fast and Efficient Parallel algorithms for the Exact
	    Inversion of Integer Matrices
	    V. Pan (State University of New York at Albany)

1330-1430   LUNCH


*Short Presentation

REGISTRATION INFORMATION

Charge   Rs. 350 per person
	 includes a copy of the conference record,
	 refreshments, lunch and the Conference Dinner

Payment  By cheque, bank draft or money order
	 in favour of Fifth Conference
	 Foundation of Software Technology and
	 Theoretical Computer Science
	 Or in cash

Advance Registration is recommended.  Foreign participants should
inform of their intention to register at the address given below, but
due to bank charges and delays are advised to pay their charges at the
time of the Conference.  Accomodation can be booked at hotels near the
conference venue.  For more information write to the address given
below.

Address for all correspondence:

Dr. Niraj Sharma
FST&TCS 5
Department of Computer Science
Indian Institute of Technology
New Delhi, 110016	
INDIA
Telegram: TECHNOLOGY NEW DELHI 110016 INDIA

-------------
TN Message #4
-------------

∂30-Oct-85  0615	PATASHNIK@SU-SUSHI.ARPA 	Special AFLB   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  06:15:29 PST
Date: Wed 30 Oct 85 06:17:04-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Special AFLB
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12155249017.7.PATASHNIK@SU-SUSHI.ARPA>

This Friday there will be a special AFLB.  Please note time (2:15) and
room (301) changes.

		***************************************

1-Nov-85 (Friday)  :  Mark Krentel (Cornell)

		The Complexity of Optimization Problems

Many important problems in computer science, such as CLIQUE, COLORING
and TRAVELLING SALESPERSON, arise naturally as optimization problems.
Typically one considers these problems as decision procedures, which
are often in NP, and one shows intractibility by showing them
NP-complete. We generalize the notion of NP-problems, in a manner
analogous to the definition of Valiant's class #P, by considering the
optimization version of the problem itself, and we show that this idea
yields a natural class of problems that we call OptP. This class
allows us to make finer distinctions on the complexity of optimization
problems than is possible in NP. For example, assuming P different
from NP, we can show that TRAVELLING SALESPERSON is strictly harder
than CLIQUE and that CLIQUE is strictly harder than BIN PACKING. We
then relate OptP to the class of functions computable in polynomial
time with an oracle for NP, by showing that every P↑{SAT} (P with an
oracle for SAT) function decomposes into an OptP function followed by
a polynomial time computation. This allows us to clear up a
misconception on the role of uniqueness for the problem of UNIQUELY
OPTIMAL TRAVELLING SALESPERSON as considered by Papadimitriou in the
1982 FOCS conference.

***** Time and place: November 1, 2:15 pm in MJ301 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

More information about future and past AFLB meetings and topics are in
the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂30-Oct-85  0931	PHILOSOPHY@SU-CSLI.ARPA 	Josh Cohen
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  09:31:49 PST
Date: Wed 30 Oct 85 09:27:51-PST
From: Eve Wasmer <PHILOSOPHY@SU-CSLI.ARPA>
Subject: Josh Cohen
To: folks@SU-CSLI.ARPA
cc: friends@SU-CSLI.ARPA

The Philosophy Department is sponsoring a talk by Joshua Cohen from
M.I.T.  The talk is titled "Structure, Choice, and Legitimacy in Locke's
Theory of Politics", and will be at 4:15 on Tuesday, November 5 in the
Philosophy Department Seminar Room, 90=92Q.
-------

∂30-Oct-85  0931	PHILOSOPHY@SU-CSLI.ARPA 	Josh Cohen
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  09:31:49 PST
Date: Wed 30 Oct 85 09:27:51-PST
From: Eve Wasmer <PHILOSOPHY@SU-CSLI.ARPA>
Subject: Josh Cohen
To: folks@SU-CSLI.ARPA
cc: friends@SU-CSLI.ARPA

The Philosophy Department is sponsoring a talk by Joshua Cohen from
M.I.T.  The talk is titled "Structure, Choice, and Legitimacy in Locke's
Theory of Politics", and will be at 4:15 on Tuesday, November 5 in the
Philosophy Department Seminar Room, 90=92Q.
-------

∂30-Oct-85  1055	PHILOSOPHY@SU-CSLI.ARPA  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  10:55:28 PST
Date: Wed 30 Oct 85 10:48:39-PST
From: Eve Wasmer <PHILOSOPHY@SU-CSLI.ARPA>
To: folks@SU-CSLI.ARPA
cc: bboard@SU-CSLI.ARPA

Does anyone have a good condition Apple IIe for sale?  Call Nancy Steege
on 497-2547 or electronic mail HF.NSS.
-------

∂30-Oct-85  1140	INGRID@SU-CSLI.ARPA 	Announcement  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  11:40:15 PST
Date: Wed 30 Oct 85 11:35:34-PST
From: Ingrid Deiwiks <INGRID@SU-CSLI.ARPA>
Subject: Announcement
To: SU-BBoards@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA


		          ANNOUNCEMENT


Soto Symposia on Artificial Intelligence
----------------------------------------

A series of three symposia on AI will be held in the Soto Lounge,
Wilbur Hall, the first three Mondays in November, at 7 p.m.  These
symposia are intended to inform Stanford students and others who are
interested about the nature of AI, its prospects, and the intellectual
and social issues to which it gives rise.

Monday, November 4, 7 p.m.

	"What is AI?"

	Barbara Grosz     Computer Scientist, SRI
		          Consulting Professor of Computer Science,
		          Stanford

	Stan Rosenschein  Director of AI Lab, SRI
		          Consulting Professor of Computer Science,
		          Stanford

	Matt Ginsberg     Department of Computer Science, Stanford


Monday, November 11, 7 p.m.

	"Can we make computers that think and talk?"

	Terry Winograd    Professor of Computer Science, Stanford

	Brian Smith       Computer Scientist, Xerox PARC
		          Consulting Professor of Philosophy,
	                  Stanford			  

	Nils Nilsson      Professor of Computer Science and
		          Chair of Department of Computer Science, 
			  Stanford


Monday, November 18, 7 p.m.

	"Star Wars and Computation"

	John McCarthy     Professor of Computer Science, Stanford

	David Redell      DEC Research Laboratory

	and possibly others

Each symposium will begin with short talks by the speakers, followed
by plenty of time for vigorous interchanges among the speakers and
with members of the audience.

Soto is the first dormitory in the Wilbur complex that one comes to
when walking out Escondido from the Undergraduate Library.

	
----------------------------------------------------------------------
<--UGLI		              Escondido  Rd.
----------------------------------------------------------------------
                                       
xxxxxxxxxxxxxx                              X XXX     XXX X 
		                   Soto --> X   Wilbur    X		
Stern Hall                                  X             X

xxxxxxxxxxxxxx                              X    Hall     X
                                            X             X
                                            X XXX     XXX X




-------

∂30-Oct-85  1155	@SU-SUSHI.ARPA:vemuri@ngp.UTEXAS.EDU 	Re:  Special AFLB
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  11:55:47 PST
Received: from ngp.UTEXAS.EDU by SU-SUSHI.ARPA with TCP; Wed 30 Oct 85 11:55:52-PST
Date: Wed, 30 Oct 85 13:28:25 cst
From: vemuri@ngp.UTEXAS.EDU (baba c. vemuri)
Posted-Date: Wed, 30 Oct 85 13:28:25 cst
Message-Id: <8510301928.AA24884@ngp.UTEXAS.EDU>
Received: by ngp.UTEXAS.EDU (4.22/4.22)
	id AA24884; Wed, 30 Oct 85 13:28:25 cst
To: PATASHNIK@SU-SUSHI.ARPA, aflb.all@SU-SUSHI.ARPA
Subject: Re:  Special AFLB


∂30-Oct-85  1639	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH  
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  16:39:32 PST
Date: Wed 30 Oct 85 16:38:45-PST
From: LANSKY@SRI-AI.ARPA
Subject: Next Monday's PLANLUNCH
To: planlunch.dis: ;

	   THOUGHTS ON AN INVITATION TO A (SURPRISE) BEHEADING

			    David Israel
  	             SRI International, AI Center	
				
	 	    11:00 AM, MONDAY, November 4
       SRI International, Building E, Room EJ228 (new conference room)

Imagine you learn - from an authoritative source - that you will be
beheaded some afternoon of this coming week; say by Saturday at the
latest.  No doubt you'd be unnerved.  What, though, if you were also
told that the beheading would only take place if the following
condition were met: you would not know, say on the night before or the
morning of the event, that it was going to take place that afternoon.
So the beheading, if it is to take place at all, is to be a surprise.
Should your fears be allayed?  Indeed, shouldn't they be completely
eliminated?  There is a very plausible-seeming argument that you have
nothing to worry about. Unhappily, there seem to be lots of things
wrong this argument.  What's up?

I want to look at this puzzle, with an eye to connecting together a
variety of considerations about belief, knowledge, time, and - of all
things - rational action.  (This is going to be a PLANlunch seminar,
after all.)  So, I invite you all to a seminar that I shall give some
day between now and next Wednesday.  But, of course, you won't know,
say on the morning before the seminar,.........


-------

∂30-Oct-85  1732	EMMA@SU-CSLI.ARPA 	Newsletter October 31, No. 52  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 30 Oct 85  17:32:20 PST
Date: Wed 30 Oct 85 16:47:32-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter October 31, No. 52
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
October 31, 1985                Stanford                       Vol. 2, No. 52
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, October 31, 1985

   12 noon		TINLunch
     Ventura Hall       The Formation of Adjectival Passives
     Conference Room    by B. Levin and M. Rappaport
			Discussion led by Mark Gawron

   2:15 p.m.		CSLI Seminar
     Redwood Hall	Foundations of Document Preparation
     Room G-19		David Levy, CSLI and Xerox PARC

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Redwood Hall	The Structure of Social Facts
     Room G-19		Prof. John Searle, Dept. of Philosophy, UC Berkeley
                              ←←←←←←←←←←←←
          CSLI ACTIVITIES FOR *NEXT* THURSDAY, November 7, 1985

   12 noon		TINLunch
     Ventura Hall       James Gibson's Ecological Revolution in Psychology
     Conference Room    by E. S. Reed and R. K. Jones
			Discussion led by Ivan Blair, CSLI
			(Abstract on page 2)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	Phonology/Phonetics Seminar
     Room G-19		Bill Poser and Paul Kiparsky
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Redwood Hall	Meaning, Information and Possibility
     Room G-19		Lofti A. Zadeh, Computer Science Division
			University of California at Berkeley
			(Abstract on page 2)
                              ←←←←←←←←←←←←

!
Page 2                     CSLI Newsletter                   October 31, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                    ABSTRACT OF NEXT WEEK'S TINLUNCH
           James Gibson's Ecological Revolution in Psychology
                       E. S. Reed and R. K. Jones

      From about 1950 until his death, James Gibson constantly argued for
   a view of and a research program for cognitive psychology that
   differed radically from the mainstream position.  Today the dominant
   view in cognitive psychology is of cognitive agents as information
   processors, a view to which the advent of the modern digital computer
   has given a considerable boost.  In the paper for this week's
   Tinlunch, Reed and Jones characterize and contrast the Gibsonian (or
   ecological) and information processing approaches.
      My intention is to use this article to lay out for discussion the
   basic principles of the ecological approach.  The issues to be
   considered include: the need for cognitive psychology to study the
   organism in a real environment; the ecological program of studying
   the environmental sources of information; and the rejection of any
   appeal to mental representations in psychological explanation.
							--Ivan Blair
                              ←←←←←←←←←←←←
                        NEXT WEEK'S CSLI SEMINAR
                 Abstract of Phonology/Phonetics seminar

      Post-lexical phonological rules are associated with a hierarchy of
   nested domains, which are systematically related to phrase structure.
   There is growing evidence in favor of recent proposals that this
   hierarchy is universal. In this talk, we show that Japanese has tonal
   rules associated with each of the postulated post-lexical domains, and
   propose a cross-linguistic account for one of the prosodic domains,
   the phonological phrase.		--Bill Poser, Paul Kiparsky
                              ←←←←←←←←←←←←
                       NEXT WEEK'S CSLI COLLOQUIUM
                  Meaning, Information and Possibility
          L.A. Zadeh, Computer Science Division, University of
                    California, Berkeley, CA 94720}

   Our approach to the connection between meaning and information is in
   the spirit of the Carnap-Bar-Hillel theory of state descriptions.
   However, our point of departure is the assumption that any proposition, 
   p, may be expressed as a generalized assignment statement of the form
   X `isr' C, where X is a variable which is usually implicit in p, C is
   an elastic constraint on the values which X can take in a universe of
   discourse U, and the suffix r in the copula `isr' is a variable whose
   values define the role of C in relation to X.  The principal roles are
   those in which r is d, in which case C is a disjunctive constraint; and 
   r is c, p and g, in which cases C is conjunctive, probabilistic, and
   granular, respectively.  In the case of a disjunctive constraint, `isd' 
   is written for short as `is', and C plays the role of a graded possibility 
   distribution which associates with each point (or, equivalently,
   state-description) the degree to which it can be assigned as a value to X.
   This possibility distribution, then, is interpreted as the information
   conveyed by p.  Based on this interpretation, we can construct a set of 
   rules of inference which allow the possibility distribution of a
   conclusion to be deduced from the possibility distributions of the
   premises.  In general, the process of inference reduces to the solution 
   of a nonlinear program and the traditional methods of deduction in
   first-order logic are explained and illustrated by examples.
!
Page 3                     CSLI Newsletter                  October 31, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                       ENVIRONMENTS GROUP MEETING
      NoteCards:  An Environment for Authoring and Idea Structuring
                         Randy Trigg, Xerox PARC
             Monday, November 4, noon, Ventura Seminar Room

      NoteCards is part of an ongoing research project in the Intelligent
   Systems Lab at Xerox PARC investigating "idea processing" tasks, such
   as interpreting textual information, structuring ideas, formulating
   arguments, and authoring complex documents.  NoteCards is intended
   primarily as an idea structuring tool, but it can also be used as a
   fairly general database system for loosely structured information.
      The basic object in NoteCards is an electronic note card containing
   an idea-sized unit of text, graphics, images, or whatever.  Different
   kinds of note cards are defined in an inheritance hierarchy of note
   card types (e.g., text cards, sketch cards, query cards, etc.).  On
   the screen, multiple cards can be simultaneously displayed, each one
   in a separate window having an underlying editor appropriate to the
   card type.
      Individual note cards can be connected to other note cards by
   arbitrarily typed links, forming networks of related cards.  At
   present, link types are simply labels attached to each link.  It is up
   to each user to utilize the link types to organize the note card
   network.
      NoteCards also includes a filing mechanism for building
   hierarchical structures using system-defined card and link types.
   There are also browser cards containing node-link diagrams (i.e.,
   maps) of arbitrary pieces of the note card network and Sketch cards
   for organizing information in the form of drawings, text and links
   spatially.
      All of the functionality in NoteCards is accessible through a set
   of well-documented Lisp functions, allowing the user to create new
   types of note cards, develop programs that monitor or process the note
   card network, and/or integrate other programs into the NoteCards
   environment.
                               ----------
                          PIXELS AND PREDICATES
                        The Caricature Generator
                              Susan Brennan
          CSLI trailers, 1:00 p.m., Wednesday, November 6, 1985

      In an investigation of primitives for image generation,
   manipulation and perception, a face is an interesting example of an
   image.  I will briefly survey psychological literature on face
   perception which treats such issues as piecemeal vs. configurational
   recognition strategies.  I'll describe an application where a
   caricature of a face serves as a form of semantic bandwidth
   compression.  Then, with additional inspiration from art, computer
   graphics and machine vision, I'll develop a theory of caricature.
      Conditions permitting, there will be a demonstration of a program
   which generates caricatures of faces from line drawings and provides
   the user with a single exaggeration control with which the distortion
   in the image (relative to a norm) can be turned up or down.  I will
   also show a videotape and refer to the work that Gill Rhodes and I
   have been doing recently on perception of these caricatures.
!
Page 4                     CSLI Newsletter                  October 31, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                  INTERACTIONS OF MORPHOLOGY AND SYNTAX
                 Case-Assignment by Nominals in Japanese
                               Masayo Iida
        Thursday, October 31, 10:00 a.m., Ventura Conference Room

      In this paper I will discuss certain peculiar properties of a class
   of Japanese deverbal nominals, which show verb-like properties in
   certain environments: specifically, they assign verbal case and can be
   modified by adverbs (`verbal case' includes nominative, accusative and
   dative, i.e., cases normally assigned by a verb).  These
   case-assignment phenomena pose a problem for current syntactic
   theories, which assume that verbs alone assign such cases, while nouns
   do not.  Now I have observed that a deverbal nominal assigns verbal
   case only when it is concatenated with a suffix bearing temporal
   information, which might be encoded with the feature [+aspect].  The
   nominal assigns case when the following two conditions are satisfied:
   (i) the nominal has a predicate-argument structure, and (ii) it is
   concatenated with a suffix which bears an aspectual feature.  I will
   propose that (syntactic) category membership is not sufficient for
   determining properties of case-assignment, adverb distribution, etc.,
   and suggest that the factors (i) and (ii) are perhaps more relevant.
						   --Masayo Iida
                               ----------
                              LOGIC SEMINAR
             ``Truth, the Liar, and Circular Propositions''
       John Etchemendy and Jon Barwise, Philosophy Dept. Stanford
      Friday, Nov. 1, noon, 383N (Math. Dept. Faculty Lounge)

       Unlike standard treatments of the Liar, we take seriously the
   intuition that truth is, first and foremost, a property of
   propositions (not of sentences), and the intuition that propositions
   (unlike sentences) can be genuinely circular or nonwellfounded.  To
   model the various semantic mechanisms that give rise to the paradox,
   we work within Peter Aczel's set theory, ZFC/AFA, a theory
   equiconsistent with ZFC but with Foundation replaced by a strong
   anti-foundation axiom.  We give two separate models; one based on an
   Austinian conception of propositions (according to which a proposition
   claims that an actual or ``historical'' situation is of a specified
   type), and one based on a Russellian conception (according to which
   propositions are complexes of objects and relations).  The models show
   that the moral of the Liar depends in a crucial way on which
   conception is adopted.
!
Page 5                     CSLI Newsletter                  October 31, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                  SUMMARY OF ENVIRONMENTS GROUP MEETING
                            October 28, 1985

      Wolfgang Pollak of Kestrel spoke on the ADA programming environment
   he helped develop at Rational Systems.  By combining dedicated special
   hardware (high-level-language oriented) with a monolingual operating
   system / command language / environment (written entirely in ADA and
   supported with specialized microcode and memory management), it was
   possible to design the environment in a unified way using the language
   itself as the structure.  All storage is handled by making it possible
   for arbitrary data objects in the language to be declared
   ``persistent,'' rather than having a separate concept of files.  These
   persistent objects are the locus of object management (access control,
   versions, etc.).  The environment is editor-based, with the commands
   extended by using arbitrary function calls in the language.  It
   incorporates a concept of unitary action, which allows the user to
   make a sequence of changes and then either commit (in which case they
   all take effect at once) or abandon (in which case the state is as if
   none of them ever happened).  Wolf described a number of techniques
   for making the environment incremental---for keeping the feel that
   each small change takes effect as it is made, rather than waiting for
   some large-scale redisplay or compile.  Discussion emphasized the way
   that a number of these issues and techniques could apply to other
   environments.
-------

∂31-Oct-85  0917	TAJNAI@SU-SCORE.ARPA 	1985 IBM Party    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Oct 85  09:16:50 PST
Date: Thu 31 Oct 85 09:13:23-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: 1985 IBM Party
To: CSD@SU-SCORE.ARPA, CSL-Everyone@SU-SIERRA.ARPA
Message-ID: <12155543259.21.TAJNAI@SU-SCORE.ARPA>


IBM is giving a party!  Yes, a party!  All those affiliated with CSD/CSL
are invited.  The Forum is making the arrangements.

Thursday, November 14, 4 p.m. to 6 p.m., Faculty Club Gold Room.

Brent Hailpern, one of our illustrious alumni, conceived the idea two
years ago.  The 1983 and 84 parties were so successful that he suggested
we do it again.  So let's do it!!


Carolyn
-------

∂31-Oct-85  0926	EMMA@SU-CSLI.ARPA 	Halloween 3
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 31 Oct 85  09:26:20 PST
Date: Thu 31 Oct 85 09:23:24-PST
From: los diablos de CSLI
Subject: Halloween 3
Sender: EMMA@SU-CSLI.ARPA
To: folks@SU-CSLI.ARPA, bboard@SU-CSLI.ARPA
Reply-To: bboard
Tel:  497-3479

                                                                              
                                                                              
                                                                              
            &                                                                  
          #######          HALLOWEEN IS HERE.  THE GHOSTS, THE DEVILS,
         ##     ###
        ##  ↑ ↑   ##       THE BASEBALL PLAYERS, RONALD REAGANS, AND
       ###  <>    ##                                                          
        ##       ##        OTHERS BEGIN THERE ONCE YEARLY WALK THROUGH        
          ########                                                            
                           THE REAL WORLD.                                    
                                                           
         THEY WILL ALL BE SEEN AT CSLI'S THURSDAY AFTERNOON TEA MINGLING

         WITH ORDINARY FOLK.


Her strong enchantments failing, 
Her towers of fear in wreck, 
Her limbecks dried of poisons 
And the knife at her neck,

The Queen of air and darkness  
Begins to shrill and cry, 
`O young man, O my slayer,
To-morrow you shall die.'

O Queen of air and darkness, 
I think 'tis truth you say.
And I shall die to-morrow; 
But you will die to-day.

             --A. E. Housman
-------

∂31-Oct-85  1323	EMMA@SU-CSLI.ARPA 	All Hallow's Eve
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 31 Oct 85  13:23:27 PST
Date: Thu 31 Oct 85 13:20:18-PST
From: Lilith
Subject: All Hallow's Eve
Sender: EMMA@SU-CSLI.ARPA
To: folks@SU-CSLI.ARPA, bboard@SU-CSLI.ARPA, consultants@SU-CSLI.ARPA
Reply-To: bboard
Tel:  497-3479


   CSLI will be starting its tea/party at 3:15 in the Trailer
Classroom. 

  Please come and celebrate Halloween, the lengthening nights, etc.

-------

∂31-Oct-85  1450	EMMA@SU-CSLI.ARPA 	Halloween tea   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 31 Oct 85  14:50:31 PST
Date: Thu 31 Oct 85 14:38:57-PST
From: Cerebrus
Subject: Halloween tea
Sender: EMMA@SU-CSLI.ARPA
To: bboard@SU-CSLI.ARPA, folks@SU-CSLI.ARPA, consultants@SU-CSLI.ARPA
Reply-To: bboard
Tel:  497-3479


Because of the atypical October weather and because the trailer room
is already booked, the tea will be held on the deck.

`Two objects cannot occupy the same place at the same time'

-------

∂31-Oct-85  1509	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Oct 85  15:09:42 PST
Date: Thu 31 Oct 85 14:46:22-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12155603875.26.LIBRARY@SU-SCORE.ARPA>

Lucid, The Dataflow Programming Language.  APIC Studies In Data Processing
no. 22 by William Wadge and Edward Ashcroft. QA76.73.L83W33 1985.

The Acquisition Of Syntactic Knowledge. MIT Pres Series in AI. by Robert
Berwick. Q335.B48 1985.

Information Systems: Theoretical And Formal Aspects. IFIP.  ed. by A.
Sernadas, J. Bubenko and A. Olive.  QA75.5.I376 1985.

Metamodeling: A Study of Approximations in Queueing Models. Research 
Reports and Notes. Computer Systems Series.  by S. C. Agrawal
QA76.9.C65A38 1985.

Logic Testing and Design For Testability. MIT Press Computer Systems Series.
by Hideo Fujiwara.  TK7868.L6F85 1985.

Model-Based Image Matching Using Location. MIT Press. ACM Distinguished
Dissertation 1984.  by Henry Baird.  TK7882.P3B35 1985.

A Manual of Intensinal Logic. CSLI Lecture Notes. by Johan van Benthem.
P51.C18 no. 1

Planning and Designing the Data Base Environment by Thomas Turk. 
QA76.9.D3T85 1985.

The Massively Parallel Processor. Research Report and Notes. Scientific
Computation Series. MIT Press. ed. by J. L. Potter. QA76.5.M24375 1985.

Performance and Evaluation of Lisp Systems. Research Reports and Notes.
Computer Systems Series. MIT Press.  by Richard Gabriel. QA76.73.L23G32 1985

SIMULA Information. Proceedings 9th Simula Users' Cong. University of
Geneva, Switzerland 1981.  QA76.73.S55S55 1981.

Intuitionistic Type Theory. Studies in Proof Theory. Lecture Notes.
by Per Martin-lof.  QA9.47.M37.

Computer Crime, Abuse, Liability and Security. A Comprehesive Bibliography
1970-1984. compiled by Reba Best and Cheryn Picquet. Z5703.4C63.B47 1985.

Microprocessors. EPO Applied Technology Series. by O A R Cornillie.
QA76.5C654 1985.

Introduction to Ada. by Paul Chirlian. QA76.73.A35C46 1984.

Getting Computers to Talk Like You and Me. Discourse Context, Focus, and
Semantics. (An ATN Model) by Rachel Reichman. P302.R45 1985.

Dictionary of Electrical, Electronics, and Computer Abreviation. 
compiled by Phil Brown.  REF TK9.B76 1985.  

Anglo-American and German Abbreviations in Data Processing. by 
Peter Wennrich.  REF  QA76.15W46 1984. 

(Chinese Dictionary.  English word alphabetized with Chinese 
explanations.)by Fan, Jen-te REF QA76.15.J46 1981.

H. Llull
-------

∂31-Oct-85  1952	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	VLSI workshop announcement
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 31 Oct 85  19:52:07 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 31 Oct 85 19:52:35-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Thu 31 Oct 85 19:52:36-PST
Received: by rsch.wisc.edu; Thu, 31 Oct 85 21:30:46 CST
Message-Id: <8510311852.AA07342@rsch.wisc.edu>
Received: from Louie.UDEL.EDU by rsch.wisc.edu; Thu, 31 Oct 85 12:52:13 CST
Received: from csnet-pdn-gw by Louie.UDEL.EDU id aa27453; 31 Oct 85 0:20 EST
Received: from utd-cs by csnet-relay.csnet id ad25325; 31 Oct 85 0:14 EST
Date: 30 Oct 85 15:36:44-CST (Wed)
From: Makedon%utd-cs.csnet@csnet-relay.ARPA
To: udi%rsch.wisc.edu@csnet-relay.ARPA
Subject: VLSI workshop announcement
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 31 Oct 85 21:30:34 CST (Thu)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

      		C A L L   F O R   P A P E R S
      
      AEGEAN WORKSHOP ON COMPUTING : VLSI ALOGORITHMS AND ARCHITECTURES
      	        	( A W O C - VLSI )
      
      	2ND INTERNATIONAL WORKSHOP ON PARALLEL COMPUTING AND VLSI
      
      			JULY 8 - 11, 1986
      			ATTICA, GREECE
      
      ORGANIZED BY : THE COMPUTER TECHNOLOGY INSTITUTE OF GREECE (C.T.I)
      
      SPONSORED BY : EATCS AND THE GENERAL SECRETARIAT OF SCIENCE AND
                     TECHNOLOGY, OF THE GREEK MINISTRY OF INDUSTRY,
                     ENERGY  AND TECHNOLOGY
      
      	PROGRAM CHAIRPERSONS:  Fillia Makedon & Paul Spirakis
      
      TOPICS
        Topics include, but are not limited to:
      
      	MODELS OF VLSI ARCHITECTURE	SYSTOLIC ALGORITHMS
      	PARALLEL ALGORITHMS		CAD ALGORITHMS
      	LAYOUT ALGORITHMS		FAULT TOLERANCE
      	DISTRIBUTED COMPUTING		VLSI TESTING
      
      PAPERS
        Authors are invited to submit six copies of a ten-page long
        extended abstract BEFORE February 1, 1986 to one of the
        following two addresses:
      
      	PROF. KURT MEHLHORN	    PROF. FILLIA MAKEDON
      	(PROGRAM CHAIRMAN)	    DEPT. OF COMPUTER SCIENCE
      	FB 10 INFORMATIK	    BOX 830688, M/S FN 3.3
      	UNIVERSITAT DES SAARLANDES  THE UNIV. OF TEXAS AT DALLAS
      	66 SAARBRUCKEN		    RICHARDSON, TX  75083-0688
      	WEST GERMANY		    TEL (214) 690-2182
      	TEL  049-681-3023428	    CSNET: Makedon@UTD-CS
      
        Authors will be notified of ACCEPTANCE/REJECTION BY MARCH 15,
        1986.  Final papers are due by May 5, 1986.
      
      PROCEEDINGS
        The proceedings of the workshop will be published in the 
        Lecture Notes of Computer Science Series of Springer-Verlag.
      
      
      PROGRAM COMMITTEE:
         K. MEHLHORN (CHAIRMAN), W. GERMANY    
         I. FILOTTI, FRANCE	                 J. REIF, USA
         S. HAMBRUSCH, USA		         A. ROSENBERG, USA
         T. LENGAUER, W. GERMANY	         P. SPIRAKIS, USA/GREECE
         U. LAUTHER, W. GERMANY	                 H. SUDBOROUGH, USA
         T. LEIGHTON, USA		         P. VITANYI, NETHERLANDS
         F. LUCCIO, ITALY		         C. PAPADIMITRIOU, USA/GREECE
         F. MAKEDON, USA                         T. PAPATHEODOROU, GREECE
      
      LOCAL ARRANGEMENTS COMMITTEE
         C. MANOLOPOULOS, CHAIR (C.T.I., GREECE)
         P. SPIRAKIS (USA, GREECE)
         F. MAKEDON (USA)
         T. PAPATHEODOROU (C.T.I. DIRECTOR, GREECE)
         E. PROTONOTARIOS (N.T.U.A., GREECE)
     












  
      FURTHER INFORMATION
         AWOC-VLSI 86 will be held at a sea-side resort near Athens.
         It will follow similar guidelines to the AMALFI-VLSI workshop
         of 1984.  The third day of the workshop will be left open for
         activities which will include discussions/meetings between
         VLSI theorists and invited VLSI designers from industry.
         Further details about the workshop (and the final program)
         will be sent to anyone who submits a paper.  Others should
         fill out the form provided below:
      
      NAME AND MAILING ADDRESS:←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
      
      ←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
      
      TELEPHONE NUMBER:←←←←←←←←←←←←←←←←←←←← CSNET ADDRESS:←←←←←←←←←←←←←←
      
      I PLAN TO: SEND A PAPER←←←←←←←← , PARTICIPATE←←←←←←←←←, (No. of
      
      accompanying persons)←←←←←←←←←← , I WISH FURTHER INFORMATION←←←←←←←←,
      
      SIGNATURE:←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
      
      PLEASE SEND TO:        PROF. PAUL SPIRAKIS
      	       	             COMPUTER TECHNOLOGY INSTITUTE OF GREECE
                             P.O. BOX 1122
                             261 10 PATRAS, GREECE 
                             TEL: 061-99-3174/3176
      

-------------
TN Message #5
-------------

∂01-Nov-85  0040	ACUFF@SUMEX-AIM.ARPA 	Explorers working at last   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  00:40:20 PST
Date: Fri 1 Nov 85 00:41:50-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorers working at last
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12155712278.85.ACUFF@SUMEX-AIM.ARPA>

   To the best of my knowledge, all of our current Explorers except for
3 (Rosenbloom, Helix, and one pool) are functional.  The other should join
the ranks in a matter of hours or a day or two.  There is a release 2.0
(read "Getting Started with Explorers at KSL" near a console, or in
Sumex:<LispM>Exp-Intro.mss) on each machine, but I haven't gotten 1.0.2
configured.  Let me know if you need it.

	Enjoy,
	-- Rich
-------

∂01-Nov-85  0631	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 5  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  06:31:48 PST
Received: from ucb-vax.berkeley.edu by SU-CSLI.ARPA with TCP; Fri 1 Nov 85 06:31:43-PST
Received: by UCB-VAX (5.29/5.14)
	id AA01165; Wed, 30 Oct 85 12:21:37 PST
Received: by cogsci (5.31/5.13)
	id AA01002; Wed, 30 Oct 85 12:05:18 PST
Date: Wed, 30 Oct 85 12:05:18 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8510302005.AA01002@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 5

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                                  Fall 1985
                     Cognitive Science Seminar - IDS 237A

                      Tuesday, November 5, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

        ``On the Intentional Contents of Mental States About Fictions''

                                 Edward Zalta
                 Postdoctoral Fellow in Philosophy at C.S.L.I.
           Acting Asst. Professor of Philosophy, Stanford University


             In this seminar, I present a theory of intentional objects
        some  of  which  seem to serve nicely as the contents of mental
        states about stories and dreams (no matter how bizarre they may
        be).  The theory yields a way of understanding utterances about
        particular fictional characters and particular  dream  objects.
        For  the  purposes  of  the  talk,  it  will make no difference
        whether one construes the  theory  ontologically  as  a  theory
        about  what  the  world  has to be like or has to have in it in
        order for us to characterize properly such  mental  states,  or
        whether  one  construes the theory as just a canonical notation
        for specifying  the  contents  of  (or  mental  representations
        involved  in)  such  states.   Either  way,  one is left with a
        domain over which operations may be defined to explain  how  we
        get  from one state to the next, and so the theory should be of
        interest to cognitive scientists.  The philosophical  basis  of
        my  work  lies in a theoretical compromise between the views of
        Edmund Husserl and Alexius Meinong, and it is  consistent  with
        classical logic.
        ---------------------------------------------------------------
        UPCOMING TALKS
        November 12:Robert Wilensky, Computer Science, UCB
        November 19:Richard Alterman, Computer Science, UCB
        November 26:Eve Clark, Linguistics, Stanford
        December   3:Bernard Baars, Langley Porter, UCSF
        ---------------------------------------------------------------
        ELSEWHERE ON CAMPUS
        Steven Pulos, UCB, will  discuss  ``Children's  conceptions  of
        computers''  at  the  SESAME  Colloquium on Monday, November 4,
        4:00pm, 2515 Tolman Hall.

        John Kruschki, UCB, will speak on ``Depth  and  the  Configural
        Orientation  Effect''  at  the Cognitive Psychology Colloquium,
        Friday, November 8, 4:00pm, Beach Room, 3105 Tolman Hall.
  

∂01-Nov-85  1030	ullman@su-aimvax.arpa 	talk today  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  10:30:10 PST
Received: by su-aimvax.arpa with Sendmail; Fri, 1 Nov 85 10:25:17 pst
Date: Fri, 1 Nov 85 10:25:17 pst
From: Jeff Ullman <ullman@diablo>
Subject: talk today
To: nail@diablo

Don't forget the talk by Naqvi at the CS545 seminar, 3:15 today.

∂01-Nov-85  1130	WINOGRAD@SU-CSLI.ARPA 	Lost coat   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  11:18:41 PST
Date: Fri 1 Nov 85 11:14:06-PST
From: Terry Winograd <WINOGRAD@SU-CSLI.ARPA>
Subject: Lost coat
To: folks@SU-CSLI.ARPA

I think I left a dark blue jacket in Ventura on Monday or Tuesday.
It has my name in the label.  If anyone has seen one around, please
let me know.  Thanks -- t
-------

∂01-Nov-85  1203	COLEMAN@SU-CSLI.ARPA 	help the Aussies  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  12:02:31 PST
Date: Fri 1 Nov 85 11:58:25-PST
From: Carolyn Coleman <COLEMAN@SU-CSLI.ARPA>
Subject: help the Aussies
To: folks@SU-CSLI.ARPA
cc: coleman@SU-CSLI.ARPA

Avery Andrews of Linguistics at the Australian National University is working
to get things set up computationally in his department.  He has sent me
several requests for information which I am not competent to provide.
Perhaps one of you helpful souls would like to fill him in:
1)Has anyone around CSLI produced a phonetic symbols font that could be used
or adapted to be used on a QMS LG1200 laser-printer?

2) Avery could send across his Scribe definitions for linguistics (including
LI, NLLT, Language and CUP bibliography formats) if anyone has use for such
things.  He would also be interested in receiving such goodies.

--They haven't set up (LA)-TEX on their computer yet, but it will happen 
fairly soon.

Avery's adress;
Avery Andrews <munnari!FACO.anu.oz!ANDALING@seismo.CSS.GOV>

Thanks in anticipation,  Carolyn.
-------

∂01-Nov-85  1328	@SU-SCORE.ARPA:LES@SU-AI.ARPA 	Computer Facility Needs 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  13:27:28 PST
Received: from SU-AI.ARPA by SU-SCORE.ARPA with TCP; Fri 1 Nov 85 13:15:14-PST
Date: 01 Nov 85  1309 PST
From: Les Earnest <LES@SU-AI.ARPA>
Subject: Computer Facility Needs 
To:   Faculty@SU-SCORE.ARPA 

We need your estimates of computer and other technical support needs over
at least the next five years.  These needs are to be compiled, argued, and
integrated into three documents that are under development:
  (1) a general 10 year plan for the department,
  (2) a description of support needs over the next 5 years that
      should be met by either the CSD Computer Facilities Group or
      outside entities such as LOTS,
  (3) a specific 2 year development plan for the Computer Facilities Group.
A response form is attached for your convenience.

Recognizing that most people's crystal balls get cloudy beyond a few
years, we think that it will suffice to get estimates from each group
covering the next five years and to use rather simple-minded extrapolation
to sketch departmental needs in the 1991-95 time period.  If anyone wishes
to predict their needs for the full 10 year term, however, we will welcome
the additional information.  Something on the order of one page of text is
about the right level of detail for this round.

We wish to use a predominently "top down" approach in formulating these
plans.  General needs should be tied to one or more departmental programs
(e.g. "Instruction," "Research," or "Administration and Support").  While
specific suggestions, recommendations, or examples of solutions are
welcome, we mainly want functional requirements.  For example, if you
recommend that we buy "25 XYZ machines," please also say something like
"in support of a course on Pernicious Programming involving about
125 students for two quarters each year with an average student needing
six hours/week on a terminal supporting PunkLisp."

You need not cover expansions needed to support the planned undergraduate
program.  We shall extract these needs from Jeff Ullman's report,
"Resource Requirements for an Undergraduate Program in Computer Science."

It is understood that research programs will generally pay their own way
as far as computer facilities are concerned.  We expect that some groups
plan to take responsibility for their own computer acquisitions and
system software development.  In such cases, descriptions of computing
resources need not be very detailed but we still need to know about
computer space needs and any known special environmental requirements such
as unusual electrical power, special communication needs, or large
archival storage needs.  Groups that plan to use departmental technical
facilities or who wish to consider this option should so state.

While we recognize that CSD technical, financial, and space resources are
limited, we plan to uninhibitedly accept and compile stated requirements
from all groups in the department at first.  This compilation will be made
available to whoever is interested and we will then attempt to reconcile
the demands with resource limitations by traditional methods: cost/benefit
analysis, argumentation and political infighting.

The Facilities Committee will start thrashing on this November 11.
A timely response will be appreciated.  Please send your projected needs,
questions or gripes to me, preferably by electronic mail.

	Les Earnest (les@SAIL)
	Facilities Committee Chair

-----------------------------------

To:  CSD Facilities Committee

Subject: Estimated Computer Support Needs, 1986-1990

Name of Activity:


Instruction:
					Terminal
					Hrs./Week
  Course	Qtr./Yr.  # Students	/Student    Languages & Special Needs







Research:

  Departmental		# of People
 Service Needed		(Full Time Equiv.)	Languages & Special Needs



 Project-Developed	Space Required		Special Needs
 Computer Facilities	(Square Feet)		(e.g. communications, archiving)



Administration and Support:

			# of People
 Service Needed		(Full Time Equiv.)	Languages & Special Needs




∂01-Nov-85  1459	CHEADLE@SU-SCORE.ARPA 	Lost glasses
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  14:58:55 PST
Date: Fri 1 Nov 85 14:53:35-PST
From: Victoria Cheadle <CHEADLE@SU-SCORE.ARPA>
Subject: Lost glasses
To: csd@SU-SCORE.ARPA
Office: Margaret Jacks 258, 497-1519
Message-ID: <12155867333.33.CHEADLE@SU-SCORE.ARPA>


If anyone finds a pair of pink-framed glasses in a maroon case
around MJH, please return them to me.  Some time between lunch
and 2:00, I misplaced them.  The last place I remember seeing
them is right outside the rear entrance into MJH (1st floor).

Thanks,

Victoria
-------

∂01-Nov-85  1604	RINDFLEISCH@SUMEX-AIM.ARPA 	IBM S.J. Research Computer Science seminars 4 - 8 NOV 85 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  16:03:32 PST
Return-Path: <calendar.sjrlvm1%ibm-sj.csnet@CSNET-RELAY.ARPA>
Received: from CSNET-RELAY.ARPA by SUMEX-AIM.ARPA with TCP; Fri 1 Nov 85 14:12:05-PST
Received: from ibm-sj by csnet-relay.csnet id ar12736; 1 Nov 85 16:19 EST
Date: Wed, 30 Oct 85 18:19:35 PST
From: IBM San Jose Research Laboratory Calendar <calendar%ibm-sj.csnet@CSNET-RELAY.ARPA>
Reply-To: IBM-SJ Calendar <CALENDAR%ibm-sj.csnet@CSNET-RELAY.ARPA>
To: CSNET <"IBM Computer Science Subset distribution list"@ibm-sj.CSNET>
Subject: IBM S.J. Research Computer Science seminars 4 - 8 NOV 85
ReSent-Date: Fri 1 Nov 85 16:03:44-PST
ReSent-From: T. C. Rindfleisch <Rindfleisch@SUMEX-AIM.ARPA>
ReSent-To: SSRG-Systems-Staff@SUMEX-AIM.ARPA, AAP@SUMEX-AIM.ARPA
ReSent-Message-ID: <12155880103.24.RINDFLEISCH@SUMEX-AIM.ARPA>

                 IBM San Jose Research Lab
                      5600 Cottle Road
                     San Jose, CA 95193
 
                          CALENDAR
                   November 4 - 8, 1985
 
 
Mon., Nov. 4   Horizon Lecture Series
1:00 P.M.      THE TELECOMMUNICATIONS ENVIRONMENT AND CHALLENGES
Auditorium     Dr. Oshman will discuss ROLM's history and the
               telecommunications environment facing IBM and ROLM.
 
 
               Dr. K. Oshman, IBM Vice President
                 and President of ROLM Corporation
               Host:  F. Mayadas
 
Tues., Nov. 5  Computer Science Seminar
2:00 P.M.      SPUR - SYMBOLIC PROCESSING USING RISCS
Aud. A         We describe the architecture of a multiprocessor
               risc workstation under development at
               University of California, Berkeley.  Each
               processor consists of a custom VLSI CPU, an
               integrated cache controller/memory management
               unit, and a floating point co-processor.  The
               processor includes architectural support for
               Common LISP, the cache controller implements a
               hardware multiprocessor cache coherency scheme,
               the memory management unit performs high
               performance virtual memory translation without a
               TLB, and the floating point unit implements the
               I.E.E.E. standard.  Six faculty and twenty
               students are involved in the project, which
               spans from i.c. design to programming
               environments research.  The project is being
               supported by DARPA under the Strategic Computing
               Initiative.
 
               R. Katz, Computer Science Division,
                 University of California, Berkeley
               Host:  L. Cabrera
 
Thurs., Nov. 7 Computer Science Seminar
2:00 P.M.      THE SPRITE NETWORK OPERATING SYSTEM
Aud. B         Sprite is a new network operating system built
               by a team of graduate students and myself as
               part of the SPUR workstation project.  The talk
               will focus on three parts of Sprite:  the
               filesystem, process offloading, and the virtual
               memory system.  The filesystem will provide a
               single shared Unix-like file hierarchy
               distributed across several servers.  Clients
               will use dynamically-constructed prefix tables
               to associate file names with servers.  Sprite
               will include a process offloading mechanism that
               allows jobs to be run on idle workstations in
               the same way that jobs may be placed in
               background in Unix.  The virtual memory system
               will be Unix-like with simple extensions to
               permit shared data segments and synchronization.
               I'll talk about how changes in technology have
               influenced the design of Sprite and try to
               convince you that a) a simple network filesystem
               eliminates the need for other network protocols,
               RPC, name servers, and the like; b) magnetic
               disks will soon be obsolete; and c) paging is
               also about to be obsolete (sort of).
 
 
               J. Ousterhout, Computer Science Division,
                 University of California, Berkeley
               Host:  L. Cabrera
 
-----------------------------------------------------------------------
 
For further information on individual talks, please contact the host
listed above.
 
Visitors, please arrive 15 minutes early.  IBM is located on U.S.
101, about 7 miles south of Interstate 280.  Exit at Monterey Road
(82) and turn right if you took 101 south (left for 101 north.)
Continue straight, ignoring the sign for 82, then follow signs for
Cottle Road.  The Research Laboratory is IBM Building 028.
For more detailed directions, please phone the Research Lab
receptionist at (408) 256-3028.
 
IBM San Jose Research mails out both the complete research calendar
and a computer science subset calendar.  Send requests for inclusion
in either electronic mailing list to CALENDAR at IBM-SJ.CSNET
(CALENDAR at SJRLVM1 on VNET), specifying computer science or general.
 

∂01-Nov-85  1644	TAJNAI@SU-SCORE.ARPA 	honors  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  16:43:43 PST
Date: Fri 1 Nov 85 16:44:30-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: honors
To: Faculty@SU-SCORE.ARPA
Message-ID: <12155887525.12.TAJNAI@SU-SCORE.ARPA>


If you received a special honor in 1985, please let me know.  
I know of John McCarthy received the first Award for Research
Excellence at the IJCAI.  George Dantzig was elected to the National
Academy of Engineering and won the 1985 Harvey Prize in the field of
science and technology by Technion.  

Don't be modest.  Let me know.

Carolyn
-------

∂01-Nov-85  1656	INGRID@SU-CSLI.ARPA 	House for Rent
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  16:55:57 PST
Date: Fri 1 Nov 85 16:53:19-PST
From: Ingrid Deiwiks <INGRID@SU-CSLI.ARPA>
Subject: House for Rent
To: SU-BBoards@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA


***********************************************************************
HOUSE FOR RENT *** HOUSE FOR RENT *** HOUSE FOR RENT *** HOUSE FOR RENT
***********************************************************************

3  1/2 bedrooms, 2 bathrooms, completely new kitchen with all gadgets.
Garden.  Located in Palo Alto, in walking distance of main library and
two parks.  Very close to schools.  For family only.  Terms of rent:
Two months starting November 11.  $1,000 per month. 
Please call 322-0187.

PLEASE DO NOT REPLY TO THIS MESSAGE BUT CALL THE ABOVE NUMBER.
-------

∂01-Nov-85  1752	COLEMAN@SU-CSLI.ARPA
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  17:52:06 PST
Date: Fri 1 Nov 85 17:49:02-PST
From: Carolyn Coleman <COLEMAN@SU-CSLI.ARPA>
To: folks@SU-CSLI.ARPA
cc: coleman@SU-CSLI.ARPA


-------

∂01-Nov-85  1822	ACUFF@SUMEX-AIM.ARPA 	Leaving Explorers 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  18:21:54 PST
Date: Fri 1 Nov 85 18:23:02-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Leaving Explorers
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12155905461.13.ACUFF@SUMEX-AIM.ARPA>

   I suggest leaving Explorers booted, but with the brightness turned
down, much as 36xx's are treated.

	-- Rich
-------

∂01-Nov-85  1829	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Call for papers - PODC    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 1 Nov 85  18:29:30 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 1 Nov 85 18:26:48-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Fri 1 Nov 85 18:26:52-PST
Received: by rsch.wisc.edu; Fri, 1 Nov 85 20:08:32 CST
Message-Id: <8511020154.AA06059@rsch.wisc.edu>
Received: from IBM-SJ.ARPA (ibm-sj.csnet) by rsch.wisc.edu; Fri, 1 Nov 85 19:54:30 CST
Date: 1 Nov 85 13:35:07 PST
From: HALPERN@IBM-SJ.ARPA
Subject: Call for papers - PODC
To: theory@rsch.wisc.edu
Status: R
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 01 Nov 85 20:08:20 CST (Fri)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

                       CALL FOR PAPERS:
5th ACM SIGACT-SIGOPS Symposium on Principles of Distributed Computing
                            (PODC)

                  Calgary, Alberta, Canada
                     August 11-13, 1986


Original research contributions are sought that address fundamental
issues in the theory and practice of distributed and concurrent systems.
Topics of interest include, but are not limited to,
the following aspects of concurrent and distributed systems:

* Principles of distributed computation derived from
  practical experience with working systems
* Algorithms and complexity
* Specification, semantics, and verification
* Programming languages and programming language constructs
* Fault tolerance

Important Dates:
Jan. 31, 1986: Abstracts due
Apr. 11, 1986: Authors informed of acceptance or rejection
May 16, 1986:  A final copy of each accepted paper due, typed on special
               forms for inclusion in the conference proceedings

Please send ELEVEN copies of a detailed abstract
(not the complete paper), with the address,
telephone number, and NET ADDRESS (if available) of a
contact author on the cover page, to the program chair:

     Dr. J. Halpern
     Department K53/801
     IBM Almaden Research Center
     650 Harry Road
     San Jose, CA 95120-6099

The abstract should be no more than 10 DOUBLE-SPACED TYPEWRITTEN PAGES.
It must include a clear description of the problem being discussed,
comparisons with extant work, and a section on major original
contributions.  There should be enough detail provided for the
program committee to make a decision.

SUBMISSIONS ARRIVING LATE OR DEPARTING SIGNIFICANTLY FROM THESE
GUIDELINES RISK REJECTION WITHOUT CONSIDERATION OF THEIR MERITS.

The Program Committee consists of:

  David Cheriton, Stanford
  Cynthia Dwork, IBM Almaden
  Nissim Francez, Technion
  Hector Garcia-Molina, Princeton
  Joseph Halpern, IBM Almaden
  Butler Lampson, DEC
  Richard Ladner, U. of Washington
  Paul Leach, Apollo Computer, Inc.
  Michael Merritt, AT&T Bell Laboratories
  Doron Rotem, Waterloo University/Lawrence Berkeley Labs
  Sam Toueg, Cornell

Conference Chair: Ernest Chang, Alberta Research Council
Publicity Chair: Tiko Kameda, Simon Fraser University

-------------
TN Message #6
-------------

∂02-Nov-85  1158	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.berkeley.edu  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 2 Nov 85  11:58:00 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Sat 2 Nov 85 11:55:25-PST
Received: from ucb-vax.berkeley.edu by SU-SCORE.ARPA with TCP; Sat 2 Nov 85 11:55:35-PST
Received: by ucb-vax.berkeley.edu (5.31/1.2)
	id AA20741; Sat, 2 Nov 85 11:40:53 PST
Received: by ernie (5.31/5.13)
	id AA14212; Sat, 2 Nov 85 11:40:39 PST
Date: Sat, 2 Nov 85 11:40:39 PST
From: karp@ernie.berkeley.edu (Richard Karp)
Message-Id: <8511021940.AA14212@ernie>
To: aflb.local@su-score.arpa, allmsgs@ernie.berkeley.edu,
        csfac@ernie.berkeley.edu, csmsgs@ernie.berkeley.edu,
        theory-b@ernie.berkeley.edu, theory-f@ernie.berkeley.edu

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley    (415)642-0143
Seminars in Computational Complexity
Tuesday, November 5    10:30  MSRI Lecture Hall
Deborah Joseph  "Separating Complexity Classes" (continued)

Tuesday, November 5    2:00   MSRI Lecture Hall
Ron Book    "Algorithmic Problems of Certain Finitely Presented Groups and
Monoids"

Tuesday, November 5    4:00    MSRI Lecture Hall
Ker-I Ko     "Applying Discrete Complexity Theory to Numerical Computation"

Wednesday, November 6  4:15     60 Evans
MSRI-EVANS WEDNESDAY LECTURES
Hendrik Lenstra   "Applied Number Theory"

Thursday, November 7  4:00      MSRI Lecture Hall
STRUCTURE IN COMPLEXITY THEORY
Juris Hartmanis  Title to be Announced

∂03-Nov-85  1923	LANSKY@SRI-AI.ARPA 	planlunch reminder -- Monday/11am/David Israel    
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 3 Nov 85  19:23:49 PST
Date: Sun 3 Nov 85 19:24:18-PST
From: LANSKY@SRI-AI.ARPA
Subject: planlunch reminder -- Monday/11am/David Israel
To: planlunch.dis: ;

	   THOUGHTS ON AN INVITATION TO A (SURPRISE) BEHEADING

			    David Israel
  	             SRI International, AI Center	
				
	 	    11:00 AM, MONDAY, November 4
       SRI International, Building E, Room EJ228 (new conference room)

Imagine you learn - from an authoritative source - that you will be
beheaded some afternoon of this coming week; say by Saturday at the
latest.  No doubt you'd be unnerved.  What, though, if you were also
told that the beheading would only take place if the following
condition were met: you would not know, say on the night before or the
morning of the event, that it was going to take place that afternoon.
So the beheading, if it is to take place at all, is to be a surprise.
Should your fears be allayed?  Indeed, shouldn't they be completely
eliminated?  There is a very plausible-seeming argument that you have
nothing to worry about. Unhappily, there seem to be lots of things
wrong this argument.  What's up?

I want to look at this puzzle, with an eye to connecting together a
variety of considerations about belief, knowledge, time, and - of all
things - rational action.  (This is going to be a PLANlunch seminar,
after all.)  So, I invite you all to a seminar that I shall give some
day between now and next Wednesday.  But, of course, you won't know,
say on the morning before the seminar,.........


-------

∂03-Nov-85  2051	BARWISE@SU-CSLI.ARPA 	Visitor 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 3 Nov 85  20:51:17 PST
Date: Sun 3 Nov 85 20:48:08-PST
From: Jon Barwise <BARWISE@SU-CSLI.ARPA>
Subject: Visitor
To: Folks@SU-CSLI.ARPA, bboard@SU-CSLI.ARPA


Professor Giuseppe Longo, from Pisa, will be visiting CSLI this week,
Monday through Thursday.  As you may know, he works in the area of
denotational semantics, models of the lambda calculus, and types.  If
you wish to contact him, send a message to Brown@csli, who will see
that it gets to him.  We will schedule a talk if a suitable time can
be found.

-Jon Barwise

-------

∂04-Nov-85  0917	INGRID@SU-CSLI.ARPA 	Symposia on AI
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  09:15:53 PST
Date: Mon 4 Nov 85 09:09:56-PST
From: Ingrid Deiwiks <INGRID@SU-CSLI.ARPA>
Subject: Symposia on AI
To: SU-Bboards@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA

	 	          ANNOUNCEMENT


             Soto Symposia on Artificial Intelligence
             ----------------------------------------

A series of three symposia on AI will be held in the Soto Lounge,
Wilbur Hall, the first three Mondays in November, at 7 p.m.  These
symposia are intended to inform Stanford students and others who are
interested about the nature of AI, its prospects, and the intellectual
and social issues to which it gives rise.

Monday, November 4, 7 p.m.

	"What is AI?"

	Barbara Grosz     Senior Computer Scientist, SRI
		          Consulting Professor of Computer Science,
		          Stanford

	Stan Rosenschein  Director of AI Lab, SRI
		          Consulting Professor of Computer Science,
		          Stanford

	Matt Ginsberg     Department of Computer Science, Stanford


Monday, November 11, 7 p.m.

	"Can we make computers that think and talk?"

	Terry Winograd    Professor of Computer Science, Stanford

	Brian Smith       Computer Scientist, Xerox PARC
		          Consulting Professor of Philosophy,
	                  Stanford			  

	Nils Nilsson      Professor of Computer Science and
		          Chair of Department of Computer Science, 
			  Stanford


Monday, November 18, 7 p.m.

	"Star Wars and Computation"

	John McCarthy     Professor of Computer Science, Stanford

	David Redell      DEC Research Laboratory

	and possibly others

Each symposium will begin with short talks by the speakers, followed
by plenty of time for vigorous interchanges among the speakers and
with members of the audience.

Soto is the first dormitory in the Wilbur complex that one comes to
when walking out Escondido from the Undergraduate Library.

	
----------------------------------------------------------------------
<--UGLI		              Escondido  Rd.
----------------------------------------------------------------------
                                       
xxxxxxxxxxxxxx                              X XXX     XXX X 
		                   Soto --> X   Wilbur    X		
Stern Hall                                  X             X

xxxxxxxxxxxxxx                              X    Hall     X
                                            X             X
                                            X XXX     XXX X




-------

∂04-Nov-85  1002	LIBRARY@SU-SCORE.ARPA 	Applied Numerical Mathematics--New Journal In Math/CS Library 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  10:01:52 PST
Date: Mon 4 Nov 85 10:00:34-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Applied Numerical Mathematics--New Journal In Math/CS Library
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12156600422.31.LIBRARY@SU-SCORE.ARPA>

We have received the first five issues of a new journals titled Applied
Numerical Mathematics.  The first issue of volume one is dated January
1985.  It is published by North-Holland and is subtitled transactions of
IMACS, International Association for Mathematics and Computers in Simulation.
The editors-in-chief are R. Vichnevetsky, Dept. of CS, Rutgers Univ. and
R. Stepleman, Exxon Research & Engineering.  

H. Llull
-------

∂04-Nov-85  1036	BARWISE@SU-CSLI.ARPA 	Talk on semantics of types in computer languages
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  10:35:53 PST
Date: Mon 4 Nov 85 10:32:52-PST
From: Jon Barwise <BARWISE@SU-CSLI.ARPA>
Subject: Talk on semantics of types in computer languages
To: BBoard@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA, CLT@SU-AI.ARPA

"The semantics of types and terms and their equivalences for various
lambda-calculi", by Prof. Giuseppe Longo, Wednesday, 12 noon, in the
Ventura Conference Room.

Abstract: Lambda calculus provides the core of functional programming
as it is based on the key notions of functional abstraction and
application.  The first part of the lecture will present an will
present an introductory account of the main type disciplines and their
semantics. First-order polymorphism and its motivations are also
surveyed.  In the second part, the semantic equivalence of typed terms
will be discussed.  The relation between types and terms gives us an
insight into second-order polymorphism (parametric types) and its
semantics.

(Please forward this to others who might be interested.)
-------

∂04-Nov-85  1140	HOLDEN@SU-CSLI.ARPA 	Friday afternoon drinking sessions
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  11:40:23 PST
Date: Mon 4 Nov 85 11:34:51-PST
From: Gary Holden <HOLDEN@SU-CSLI.ARPA>
Subject: Friday afternoon drinking sessions
To: Folks@SU-CSLI.ARPA, Linguists@SU-CSLI.ARPA

THIS IS YOUR FIRST (THOUGH NOT LAST) CHANCE TO CHECK OUT THE NEW
STUDENTS! (In a socially acceptable manner.)

COME ONE, COME ALL! TO THE WEEKLY SOIREE.

Place: The Greenberg Room, Linguistics Dept.

Date: Friday November 8 (and subsequent Fridays ad infinitum)

Time: 5 to 6 pm.
-------

∂04-Nov-85  1229	ACUFF@SUMEX-AIM.ARPA 	What Explorer files to keep 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  12:29:01 PST
Date: Mon 4 Nov 85 12:30:07-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: What Explorer files to keep
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12156627647.24.ACUFF@SUMEX-AIM.ARPA>

People with "personal" Explorers--

   Each Explorer has all the system source files and such on it.  For
people with Explorers in their offices, you can safely delete most if
not all of these directories, especially since these files will be
sought after on other hosts anyway.  You might be interested in the
TI-EXAMPLES and UNSUPPORTED directories, though you needn't keep a
copy on your own machine.

	-- Rich
-------

∂04-Nov-85  1250	ACUFF@SUMEX-AIM.ARPA 	Booting Lisp Machines  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  12:49:57 PST
Date: Mon 4 Nov 85 12:50:36-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Booting Lisp Machines
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12156631376.24.ACUFF@SUMEX-AIM.ARPA>

   We have arrived at the time when simply shouting "Can I boot?" is
no longer sufficient to insure that rebooting a TI or Symbolics lisp
machine at Welch Rd. will not burn someone else.  Given that there is
still a possibility of a machine you need to boot being in use by
someone else via the network (even in office machines, though this is
more remote), I suggest that the following protocol be followed when
booting:

1. Use Peek's Server display (System or Select P and then S) to see
   who, if anyone, is using your machine as a server.

2. Contact the person(s) using your machine, and tell them you'd like
   to reboot so that they have a chance to close down gracefully.
   Voice, intercom, and Converse are all good potential ways to
   contact the other users.

3. After the other users have given you the go ahead, use (si:shutdown).

   Please keep me informed of any problems with this scheme, and, if
necessary, I'll write software to automate the process somewhat.

	-- Rich
-------

∂04-Nov-85  1436	ULLMAN@SU-SCORE.ARPA 	new contract 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  14:36:34 PST
Date: Mon 4 Nov 85 14:17:23-PST
From: Jeffrey D. Ullman <ULLMAN@SU-SCORE.ARPA>
Subject: new contract
To: faculty@SU-SCORE.ARPA
cc: hanrahan@SU-SCORE.ARPA
Message-ID: <12156647176.15.ULLMAN@SU-SCORE.ARPA>

In discussions with Clive Liston it appeared to me that we do
not have to sign the new form if we have
a) signed an old PATENT agreement, and
b) are not taking UNIVERSITY funds for production of copyrightable
	material, e.g., writing code for hire for the library.
				---Jeff
-------

∂04-Nov-85  1546	GINSBERG@SUMEX-AIM.ARPA 	change to SIGlunch mailing list    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  15:46:17 PST
Date: Mon 4 Nov 85 15:39:57-PST
From: Matt Ginsberg <GINSBERG@SUMEX-AIM.ARPA>
Subject: change to SIGlunch mailing list
To: aam@SU-AI.ARPA, abarbanel@SUMEX-AIM.ARPA, adrienne@SU-SUSHI.ARPA,
    aiello@SUMEX-AIM.ARPA, ali@SU-SCORE.ARPA, alpert@SU-SCORE.ARPA,
    altenberg@SUMEX-AIM.ARPA, altman@SUMEX-AIM.ARPA, amarel@SUMEX-AIM.ARPA,
    andy@SU-SUSHI.ARPA, arun@SU-SCORE.ARPA, ashok@SU-SCORE.ARPA,
    bach@SU-SCORE.ARPA, banks@SUMEX-AIM.ARPA, barnes@SU-SCORE.ARPA,
    barnhouse@SUMEX-AIM.ARPA, barr@SUMEX-AIM.ARPA, bennett@SUMEX-AIM.ARPA,
    berlin@SU-SCORE.ARPA, bertrand@SUMEX-AIM.ARPA, bhayes-roth@SU-SCORE.ARPA,
    bischoff@SUMEX-AIM.ARPA, blum@SUMEX-AIM.ARPA, breese@SUMEX-AIM.ARPA,
    bright@SU-SCORE.ARPA, brinkley@SUMEX-AIM.ARPA, brouillet@SUMEX-AIM.ARPA,
    brown@SUMEX-AIM.ARPA, brutlag@SUMEX-AIM.ARPA, buchanan@SUMEX-AIM.ARPA,
    cck@SU-AI.ARPA, chou@SUMEX-AIM.ARPA, clancey@SUMEX-AIM.ARPA,
    clarkson@SU-SCORE.ARPA, combs@SUMEX-AIM.ARPA, condran@SUMEX-AIM.ARPA,
    cooper@SUMEX-AIM.ARPA, cwc@SU-AIMVAX.ARPA, croft@SU-SAFE.ARPA,
    damon@SU-SCORE.ARPA, davies@SU-SIERRA.ARPA, davies@SUMEX-AIM.ARPA,
    de2smith@SUMEX-AIM.ARPA, delagi@SUMEX-AIM.ARPA, dfeldman@SUMEX-AIM.ARPA,
    dirk@SU-PSYCH.ARPA, dlo@SU-AI.ARPA, dmc@SU-AI.ARPA, downs@SUMEX-AIM.ARPA,
    drogers@SUMEX-AIM.ARPA, edozien@SU-SCORE.ARPA, ejs@SU-AI.ARPA,
    engelmore@SUMEX-AIM.ARPA, erlbaum@SUMEX-AIM.ARPA, fagan@SUMEX-AIM.ARPA,
    feigenbaum@SUMEX-AIM.ARPA, frayman@SUMEX-AIM.ARPA, frd@SU-AI.ARPA,
    friedland@SUMEX-AIM.ARPA, fu@SUMEX-AIM.ARPA, gardner@SUMEX-AIM.ARPA,
    Gelman@SUMEX-AIM.ARPA, genesereth@SU-SCORE.ARPA, Gibbons@SU-SCORE.ARPA,
    Ginsberg@SUMEX-AIM.ARPA, ginn@SUMEX-AIM.ARPA, gluck@SU-PSYCH.ARPA,
    golding@SUMEX-AIM.ARPA, goldstein@SU-SCORE.ARPA, greep@SU-CSLI.ARPA,
    grosof@SU-SCORE.ARPA, haggerty@SUMEX-AIM.ARPA, hahn@SU-SCORE.ARPA,
    haken@SUMEX-AIM.ARPA, harvey@SUMEX-AIM.ARPA, hasling@SUMEX-AIM.ARPA,
    henager@SUMEX-AIM.ARPA, henjum@SU-SUSHI.ARPA, hewett@SUMEX-AIM.ARPA,
    hirsh@SUMEX-AIM.ARPA, holtzman@SUMEX-AIM.ARPA, horvitz@SUMEX-AIM.ARPA,
    hsu@SUMEX-AIM.ARPA, JED@SU-AI.ARPA, jimison@SUMEX-AIM.ARPA,
    jjf@SU-SCORE.ARPA, jk@SU-AI.ARPA, jmc-LISTS@SU-AI.ARPA, jmm@SU-AI.ARPA,
    johnmark@SU-WHITNEY.ARPA, Johnson@SUMEX-AIM.ARPA, jordan@SU-PSYCH.ARPA,
    joyce@SUMEX-AIM.ARPA, jvc@SU-SCORE.ARPA, kahn@SUMEX-AIM.ARPA,
    kahoun@SUMEX-AIM.ARPA, kapsner@SUMEX-AIM.ARPA, karnicky@SU-SCORE.ARPA,
    King@SUMEX-AIM.ARPA, koo@SUMEX-AIM.ARPA, kramer@SUMEX-AIM.ARPA,
    kuhn@SUMEX-AIM.ARPA, Kunz@SUMEX-AIM.ARPA, langlotz@SUMEX-AIM.ARPA,
    Lenat@SUMEX-AIM.ARPA, leu@SUMEX-AIM.ARPA, lgd@SU-AI.ARPA,
    lichtarge@SUMEX-AIM.ARPA, London@SUMEX-AIM.ARPA, mackinlay@SUMEX-AIM.ARPA,
    MAFFLY@SUMEX-AIM.ARPA, maslin@SUMEX-AIM.ARPA, mclaughlin@SUMEX-AIM.ARPA,
    meindl@SU-SIERRA.ARPA, merchant@SUMEX-AIM.ARPA,
    mis-students@SUMEX-AIM.ARPA, mis%su-psych@SU-SCORE.ARPA, ML@SU-AI.ARPA,
    msgs@SU-PSYCH.ARPA, musen@SUMEX-AIM.ARPA, nagle@SU-SCORE.ARPA,
    navarro@SUMEX-AIM.ARPA, nii@SUMEX-AIM.ARPA, Nilsson@SU-SCORE.ARPA,
    nishimura@SUMEX-AIM.ARPA, ng@SU-SUSHI.ARPA, nsingh@SUMEX-AIM.ARPA,
    novak@SUMEX-AIM.ARPA, pchen@SU-SCORE.ARPA, PCohen@SUMEX-AIM.ARPA,
    pednault@SU-SCORE.ARPA, petrone@SUMEX-AIM.ARPA, pkr@SU-AI.ARPA,
    qian@SU-SCORE.ARPA, rac@SU-AI.ARPA, rappaport@SUMEX-AIM.ARPA,
    RDG@SU-AI.ARPA, rennels@SUMEX-AIM.ARPA, rep@SU-AI.ARPA,
    richer@SUMEX-AIM.ARPA, Rindfleisch@SUMEX-AIM.ARPA, rohn@SUMEX-AIM.ARPA,
    rosenschein@SUMEX-AIM.ARPA, russell@SUMEX-AIM.ARPA, rww@SU-AI.ARPA,
    saraiya@SUMEX-AIM.ARPA, sbarnes@SUMEX-AIM.ARPA, scales@SUMEX-AIM.ARPA,
    selig@SU-SUSHI.ARPA, sg@SU-AI.ARPA, shawn@SU-SCORE.ARPA,
    shng@SUMEX-AIM.ARPA, Shortliffe@SUMEX-AIM.ARPA, sleeman@SUMEX-AIM.ARPA,
    smith@SUMEX-AIM.ARPA, Su-bboard@SU-SCORE.ARPA, Subramanian@SUMEX-AIM.ARPA,
    Suwa@SUMEX-AIM.ARPA, takahashi@SU-AI.ARPA, Teach@SUMEX-AIM.ARPA,
    Terry@SUMEX-AIM.ARPA, theodorou@SU-SCORE.ARPA, tthompson@SUMEX-AIM.ARPA,
    TOB@SU-AI.ARPA, treitel@SUMEX-AIM.ARPA, tsuji@SUMEX-AIM.ARPA,
    tthompson@SUMEX-AIM.ARPA, Tu@SUMEX-AIM.ARPA, Turner@SUMEX-AIM.ARPA,
    unruh@SUMEX-AIM.ARPA, val@SU-AI.ARPA, VGA@SU-AI.ARPA, Vian@SUMEX-AIM.ARPA,
    vp@SU-SCORE.ARPA, Walker@SUMEX-AIM.ARPA, wallace@SU-SCORE.ARPA,
    waltzman@SU-SCORE.ARPA, Warren@SRI-AI.ARPA, washington@SUMEX-AIM.ARPA,
    Welch-road@SUMEX-AIM.ARPA, whitney@SUMEX-AIM.ARPA,
    Wiederhold@SUMEX-AIM.ARPA, wilkins@SUMEX-AIM.ARPA, wulfman@SUMEX-AIM.ARPA,
    yan@SU-SCORE.ARPA, ym@SU-AI.ARPA, yue@SU-SCORE.ARPA
Message-ID: <12156662207.75.GINSBERG@SUMEX-AIM.ARPA>

A.#Internetλ,
    genesereth@λSU-SCORE.ARPA.#Internetλ,
    Gibbons@λSU-SCORE.ARPA.#Internetλ,
    Ginsberg@λSUMEX-AIM.ARPA.#Internetλ,
    ginn@λSUMEX-AIM.ARPA.#Internetλ,
    gluck@λSU-PSYCH.ARPA.#Internetλ,
    golding@λSUMEX-AIM.ARPA.#Internetλ,
    goldstein@λSU-SCORE.ARPA.#Internetλ,
    greep@λSU-CSLI.ARPA.#Internetλ,
    grosof@λSU-SCORE.ARPA.#Internetλ,
    haggerty@λSUMEX-AIM.ARPA.#Internetλ,
    hahn@λSU-SCORE.ARPA.#Internetλ,
    haken@λSUMEX-AIM.ARPA.#Internetλ,
    harvey@λSUMEX-AIM.ARPA.#Internetλ,
    hasling@λSUMEX-AIM.ARPA.#Internetλ,
    henager@λSUMEX-AIM.ARPA.#Internetλ,
    henjum@λSU-SUSHI.ARPA.#Internetλ,
    hewett@λSUMEX-AIM.ARPA.#Internetλ,
    hirsh@λSUMEX-AIM.ARPA.#Internetλ,
    holtzman@λSUMEX-AIM.ARPA.#Internetλ,
    horvitz@λSUMEX-AIM.ARPA.#Internetλ,
    hsu@λSUMEX-AIM.ARPA.#Internetλ,
    JED@λSU-AI.ARPA.#Internetλ,
    jimison@λSUMEX-AIM.ARPA.#Internetλ,
    jjf@λSU-SCORE.ARPA.#Internetλ,
    jk@λSU-AI.ARPA.#Internetλ,
    jmc-LISTS@λSU-AI.ARPA.#Internetλ,
    jmm@λSU-AI.ARPA.#Internetλ,
    johnmark@λSU-WHITNEY.ARPA.#Internetλ,
    Johnson@λSUMEX-AIM.ARPA.#Internetλ,
    jordan@λSU-PSYCH.ARPA.#Internetλ,
    joyce@λSUMEX-AIM.ARPA.#Internetλ,
    jvc@λSU-SCORE.ARPA.#Internetλ,
    kahn@λSUMEX-AIM.ARPA.#Internetλ,
    kahoun@λSUMEX-AIM.ARPA.#Internetλ,
    kapsner@λSUMEX-AIM.ARPA.#Internetλ,
    karnicky@λSU-SCORE.ARPA.#Internetλ,
    King@λSUMEX-AIM.ARPA.#Internetλ,
    koo@λSUMEX-AIM.ARPA.#Internetλ,
    kramer@λSUMEX-AIM.ARPA.#Internetλ,
    kuhn@λSUMEX-AIM.ARPA.#Internetλ,
    Kunz@λSUMEX-AIM.ARPA.#Internetλ,
    langlotz@λSUMEX-AIM.ARPA.#Internetλ,
    Lenat@λSUMEX-AIM.ARPA.#Internetλ,
    leu@λSUMEX-AIM.ARPA.#Internetλ,
    lgd@λSU-AI.ARPA.#Internetλ,
    lichtarge@λSUMEX-AIM.ARPA.#Internetλ,
    London@λSUMEX-AIM.ARPA.#Internetλ,
    mackinlay@λSUMEX-AIM.ARPA.#Internetλ,
    MAFFLY@λSUMEX-AIM.ARPA.#Internetλ,
    maslin@λSUMEX-AIM.ARPA.#Internetλ,
    mclaughlin@λSUMEX-AIM.ARPA.#Internetλ,
    meindl@λSU-SIERRA.ARPA.#Internetλ,
    merchant@λSUMEX-AIM.ARPA.#Internetλ,
    mis-students@λSUMEX-AIM.ARPA.#Internetλ,
    mis%su-psych@λSU-SCORE.ARPA.#Internetλ,
    ML@λSU-AI.ARPA.#Internetλ,
    msgs@λSU-PSYCH.ARPA.#Internetλ,
    musen@λSUMEX-AIM.ARPA.#Internetλ,
    nagle@λSU-SCORE.ARPA.#Internetλ,
    navarro@λSUMEX-AIM.ARPA.#Internetλ,
    nii@λSUMEX-AIM.ARPA.#Internetλ,
    Nilsson@λSU-SCORE.ARPA.#Internetλ,
    nishimura@λSUMEX-AIM.ARPA.#Internetλ,
    ng@λSU-SUSHI.ARPA.#Internetλ,
    nsingh@λSUMEX-AIM.ARPA.#Internetλ,
    novak@λSUMEX-AIM.ARPA.#Internetλ,
    pchen@λSU-SCORE.ARPA.#Internetλ,
    PCohen@λSUMEX-AIM.ARPA.#Internetλ,
    pednault@λSU-SCORE.ARPA.#Internetλ,
    petrone@λSUMEX-AIM.ARPA.#Internetλ,
    pkr@λSU-AI.ARPA.#Internetλ,
    qian@λSU-SCORE.ARPA.#Internetλ,
    rac@λSU-AI.ARPA.#Internetλ,
    rappaport@λSUMEX-AIM.ARPA.#Internetλ,
    RDG@λSU-AI.ARPA.#Internetλ,
    rennels@λSUMEX-AIM.ARPA.#Internetλ,
    rep@λSU-AI.ARPA.#Internetλ,
    richer@λSUMEX-AIM.ARPA.#Internetλ,
    Rindfleisch@λSUMEX-AIM.ARPA.#Internetλ,
    rohn@λSUMEX-AIM.ARPA.#Internetλ,
    rosenschein@λSUMEX-AIM.ARPA.#Internetλ,
    russell@λSUMEX-AIM.ARPA.#Internetλ,
    rww@λSU-AI.ARPA.#Internetλ,
    saraiya@λSUMEX-AIM.ARPA.#Internetλ,
    sbarnes@λSUMEX-AIM.ARPA.#Internetλ,
    scales@λSUMEX-AIM.ARPA.#Internetλ,
    selig@λSU-SUSHI.ARPA.#Internetλ,
    sg@λSU-AI.ARPA.#Internetλ,
    shawn@λSU-SCORE.ARPA.#Internetλ,
    shng@λSUMEX-AIM.ARPA.#Internetλ,
    Shortliffe@λSUMEX-AIM.ARPA.#Internetλ,
    sleeman@λSUMEX-AIM.ARPA.#Internetλ,
    smith@λSUMEX-AIM.ARPA.#Internetλ,
    Su-bboard@λSU-SCORE.ARPA.#Internetλ,
    Subramanian@λSUMEX-AIM.ARPA.#Internetλ,
    Suwa@λSUMEX-AIM.ARPA.#Internetλ,
    takahashi@λSU-AI.ARPA.#Internetλ,
    Teach@λSUMEX-AIM.ARPA.#Internetλ,
    Terry@λSUMEX-AIM.ARPA.#Internetλ,
    theodorou@λSU-SCORE.ARPA.#Internetλ,
    tthompson@λSUMEX-AIM.ARPA.#Internetλ,
    TOB@λSU-AI.ARPA.#Internetλ,
    treitel@λSUMEX-AIM.ARPA.#Internetλ,
    tsuji@λSUMEX-AIM.ARPA.#Internetλ,
    tthompson@λSUMEX-AIM.ARPA.#Internetλ,
    Tu@λSUMEX-AIM.ARPA.#Internetλ,
    Turner@λSUMEX-AIM.ARPA.#Internetλ,
    unruh@λSUMEX-AIM.ARPA.#Internetλ,
    val@λSU-AI.ARPA.#Internetλ,
    VGA@λSU-AI.ARPA.#Internetλ,
    Vian@λSUMEX-AIM.ARPA.#Internetλ,
    vp@λSU-SCORE.ARPA.#Internetλ,
    Walker@λSUMEX-AIM.ARPA.#Internetλ,
    wallace@λSU-SCORE.ARPA.#Internetλ,
    waltzman@λSU-SCORE.ARPA.#Internetλ,
    Warren@λSRI-AI.ARPA.#Internetλ,
    washington@λSUMEX-AIM.ARPA.#Internetλ,
    Welch-road@λSUMEX-AIM.ARPA.#Internetλ,
    whitney@λSUMEX-AIM.ARPA.#Internetλ,
    Wiederhold@λSUMEX-AIM.ARPA.#Internetλ,
    wilkins@λSUMEX-AIM.ARPA.#Internetλ,
    wulfman@λSUMEX-AIM.ARPA.#Internetλ,
    yan@λSU-SCORE.ARPA.#Internetλ,
    ym@λSU-AI.ARPA.#Internetλ,
    yue@λSU-SCORE.ARPA.#Internetλ
Message-ID: <12156662207.75.GINSBERG@SUMEX-AIM.ARPA>

∂04-Nov-85  1607	ACUFF@SUMEX-AIM.ARPA 	Symbolics shuffling and installation  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  16:07:42 PST
Date: Mon 4 Nov 85 16:06:39-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Symbolics shuffling and installation
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12156667068.26.ACUFF@SUMEX-AIM.ARPA>

   Due to hardware limitations (ie. short cables), 3674 has been
moved to the other end of the line of cubicles.  Then new 3640
should 74's old spot tomorrow, if all goes well.  It isn't clear when
the 3640 will be ready to be used, however, since it will probably
have 6.1, and otherwise take a lot of time to get synced with the
other Symbolics.

   74 persists in its network problems, so beware.  I'm trying to get
some likely boards swapped.

	-- Rich
-------

∂04-Nov-85  1618	SJG  	change to SIGlunch mailing list   
To:   "@TOTALS.DIS[1,SJG]"@SU-AI.ARPA 

I apologize for the previous message.  The content was intended to be the
following:

I have received several inquiries recently regarding the new policy to
send SIGlunch messages only to members of the KSL.  This is, I fear,
a pretty hard and fast rule, since we are trying to make these meetings
a bit smaller and more informal.  (But everyone is still welcome ...)
The only exceptions I am prepared to make are for SIGlunch speakers.
So, if you want to hear about the SIGlunches, come and give one!

					Matt Ginsberg

∂04-Nov-85  1651	ullman@su-aimvax.arpa 	more papers received today 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  16:50:48 PST
Received: by su-aimvax.arpa with Sendmail; Mon, 4 Nov 85 16:46:06 pst
Date: Mon, 4 Nov 85 16:46:06 pst
From: Jeff Ullman <ullman@diablo>
Subject: more papers received today
To: nail@diablo

"Safety and compilation of non-recursive Horn clauses," by Zaniolo
"On the implementation of a simple class of logic queries for
databases," by Sacca and Zaniolo

∂04-Nov-85  1702	ullman@su-aimvax.arpa 	paper received   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  16:31:34 PST
Received: by su-aimvax.arpa with Sendmail; Mon, 4 Nov 85 16:27:01 pst
Date: Mon, 4 Nov 85 16:27:01 pst
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

"Recognizing Pseudo-recursion in deductive databases,"
by Chomicki and Imielinski

Jeff N. need not quake.  They dwell on typed formulas, not
untyped as he does.  In fact, this paper seems to relate
both to the Maier/Ullman/Vardi "revenge of the JD" (not referenced)
and Sagiv's "real revenge of the JD" (referenced).

Anybody want copies?
				---Jeff U.

∂04-Nov-85  1703	REGES@SU-SCORE.ARPA 	lunch reminder
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  16:43:21 PST
Date: Mon 4 Nov 85 16:39:34-PST
From: Stuart Reges <REGES@SU-SCORE.ARPA>
Subject: lunch reminder
To: faculty@SU-SCORE.ARPA
Office: Margaret Jacks 210, 497-9798
Message-ID: <12156673059.34.REGES@SU-SCORE.ARPA>

Even though Nils is out of town, we will still have a faculty lunch tomorrow
starting at 12:15.  I have invited Mary Lou Allen, the director of the Stanford
Instructional Television Network, to come and discuss TV with us.  I believe
she plans to share some of her ideas about satellite hookups and other possible
future directions of the network.  I'm sure she will be willing to discuss any
issues surrounding TV as well.  I hope you can all attend.
-------

∂04-Nov-85  1720	@SU-SCORE.ARPA:WINOGRAD@SU-CSLI.ARPA 	Forum speakers   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  17:20:21 PST
Received: from SU-CSLI.ARPA by SU-SCORE.ARPA with TCP; Mon 4 Nov 85 17:21:15-PST
Date: Mon 4 Nov 85 17:19:21-PST
From: Terry Winograd <WINOGRAD@SU-CSLI.ARPA>
Subject: Forum speakers
To: faculty@SU-SCORE.ARPA, csl-faculty@SU-SIERRA.ARPA
cc: tajnai@SU-SCORE.ARPA


The following is the list of speakers suggested to date, as far as I
know.  If your entry does not correspond to what you think it should be,
let me know by the end of this week.  A "??" following a name means I
got no response.   It will transform into a "nobody" by next week unless
told otherwise.  This is the proposal list, and everyone on it may not
end up speaking. Also there have been some special sessions (e.g., a KSL
open house) suggested, yet to be worked into the schedule. --t

------------

 Binford
	Jean Ponce; new Research Associate
	David Chelberg 
	Jitendra Malik 
	Ron Fearing
	Michael Lowry: update
 Buchanan - David Wilkins
 Cheriton - Ross Finlayson
 Feigenbaum - ??
 Floyd  - ??
 Flynn - Chad Mitchell
 Genesereth - Jeff Finger
 Gill - ??
 Golub - ??
 Hennessy - Anant Agarwal
 Horowitz - Himself (new faculty)
 Knuth - Scott Kim
 Lantz - ??
 Linton - ??
 Luckham - ??
 Manna - Nobody
 Mayr - ??
 McCarthy - Ian Mason
 McCluskey - ??
 Nilsson - ??
 Oliger - Christina Fraley, Wei-Pai Tang
 Owicki - nobody
 Papadimetriou - Joe Mitchell, Esther Arkin, Vlad Rutenberg
 Pratt - ??
 Reid - ??
 Rosenbloom - ??
 Shortliffe - Glenn Rennels
 Ullman -
	Allen Van Gelder 
	Anna Karlin
	Jeff Naughton
	Kathy Morris
 Ungar - Himself (new faculty)
 Tobagi - separate session?
	Yitzhak Birk
	Hemant Kanakia
	M. Mehdi Nassehi, Fouad A. Tobagi and Michel E. Marhic
	David Shur
	James Storey
	Randy Neff
 Wiederhold - Winslett, Pednault?
 Winograd - nobody
 Yao - nobody --------
-------

∂04-Nov-85  2019	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Talk on semantics of types in computer languages  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  20:12:08 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 4 Nov 85 20:09:04-PST
Date: 04 Nov 85  1937 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Talk on semantics of types in computer languages 
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    



Speaker:  Professor Giuseppe Longo, visiting CSLI from Pisa

Title: The semantics of types and terms and their equivalences
       for various lambda-calculi

Time: Wednesday, November 6th, 12 noon

Place:  Ventura Conference Room.

Abstract: Lambda calculus provides the core of functional programming
as it is based on the key notions of functional abstraction and
application.  The first part of the lecture will present an will
present an introductory account of the main type disciplines and their
semantics. First-order polymorphism and its motivations are also
surveyed.  In the second part, the semantic equivalence of typed terms
will be discussed.  The relation between types and terms gives us an
insight into second-order polymorphism (parametric types) and its
semantics.



∂04-Nov-85  2048	@SU-CSLI.ARPA:dirk@SU-PSYCH 	Psychology Department Friday Seminar.    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  20:47:53 PST
Received: from SU-PSYCH by SU-CSLI.ARPA with TCP; Mon 4 Nov 85 20:45:31-PST
From: dirk@SU-PSYCH (Dirk Ruiz)
Received: from  by SU-PSYCH with TCP; Mon, 4 Nov 85 20:45:08 pst
Date: 04 Nov 85 20:44:59 PST (Mon)
To: friends@su-csli.ARPA
Subject: Psychology Department Friday Seminar.

Our speaker this week is Gyorgy Gergely.  Time and place are 3:15, Friday
(November 8, 1985) in room 100 of Jordan Hall.  Title and abstract follow:

------------------------------------------------------------------------

       Discourse Integrational Processes in Sentence Comprehension

                           Gyorgy Gergely

Classical models of sentence processing (e.g., Fodor, Bever & Garrett,
1974) developed in the universalist framework of Chomskian generative
grammar are examined critically from a functionalist comparative
perspective.  It is argued that earlier interpretations of on-line measures
of clausal processing (e.g., of the local increase of processing load
before the clause-boundary) lose their plausibility when considering a
class of languages that are typologically radically different from English.
Several experiments will be reported that examine clausal processing in
Hungarian, a non-Indo-European language, which, unlike English, has a)
`free' word order, b) marks underlying structural roles of NPs locally
unambiguously by case-marker suffixes, and c) encodes the discourse
functions of surface constituents syntactically.

The experiments demonstrate the existence of several kinds of discourse
integrational processes (such as `topic foregrounding' or focus-based
`inferential priming') which determine on-line measures of clausal
processing.  The results suggest that the local increase in processing load
at the end of the clause serves, to a large extent, across-clause discourse
integrational functions rather than within-clause functions of assigning
underlying structural representations, as previously supposed.  It is shown
that, during on-line processing, discourse segmentational cues, identifying
the informational focus (i.e., `new' information) and topic (i.e., `given'
information) of the clause, play a crucial role in directly mapping surface
sequences onto discourse representational structures.

------------------------------------------------------------------------



------- End of Forwarded Message

∂04-Nov-85  2138	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	POPL 86    
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  21:37:47 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Mon 4 Nov 85 21:37:23-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Mon 4 Nov 85 21:37:27-PST
Received: by rsch.wisc.edu; Mon, 4 Nov 85 23:01:08 CST
Received: from crys.wisc.edu by rsch.wisc.edu; Mon, 4 Nov 85 14:55:19 CST
Message-Id: <8511042056.AA28603@crys.wisc.edu>
Received: from CS.COLUMBIA.EDU by crys.wisc.edu; Mon, 4 Nov 85 14:56:22 CST
Date: Mon 4 Nov 85 15:47:27-EST
From: "Debra A. Jenkins" <JENKINS@CS.COLUMBIA.EDU>
Subject: POPL 86
To: theory@CRYS.WISC.EDU
Cc: galil@CS.COLUMBIA.EDU
Status: RO
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 04 Nov 85 23:00:40 CST (Mon)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

		THIRTEENTH ANNUAL ACM SYMPOSIUM 

			     ON
		
                PRINCIPLES OF PROGRAMMING LANGUAGES

		SPONSORED BY ACM SIGACT AND SIGPLAN

		        JANUARY 13-15, 1986
			
			   TRADE WINDS
		 SAINT PETERSBURG BEACH, FLORIDA



		    PREREGISTRATION FORM
		        ACM POPL 86


Fill in and mail to: Dr. Jerald Schwarz
		     AT&T
		     190 River Road, Rm. E-330
		     Summit, N.J. 07901

				       By Jan. 1	After Jan. 1

ACM and SIGACT or SIGPLAN member       $155←←←          $215←←←
Either ACM or SIG member........        165←←←		 225←←←
Nonmember.......................        190←←←		 250←←←
Student.........................	 40←←←		  60←←←

Name............................................................
Affiliation.....................................................
Address.........................................................
................................................................
................................................................

Telephone.......................................................

Payment must be in the form of a check made payable to ``ACM POPL 86
Symposium''.  No vouchers, purchase orders, or credit cards can be
accepted.  Requests for refunds received before January 1, 1986, will
be honored.
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

                       HOTEL RESERVATION FORM               
		           ACM POPL 86

Fill in and mail to: Trade Winds
		     5500 Gulf Blvd.
		     St. Petersburg Beach, Florida 33706
            Phone: 1-800-237-0707 or 813-367-6461

Reservations must be made before December 23, 1985, after which
requests will be on a space-available basis.  Children under 18 in
parent's room stay free.

Accomodations desired:	                   Dates:
←←←single or double at $58.95		   Arrival..........
←←←additional person at $5.00 each	   Departure........

Name........................................................
Affiliation.................................................
Address.....................................................
............................................................
............................................................

To guarantee for late arrival after 2 pm enclose check for one night
payable to ``Trade Winds'', or use (circle) one of American Express,
Diner's Club, Carte Blanche, Master Card, or Visa.

Card No......................... Exp. Date..................
Bank # if Master Card.......................................

			  POPL 86 PROGRAM

         Session 1: Monday, Jan. 13. 9:00 - 10:30
Chair: Charles N. Fischer, Univ. of Wisconsin---Madison

Remote Attribute Updating for Language-Based Editors, T. Reps, Univ.
of Wisconsin---Madison; C. Marceau and T. Teitelbaum, Cornell Univ.

Dynamically Bypassing Copy Rule Chains in Attribute Grammers,
R. Hoover, Cornell Univ.

Global Storage Allocation in Attribute Evaluation, H. Sasaki and T.
Katayama, Tokyo Institute of Technology

	  Session 2: Monday Jan. 13, 11:00 - 12:30
Chair: Charles N. Fischer, Univ. of Wisconsin---Madison

Finding the Source of Type Errors, M. Wand, Northeastern Univ.

A Maximum Flow Approach to Anomaly Isolation in Unification-Based
Incremental Type Inference, J. A. Walz, Cornell Univ.: G. F. Johnson,
Univ. of Maryland

Hierarchical VLSI Design Systems Based On Attribute Grammars, L. G.
Jones and J. Simon, Pennsylvania State Univ.

	  Session 3: Monday, Jan. 13, 2:00 - 3:30
Chair: Steven S. Muchnick, Sun Microsystems

Code Motion of Control Structures in High Level Languages, K. Zadeck,
R. Cytron and A. Lowry. IBM T. J. Watson Research Center, Yorktown
Heights, New York

Compilers and Staging Transformations, U. Jorring and W. L. Scherlis,
Carnegie-Mellon Univ.

A Set-Theoretic Characterization of Function Strictness in the Lambda
Calculus, P. Hudak and J. Young, Yale University

	  Session 4: Monday, Jan. 13, 4:00 - 6:00
Chair: Robert Henry, Univ. of Washington

Retargetable High-Level Alias Analysis, D. S. Coutant,
Hewlett-Packard, Cupertino, California

High-Quality Code Generation Via Bottom-Up Tree Pattern Matching, P.
J. Hatcher and T. W. Christopher, Illinois Institute of Technology

A Parallel Language and Its Compilation to Multiprocessor Machines or
VLSI, M. Chen, Yale Univ.

	       
	          REPORT FROM THE PROGRAM COMMITTEE

	   Survey/Tutorial: Tuesday, Jan. 14, 8:15 - 9:00
		     To be announced

	   Session 5: Tuesday, Jan. 14, 9:00 - 10:30
Chair: Fred B. Schneider, Cornell Univ.

Towards Programming with Knowledge Expressions, R. KurkiSuonio,
Tampere Univ. of Technology, Finland

Limitations of Synchronous Communication with Static Process Structure
in Languages for Distributed Computing, B. Liskov, M.I.T.; M. Herlihy,
C.M.U.; L. Gilbert, Autographix Inc.

Atomic Data Abstractions in a Distributed Collaborative Editing
System, I. Grief, R. Seliger and W. Weihl, M.I. T.

	   Session 6: Tuesday, Jan. 14, 11:00 - 12:30
Chair: Fred B. Schneider, Cornell Univ.

A Really Abstract Concurrent Model and Its Temporal Logic, H.
Barringer and R. Kuiper, Univ. of Manchester, England; A. Pneuli,
Weizmann Institute of Science, Isreal

Expressing Interesting Properties of Programs in Propositional
Temporal Logic, P. W@olper, AT&T Bell Laboratories

Operational Semantics of a Parallel Object-Oriented Language, P.
America, Philips Research Laboratories, The Netherlands; J. de Bakker,
J. N. Kok and J. Rutten, Centre for Mathematics and Computer Science,
Amsterdam, The Netherlands

	   Session 7: Tuesday, Jan. 14, 2:00 - 3:30
Chair: Michael O'Donnell, Univ. of Chicago

Equational Logic Programming: An Extension to Equational Programming,
J. You, Univ. of Utah; ,P. A. Subrahmanyam, AT&T Bell Laboratories

Logic Programming  and Inheritance, H. Ait-Kaci and R. Nasr,
Microelectronmics and Computer Technology Corp., Austin, Texas

Unification in Many-Sorted Algebras as a Device for Incremental
Semantic Analysis, G. Snelting and W. Henhapl, Technische Hochshule
Darmstadt, West Germany

	   Session 8: Tuesday, Jan. 14, 4:00 - 5:30
Chair: Peter Weinberger, AT&T Bell Laboratories

Distributed Data Structures in Linda, N. Carriero, D. Gelernter and J.
Leichter, Yale University

Para-Functional Programming: A Paradigm for Programming Multiprocessor
Systems, P. Hudak and L. Smith, Yale Univ.

Annotations for Distributed Programming in Logic, R. Ramakrishnan and
A. Silberschatz, Univ. of Texas

	   Survey/Tutorial: Wednesday, Jan. 15, 8:15 - 9:00
	              To be announced

	   Session 9: Wednesday, Jan. 15, 9:00 - 10:30
Chair: Dexter Kozen, Cornell University

Representation Independence and Data Abstraction, J. C. Mitchell, AT&T
Bell Laboratories

Using Dependent Types to Express Modular Structure: Experience with
Pebble and ML, D. B. MacQueen, AT&T Bell Laboratories

`Type' Is Not a Type, A. R. Meyer and M. B. Reinhold, M. I. T.

           Session 10: Wednesday, Jan. 15, 11:00 - 12:30
Chair: Steven S. Muchnick, Sun Microsystems

Data Flow Analysis of Applicative Programs Using Minimal Function
Graphs, N. D. Jones, DIKU, Denmark; A. Mycroft, Cambrdige Univ.,
England

A Mechanically Certified Theorem About Optimal Concurrency of Sorting
Netwroks, C. Huang and C. Lengauer, Univ. of Texas at Austin

Executable Specifications with Quantifiers in the FASE System, S.
Jefferson and S. Kamin, Univ. of Illinois at Urbana-Champaign


			GENERAL INMFORMATION
			       POPL 86

Location

The Thirteenth Annual ACM Symposium on Principles of Programming
Langauges, sponsored by ACM SIGACT and SIGPLAN, will be held January
13-15 (Monday - Wednesday) 1986, at the TradeWinds Resort, 5500 Gulf
Blvd., St. Petersburg Beach, Florida 33706.  A limousine service -
Limo - provides transportation 24 hours a day from Tampa International
Airport to the conference sight, approximately 30 miles away.  Go to
the booth just outside the baggage claim area.  Cost about $9.50.

By car, take I-75 to I-275 and south the St. Petersburg, exiting at
Pinellas Bayway to St. Petersburg Beach and Eckerd College.  Continue
across the Bayway to St. Petersburg Beach, at Gulf Blvd. ( the end of
the Bayway) turn right (north); the conference site is about one mile
on the left at the TradeWinds.

Accomodations

A block of rooms is reserved at the TradeWinds for POPL attendees.
All reservations must be made prior to December 23, 1985, either by
using the attached reservation form or by telephoning 800-237-0707 or
813-367-6461.  In the latter case you must mention that you will be
attending the ACM POPL symposium to get the conference rate.

Registration

The regular registration rates include two lunches, the reception
Sunday Evening January 12 from 8:00 pm to 10:00 pm, the dinner-cruise
on Tuesday, coffee breaks, and a copy of the proceedings.  The student
rate includes only the reception on Sunday, the coffee breaks, and a
copy of the proceedings.  The facilties for social functions will be
selected on the basis of the number of advance registrations ; there
is a possibility that some of the social functions will not have room
for you if you do not pre-register.  Registration fees rise by $60 on
January 1, 1986.

Conference Organization

Conference Co-Chairs: Mark Scott Johnson and Ravi Sethi: Program
Chair: Charles N. Fischer; Treasurer; Jerald Schwarz; Local
Arrangements: Edmund Gallizzi, Eckerd College, P. O. Box 12560, St.
Petersburg, FL 33733, 813-867-1166 ext. 272 .

About the Location

The TradeWinds is a resort on St. Petersburg Beach directly on the
Gulf of Mexico.  It has a breezy tropical ambiance, pools, saunas,
whirlpools, fully equipped exercise room, tennis and outdoor
racquetball courts, all beach and water sports, including sailing,
windsurfing, and windsurfing lessons.  Temperatures in January rabge
from the high 70's to the low 50's.  People go in the ocean all year
round.

Excursion

After the sessions on Tuesday, buses will take people from the hotel
to the Capt. Anderson for a three hour inland water cruise, including
live band and a sit-down dinner.
-------

-------------
TN Message #7
-------------

∂04-Nov-85  2213	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	BATS at DEC - November 22  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 4 Nov 85  22:13:43 PST
Received: from decwrl.DEC.COM by SU-SUSHI.ARPA with TCP; Mon 4 Nov 85 22:12:47-PST
Received: from magic.ARPA (magic) by decwrl.DEC.COM (4.22.01/4.7.34)
	id AA06338; Mon, 4 Nov 85 22:13:02 pst
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA14992; Mon, 4 Nov 85 22:12:36 pst
From: broder@decwrl.DEC.COM (Andrei Broder)
Message-Id: <8511050612.AA14992@magic.ARPA>
Date:  4 Nov 1985 2212-PST (Monday)
To: aflb.all@su-sushi
Cc: 
Fcc: sent
Subject: BATS at DEC - November 22


DEC-SRC in Palo Alto will host the next Bay Area Theory Seminar on
November 22.  Mark your calendar and spread the word.  I will post the
schedule (and maybe even the lunch menu) in about one week.

- Andrei

∂05-Nov-85  0046	DAVIES@SUMEX-AIM.ARPA 	No AAP meeting Wednesday   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  00:45:54 PST
Date: Tue, 5 Nov 1985  00:47 PST
Message-ID: <DAVIES.12156761908.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No AAP meeting Wednesday

Due to the lack of a speaker, there will be no Architecture Project
meeting this week.  A volunteer is eagerly sought for next week.

	-- Byron

∂05-Nov-85  1148	BERG@SU-SCORE.ARPA 	technical reports   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  11:45:14 PST
Date: Tue 5 Nov 85 11:43:43-PST
From: Kathy Berg <BERG@SU-SCORE.ARPA>
Subject: technical reports
To: faculty@SU-SCORE.ARPA, ras@SU-SCORE.ARPA
cc: berg@SU-SCORE.ARPA
Stanford-Phone: (415) 497-4776
Message-ID: <12156881346.33.BERG@SU-SCORE.ARPA>

In the interest of providing faster service, and to better anticipate
the space and resource needs of the publications office, I would like
to have some idea of the number of research papers that will be
submitted over the next two or three months.

If you are planning to have technical reports published, please
let me know the number of reports you will be preparing, and
the approximate number of pages of each report.

My thanks for your kind attention to this request.

Kathryn Berg
-------

∂05-Nov-85  1438	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  14:38:03 PST
Received: from ames-vmsb.ARPA by SU-SCORE.ARPA with TCP; Tue 5 Nov 85 14:36:06-PST
Date: 4 Nov 85 18:57:00 PST
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA


               ASSOCIATION FOR COMPUTING MACHINERY
                San Francisco Golden Gate Chapter
               "SIGBIG" Special Interest Committee
                 For Large High Speed Computers

 Meetings on  the first Wednesday of each month at 7:30 PM.   Speakers 
 who  can give insights to various aspects of  SUPERCOMPUTING are 
 featured each month.

 Next meeting:
     Wednesday, November 6,1985,  7:30 PM

     Topic:  Performance Measurement for High Speed Systems: A Panel

     Location:  Evans Hall  Room 608-7
		University of California Berkeley

     Directions:  Just North of the Campanile (belltower) 
		  at East entrance of campus.
     Panel:
     E. Miya, NASA Ames Research Center
     F. McMahon, LLNL
     D. Bailey, Informatics General Corp, NASA ARC
     K. Rowley, NASA Ames Research Center

   Metrics and tools of measuring high speed systems are critical for
 determining the characteristics of large systems.  There are no agreed
 upon measures: Whetstones, Dhrystones, MIPS, MFLOPS, LOPS, LIPS,
 and many more.  Current strategies include: running an arbitrary
 set of operations which "characterises" [from the English] a job mix,
 a sets of kernels or fundamental primitives used in scientific
 calculation, just running existing code, or specialized tests
 for functional units (vector), or just running a single subroutine
 in as identical an environment as possible like the Dongarra benchmark.
 Problems exist in characterizing the complete system versus portions
 of a system, measurement strategies and design methodologies, to adding
 multiple processors.  Three of the speakers will cover some of the better
 known benchmarking programs: the Livermore Loops and the newer NAS
 vector kernels.  A fourth panel member will discuss a new approach to
 performance measurement.

 ---------------------------------------------------------------
 Tape-recordings  of  most of the previous  may  be obtained
 in exchange for a tape cassette or $5.00 by contacting: 
                Mary Fowler (415)261-4058 (rec)
                Supercomputing  #192, BOX 2787
                Alameda, CA. 94501-0787

 For information contact Mary Fowler, Chairperson (415) 839-6547
                     or  Mike Austin, Publ. Chair (415) 423-8446

------

∂05-Nov-85  1519	LIBRARY@SU-SCORE.ARPA 	Socrates:  Searching By Call Number In The Technical Reports File  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  15:19:19 PST
Date: Tue 5 Nov 85 15:18:40-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates:  Searching By Call Number In The Technical Reports File
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA,
    : ;
Message-ID: <12156920477.17.LIBRARY@SU-SCORE.ARPA>

The Technical Reports file has a call number index.  The accession numbers
we assign to Math/CS technical reports are in this index.  However if you
wish to search by our accession number it must always be a six digit number.
Therefore if the report you are looking for is number 18203 you need to 
search by the number 018203.  Report number 9405 should be searched with
009405.

We will be investigating the possibility of dropping the leading 0's.

I would be interested to know if anyone finds it useful to be able to
search by our accession number.  Under what circumstances would you
have the accession number and not an author and title?

Harry Llull
-------

∂05-Nov-85  1543	LIBRARY@SU-SCORE.ARPA 	Socrates: How to Telnet--A Reminder  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  15:42:26 PST
Date: Tue 5 Nov 85 15:42:12-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates: How to Telnet--A Reminder
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12156924761.17.LIBRARY@SU-SCORE.ARPA>

To use Socrates from your departmental computers you can telnet for
forsythe.  First you give the command telnet.  Telnet> Forsythe.
Then when you receive -- login: -- type Forsythe again. login: Forsythe.
When you receive "enter class" type 20 and hit the return when it says
"class start".  When you use your Socrates account number and password to
login it will take you autormatically into Socrates.

If you don't have a Socrates account or an ITS account, you need to come
to the Math/CS Library to fill out a form for a Socrates account.

H. Llull
-------

∂05-Nov-85  1551	LIBRARY@SU-SCORE.ARPA 	Socrates: Telnet message correction  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  15:51:30 PST
Date: Tue 5 Nov 85 15:46:10-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates: Telnet message correction
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12156925481.17.LIBRARY@SU-SCORE.ARPA>

When you telnet there is no need to enter 20 for class.  This is done 
automatically.  When using a Gandalf you do need to enter class 20.
HL
-------

∂05-Nov-85  1620	@SU-SCORE.ARPA:EVAN@SU-CSLI.ARPA 	Re: Socrates:  Searching By Call Number In The Technical Reports File  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  16:20:43 PST
Received: from SU-CSLI.ARPA by SU-SCORE.ARPA with TCP; Tue 5 Nov 85 16:11:22-PST
Date: Tue 5 Nov 85 16:09:01-PST
From: Evan Kirshenbaum <evan@SU-CSLI.ARPA>
Subject: Re: Socrates:  Searching By Call Number In The Technical Reports File
To: LIBRARY@SU-SCORE.ARPA
cc: su-bboards@SU-SCORE.ARPA, faculty@SU-SCORE.ARPA, evan@SU-CSLI.ARPA
In-Reply-To: Message from "C.S./Math Library <LIBRARY@SU-SCORE.ARPA>" of Tue 5 Nov 85 15:26:03-PST

I actually have used this feature.  The circumstances were that I had done
a search and jotted down the numbers of the 10 or so reports that looked
promising.  When I went to the shelves, one of them was gone and three of
them were in boxes.  I then went back to Socrates with the call numbers
(all I had written down) to get more info so that I could look to see if
the reports had been published elsewhere (two had).

					evan
-------

∂05-Nov-85  1632	@SU-SCORE.ARPA:EVAN@SU-CSLI.ARPA 	Last Message (an apology) 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  16:32:51 PST
Received: from SU-CSLI.ARPA by SU-SCORE.ARPA with TCP; Tue 5 Nov 85 16:12:56-PST
Date: Tue 5 Nov 85 16:10:39-PST
From: Evan Kirshenbaum <evan@SU-CSLI.ARPA>
Subject: Last Message (an apology)
To: faculty@SU-SCORE.ARPA, su-bboards@SU-SCORE.ARPA

That was just supposed to go to the Library.  I really have to change
my default from REPLY ALL.

					evan
-------

∂05-Nov-85  1748	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #27  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  17:47:47 PST
Date:  5 Nov 85 1741-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #27
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 6 Nov 1985      Volume 1 : Issue 27

Today's Topics:

   Second PARSYM Survey: Data Abstractions for Parallel Programming

----------------------------------------------------------------------

Date: Tue, 5 Nov 1985  17:39 PST
From: DAVIES@SUMEX-AIM.ARPA
Subject: Second PARSYM Survey


Now that replies to the survey have tapered off, I offer summaries of
some other work involving data abstractions for parallel computing.
The Actors model provides a very general data abstraction for parallel
computing: the actor.  The Connection Machine notion of a xector
provides a powerful representation of aggregations of data for SIMD
computation.  Neither of these concepts is strictly a data
abstraction.  I think this is endemic to parallel computing: our data
abstractions need to be closely linked to the processes that compute
with them.  Both actors and xectors involve aspects of processing.
Actors also incorporate a powerful form of message-passing for
communication.


                           ACTORS

Dan Theriault, in his MIT thesis, discusses actors as data
abstractions (among other things):
	
     The actor model of computation is one in which many active,
     self-contained computing entities, called actors, process
     communications in parallel.  Each actor has its own
     processing power and storage.  Instead of having a notion of
     control flow, the actor model makes use of a more flexible
     idea of cooperation -- communication among entities which
     are under their own control.  Actors interact by
     transmitting information in "communications" to each other.

     An actor cannot directly view or manipulate the contents or
     implementation of another actor.  All it can do is
     communicate with the actor, asking it for information or
     requesting it to change.  Only the actor itself can alter
     its behavior.  This property is know by several names,
     including representation abstraction, protrection,
     encapsulation, opacity, and information-hiding. ...

     In addition to being opaque, an actor is entirely
     self-contained.  It can only communicate with its
     acquaintances and with the acquaintances of the
     communication it is processing. There is no notion of global
     state to put restrictions on the existence and location of
     the actor. ...

     Actors' properties of representation abstraction and
     absolute containment suggest the modularity inherent in the
     actor model.  The model goes beyond this, unifying data,
     control, and procedural abstractions.  [Since] an
     actor contains both data and procedural information (its
     acquaintances and script), [it] is ... sufficient for
     representing both procedures and data structures.  The
     model's emphasis on communication blurs the distinction
     between them.

Each actor has a set of acquaintances with whom it may communicate.
The acquaintance relationships between actors determine the topology
of the actor system.  The set of acquaintances of an actor may be
static or may change dynamically.  Actors may be composed
hierarchically: that is, an arbitrary network of communicating actors
may be packaged up as a single actor.

A list may be represented by a linked collection of actors (i.e.,
conses) each of which has two acquaintances, its car and its cdr, and
responds to two messages, CAR and CDR.  A pipeline of processes may be
represented by a collection of actors, each with one acquaintance.
Each pipeline element receives messages and does some processing
(executes its "script"), sending a message to its acquaintance.

A thesis by Gul Agha presents the Actors philosophy and explains why
Actors are a powerful mechanism for describing and implementing
concurrent processes on distributed computers.

References:

Gul A. Agha: Actors: A Model of Concurrent Computation in
Distributed Systems, MIT AI Lab Technical Report 844, March
1985.

Daniel G. Theriault: Issues in the Design and Implementation of Act2,
MIT EECS Thesis, 1983.



			        XECTORS

From "The Connection Machine":

     Connection Machine Lisp (CmLisp) is an extension of Common
     Lisp, designed to support the parallel operations of the
     Connection Machine. ...  In CmLisp, as in the Connection
     Machine itself, parallelism is achieved through simultaneous
     operations over composite data structures rather than
     through concurrent control structures.

     ...

     All concurrent operations in CmLisp involve a simple data
     structure called a xector (pronounced zek'tor).  A xector
     corresponds roughly to a set of processors with a value
     stored at each processor.  Because a xector is distributed
     across many processors, it is possible to operate on all its
     elements simultaneously.  To add two xectors together, for
     example, the Connection Machine directs each processor to
     add the corresponding values locally, producing a third
     xector of the sums.  This requires only a single addition
     time, even though the xector may have hundreds of thousands
     of elements.

     CmLisp supports many potentially concurrent operations to
     combine, create, modify, and reduce xectors.  These
     operations could be implemented on a conventional computer,
     but they would be much slower, perhaps tens of thousands of
     times slower, than they are on the Connection Machine.
     CmLisp also allows the programmer to define new xector
     operations that execute concurrently.  This is the source of
     its power.

To CmLisp, the world is composed of collections of thousands of slaves who
will carry out commands.  Xectors make it easy to describe groups
of slaves.  This is quite different from the Actors view.  To Actors,
the world is composed of many independent agents, each negotiating
with the others rather than commanding them.

The book gives a more complete description of the Connection Machine,
CmLisp, xectors as computational objects (to be read, printed,
operated upon, operated with) and a version of DEFSTRUCT for describing
xector entries.

In a chapter titled "Data Structures for the Connection Machine",
Hillis introduces "active data structures", a combination of data
structure and processor optimized for specific tasks.  In particular,
Hillis describes active versions of sets, trees, strings, arrays,
matrices, and graphs, all of which can be implemented as xectors.  For
example, the active version of a string contains a character per
processor.  Searching for a string S1 within another string S2 takes
time proportional to the length of S1.  On the first cycle, the first
character of S1 is used as a probe.  Each matching character cell in
S2 activates its successor in the string.  The successors are probed
to see if they match the second character in S1 and so on.  Since each
cycle happens in constant time, the search time is proportional to the
length of S1.  (What is the constant of proportionality?  What is the
appropriate cycle time for the Connection Machine, and how does it
scale with the total number of cells in the Connection Machine?)

Hillis's book is a masterpiece.  Beg, borrow, or steal it, or better
yet, buy it, but READ IT.


References:

W. Daniel Hillis: The Connection Machine.  MIT Press, 1985.

------------------------------

End of PARSYM Digest
********************

∂05-Nov-85  1914	ullman@su-aimvax.arpa 	meeting tomorrow 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  19:14:28 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 5 Nov 85 19:11:24 pst
Date: Tue, 5 Nov 85 19:11:24 pst
From: Jeff Ullman <ullman@diablo>
Subject: meeting tomorrow
To: nail@diablo

Shuky will discuss the Kanellakis/Cosmodakis paper.
Meet at 11AM in 301 MJH.

∂05-Nov-85  2305	ACUFF@SUMEX-AIM.ARPA 	3640 setup and ready for use
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 5 Nov 85  23:05:27 PST
Date: Tue 5 Nov 85 23:06:40-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: 3640 setup and ready for use
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12157005671.29.ACUFF@SUMEX-AIM.ARPA>

   KSL-3640-7 (also 3647 and 7) is now ready for use.  Its console
is located where 3674 used to be.  It has a very small disk, so there
is only one band (release 6.0 with IP and Vt100), and swapping; no
file space.  Enjoy, and let me know if it has problems...

	-- Rich
-------

∂06-Nov-85  0631	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 6 Nov 85  06:30:55 PST
Date: Wed 6 Nov 85 06:32:02-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: AFLB abstracts
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12157086748.13.PATASHNIK@SU-SUSHI.ARPA>

Here are abstracts for the next two AFLBs:

		**************************************

7-Nov-85  :  Alejandro A. Sch\"affer (Stanford)

		Shortest Prefix Strings Containing All
		k-Element Permutations as Subsequences

What is the length of the shortest string consisting of elements of
{1,...,n} that contains as subsequences all permutations of any
$k$-element subset? Many authors have considered the special case
where k=n. In fact, this special case was posed by D.E. Knuth
(attributed to R. M. Karp) as problem 36 in the technical report
STAN-CS-72-292. This problem arose in analyzing shortest path
algorithms and flowgraph reduction algorithms.

I shall instead consider an incremental variation on the problem first
proposed by P.J. Koutas and T.C. Hu (Discrete Math. 11(1975) pp.
125-132).  For a fixed value of n they ask for a string on the
alphabet {1,...,n} such that for all values of k <= n, the prefix
containing all permutations of any k-element subset as subsequences is
as short as possible.  The problem can also be viewed as follows.  For
k = 1 one needs n distinct digits to find each of the n possible
permutations. In going from k to k + 1, one starts with a string
containing all k-element permutations as subsequences, and one adds as
few digits as possible to the end of the string so that the new string
contains all k+1-element permutations.  The catch is that for all
values of k most shortest prefixes for k cannot be extended minimally
to prefixes for k+1.

Until now only very weak results were known about this version with
the prefix restriction posed by Koutas and Hu.  I shall describe an
*exact* solution using only *elementary* techniques.

I intend to make occasional use of the side blackboard in MJH352.

***** Time and place: November 7, 12:30 pm in MJ352 (Bldg. 460) ******

14-Nov-85  :  Richard Anderson (Stanford & MSRI)

             A Parallel Algorithm for Depth First Search

A new parallel algorithm for constructing a depth first search tree in
an undirected graph will be described.  The algorithm is a P-RAM
algorithm and uses several probabilistic algorithms as subroutines.
The run time of the algorithm is 2↑{\sqrt{\log n}}.  This makes it an
almost RNC algorithm, since the run time is less than n↑e for any e>0.

The standard sequential algorithm for depth first search can be shown
to be ``inherently sequential'', so this shows that substantial speed
up for depth first search is possible when a different approach is
taken.

***** Time and place: November 14, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

More information about future and past AFLB meetings and topics are in
the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂06-Nov-85  1409	ullman@su-aimvax.arpa 	paper received   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 6 Nov 85  14:09:44 PST
Received: by su-aimvax.arpa with Sendmail; Wed, 6 Nov 85 14:08:38 pst
Date: Wed, 6 Nov 85 14:08:38 pst
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

"The Inconsistency of Relational Database Theory" by
Roger Flynn of Univ. of Pittsburgh.

∂06-Nov-85  1742	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH  
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 6 Nov 85  17:42:27 PST
Date: Wed 6 Nov 85 17:41:42-PST
From: LANSKY@SRI-AI.ARPA
Subject: Next Monday's PLANLUNCH
To: planlunch.dis: ;

		MODEL THEORY FOR KNOWLEDGE AND BELIEF

			   Moshe Vardi
			   IBM San Jose
				
	 	    11:00 AM, MONDAY, November 11
       SRI International, Building E, Room EJ228 (new conference room)


Recently, there has  been  a surge of interest in the modal logic of
knowledge and belief, which has applications in many area of computer
science.  The  standard semantics for modal logic is Kripke semantics.
In this semantics, possible worlds and the possibility relation are both
primitive notions. This has both technical and conceptual shortcomings.
From a technical point of view, the mathematics associated with Kripke
semantics is often quite complicated.  From a conceptual point of view, 
it is not clear how to use Kripke structures to model knowledge and belief,
where one wants a clearer understanding of the notions that are taken as
primitive in Kripke semantics.

We introduce modal structures as models for modal logic. We use the idea
of possible worlds, but by directly describing the internal semantics of
each possible world.  It is much easier to study the standard logical
questions, such as completeness, decidability, and compactness, using
modal structures.  Furthermore, modal structures offer a much more
intuitive approach to modelling knowledge and belief.

As an application, we present a semantic model for knowledge
with the following properties: 

   (1) Knowledge is necessarily correct 

   (2) agents are logically omniscient, i.e., they know all 
       the consequences of their knowledge

   (3) agents are positively introspective, i.e., they are aware of
       their knowledge, but not negatively introspective, 
       i.e., they may not be aware of their ignorance.

We argue that this is the appropriate model for implicit knowledge.
We investigate the properties of the model, and use it to formalize
notions such as "to know more" and "all that is known is".


-------

∂06-Nov-85  2010	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	MFCS '86 Call for Papers  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 6 Nov 85  20:10:43 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 6 Nov 85 20:09:52-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Wed 6 Nov 85 20:09:26-PST
Received: by rsch.wisc.edu; Wed, 6 Nov 85 21:41:02 CST
Received: from crys.wisc.edu by rsch.wisc.edu; Wed, 6 Nov 85 10:20:35 CST
Message-Id: <8511061620.AA02294@crys.wisc.edu>
Received: from CS.COLUMBIA.EDU by crys.wisc.edu; Wed, 6 Nov 85 10:20:13 CST
Date: Wed 6 Nov 85 11:17:25-EST
From: "Debra A. Jenkins" <JENKINS@CS.COLUMBIA.EDU>
Subject: MFCS '86 Call for Papers
To: theory@CRYS.WISC.EDU
Cc: galil@CS.COLUMBIA.EDU
Status: R
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 06 Nov 85 21:40:45 CST (Wed)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

		       CALL FOR PAPERS

               TWELFTH INTERNATIONAL SYMPOSIUM		
		           MFCS '86
           MATHEMATICAL FOUNDATIONS OF COMPUTER SCIENCE

	            AUGUST 25 - 29, 1986
		 BRATISLAVA, CZECHOSLOVAKIA    
        ASSOCIATION OF SLOVAK MATHEMATICIANS AND PHYSICISTS 
                 SLOVAK ACADEMY OF SCIENCES

The International Symposium on Mathematical Foundations of Computer Science
		            MFCS '86

is the twelfth in the series of MFCS symposia organized in
Czechoslovakia and Poland.  The aim of the these symposia is to bring
together specialists in theoretical fields of computer science from
various counntries and to stimulate mathematical research in
theoretical computer science.

Principle areas of interest include:

- algorithms and data structures

- models of computation

- complexity and computability theory

- automata and formal langauges

- theory of logical design and layout of VLSI systems

- parallel and distributed computing

- semantics and logic of programs

- cryptography

- theory of data bases and knowledge based systems

- theory of robotics and artificial intelligence

This is not intended to be an exhaustive list.

Then scientific program of the symposium will include

- invited lectures covering areas of current interest

- short communications describing original research

The proceedings of the symposium will be available at the meeting.

LOCATION AND TIME

The symposium will be held in Bratislava, on August 25 - 29, 1986

SYMPOSIUM CHAIRMAN

          Branislzav Rovan

ORGANIZING SECRETARY

          Juraj Hromkovic

MAILING ADDRESS OF THE ORGANIZING COMMITTEE

          MFCS '86
	  Department of Theoretical Cybernetics
	  Komensky University
	  Mlysnka dolina
          842 15 Bratislava
	  Czechoslovakia

Telephone:320 003

ORGANIZING COMMITTEE

J. Bocek, A. Cerny, Z. Durayova, J. Gruska,
J. Hromkovic, S. Hudak, E. Kubincova,
P. Mikulecky, B. Rovan /chairman/ P. Ruzicka,
H. Stefanova, J. Wiederman

COOPERATING INSTITUTIONS

Komensky University, Bratislava
Institute of Technical Cybernetics of
the Slovak Academy of Sciences, Bratislava
VUSEI-AR, Bratislava
Purkyne University, Brno
Safarik University, Kosice
Charles University, Prague
Slovak Technical University, Bratislava

                    -------------

Please, use the attached REPLY CARD for a preliminary information to
the organizers and return it at your earliest convenience.

The number of participants is limited.

SUBMISSION OF PAPERS

Authors are invited to submit six copies of a full draft paper of the
size of 7-15 LNCS-like pages in English to the chairman of the Program
Committee BY JANUARY 15, 1986

Too short /long/ submissions risk that the Program Committee will not
be able to fully appreciate their merit because of the lack of
information /time/

Authors will be notified of acceptance or rejection on APRIL 1, 1986

The final text of accepted papers, typed on special forms, is due by
MAY 15, 1986 

PROGRAM COMMITTEE

A. Adachi /Tokyo/
G. Aussiello /Rome/
E. Borger /Dortmund/
L. Budach /Berlin/
J. W. de Bakker /Amsterdam/
A. P. Ershov /Novosibirsk/
R. Freivalds /Riga/
T. Gergely /Budapest/
J. Gruska /Bratislava/ -chairman
I. Guessarian /Paris/
P. Hajek /Prague/
M. Nivat /Paris/
T. Ottmann /Karlsruhe/
F. P. Preparata  /Urbana/
H. Rasiowa /Warsaw/
B. Rovan /Bratislava/
A. Salomaa /Turku/
I. H. Sudborough /Evanston/
J. van Leeuwen /Utrecht/
E. Wagner /Yorktown Heights/
W. Wechler /Dresden/
J. Wiedermann /Bratislava/ - secretary
D. Wood /Waterloo/

CHAIRMAN OF THE PROGRAM COMMITTEE

      Jozef Gruska
      MFCS '86
      Institute of Technical Cybernetics
      Slovak Academy of Sciences
      Dubravska 9
      842 37 Bratislava
      Czechoslavakia

Telephone: 375 000   Telex: 093355 tkasav



REPLY CARD                                       MFCS '86

Name←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←Ms/Mr/Dr/Prof
Address←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
       ←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
       Telephone:                      Telex:

Yes    No
←←←    ←←   I wish to attend the symposium.  Please send me further
            details and the registration form.

←←←    ←←   I wish to submit a paper. Its provisional title or subject.
            ←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

I expect to be accompanied by ←←←←←←←←←←←←←persons.
-------


-------------
TN Message #8
-------------

∂08-Nov-85  1647	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 12 (R. Wilensky, UCB)  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 8 Nov 85  16:47:06 PST
Received: from ucbvax.berkeley.edu by SU-CSLI.ARPA with TCP; Thu 7 Nov 85 19:32:59-PST
Received: by ucbvax.berkeley.edu (5.31/1.2)
	id AA16621; Thu, 7 Nov 85 17:27:35 PST
Received: by cogsci (5.31/5.13)
	id AA04733; Thu, 7 Nov 85 17:29:20 PST
Date: Thu, 7 Nov 85 17:29:20 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8511080129.AA04733@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 12 (R. Wilensky, UCB)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                     Cognitive Science Seminar - IDS 237A

                      Tuesday, November 12, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

             ``Knowledge Representation and a Theory of Meaning''
                                Robert Wilensky
                       Computer Science Division, U.C.B.

        Knowledge representation is central to most Artificial Intelli-
        gence   endeavors.    However,  most  knowledge  representation
        schemes are incomplete in a number  of  ways.   In  particular,
        their  coverage is inadequate, and they do not capture signifi-
        cant aspects of meanings.  Many do not  even  adhere  to  basic
        criteria of well-formedness for a meaning representation.

        KODIAK is a theory of  knowledge  representation  developed  at
        Berkeley  that  attempts to address some of these deficiencies.
        KODIAK incorporates representational ideas  that  have  emerged
        from  different  schools of thought, in particular from work in
        semantic networks, frames,  Conceptual  Dependency,  and  frame
        semantics.   In  particular,  KODIAK  eliminates the frame/slot
        distinction  found  in  frame-based  languages  (alternatively,
        case/slot distinction found in semantic network-based systems).
        In  its  place  KOKIAK  introduces  a  new  notion  called  the
        absolute/aspectual  distinction.   In addition, the theory sup-
        ports ``non-literal'' representations, namely, those  motivated
        by  metaphoric  and metonymic considerations.  Using these dev-
        ices, the theory allows for the representation  of  some  ideas
        that  in  the  past  have  only  been represented procedurally,
        informally, or not at all.

        KODIAK is being used to represent both linguistic  and  concep-
        tual   structures.   When  applied  to  the  representation  of
        linguistic knowledge, a new framework for talking about meaning
        emerges.   Five aspects of meaning have been identified.  These
        appear to  be  useful  in  describing  processing  theories  of
        natural language use.
        ----------------------------------------------------------------
        UPCOMING TALKS
        November 19: Richard Alterman, Computer Science, UCB
        November 26: Eve Clark, Linguistics, Stanford
        December  3: Bernard Baars, Langley Porter, UCSF
        ----------------------------------------------------------------
        ELSEWHERE ON CAMPUS
        Peter Pirolli will speak on ``Intelligent Tutoring Systems'' at
        the SESAME Colloquium on November 18, 2515 Tolman Hall, 4:00pm.
 

∂08-Nov-85  1649	@SU-CSLI.ARPA:barwise.pa@Xerox.ARPA 	"Machines and the mental"   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 8 Nov 85  16:49:05 PST
Received: from Xerox.ARPA by SU-CSLI.ARPA with TCP; Fri 8 Nov 85 09:11:26-PST
Received: from Cabernet.ms by ArpaGateway.ms ; 08 NOV 85 09:10:04 PST
Date: 8 Nov 85 08:58 PST
From: barwise.pa@Xerox.ARPA
Subject: "Machines and the mental"
To: Folks@SU-CSLI.ARPA
cc: Newsletter@SU-CSLI.ARPA
Message-ID: <851108-091004-1300@Xerox>

For complicated reasons, my TINLunch has been moved from Dacember to
next Thursday.  As a result, no abstract appeared in yesterday's
newsletter.  Here is one:

"Machines and the mental" by Fred Dretske

The paper argues that current computers do not exhibit anything that
deserves to be called rational cognitive activity.  Dretske even claims
that they can't add!  He then goes on to discuss how one might build
something that deserves to be called a rational machine.

This short, well-written paper is Dretske's Presidential Address to the
APA.  Read it can come prepared for a lively session.

∂08-Nov-85  1704	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #28  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 8 Nov 85  17:03:28 PST
Date:  8 Nov 85 0034-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #28
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest             Friday, 8 Nov 1985       Volume 1 : Issue 28

Today's Topics:

                   Automated Reasoning on the Cray
                         Actors = Smalltalk?
   Seminars: SPUR - Symbolic Processing Using RISCs (IBM San Jose)

----------------------------------------------------------------------

Date: Wednesday, 30 October 1985  18:16-PST
From: Bruce Smith <unc!bts%unc.csnet at CSNET-RELAY.ARPA>

(copied without permission from "Association of Automated
Reasoning Newsletter No. 5")

              Automated Reasoning on the CRAY
                      (from W. McCune)

     The automated reasoning system ITP was recently ported
to the CRAY X-MP-22 at the National Magnetic Fusion Energy
Computer Center in Livermore, California.  NMFECC has its
own operating system (CTSS), but the standard CRAY Research
Pascal compiler was used.  On a set of five benchmark
problems, speedups from a VAX 11/780 running Berkeley UNIX
4.2 to the CRAY ranged from 5.8 to 9.4.  This performance is
especially disappointing, considering that the Berkeley UNIX
Pascal compiler does not generate fast code.
     The following factors may explain the poor performance;
all are related to the fact that CRAYs are designed for
numeric computation.  First, the non-numeric sequential
computation of ITP cannot make use of the vectorization or
multitasking capabilities of the CRAY.  Second, the stack-
based execution required by the use of Pascal is rumored to
cause a slowdown of 2.  Finally, it seems likely that the
CTSS operating system does not use sophisticated techniques
for managing dynamically allocated memory.
     This experience supports the argument that there is a
need for ultra-high-performance hardware and software
tailored specifically for symbolic computing.

[Similar results have been reported with PSL (Portable Standard Lisp)
and the Gabriel benchmarks on the Cray (sorry, no reference handy, but
the work was done at Los Alamos, I believe).  With PSL, no vector
processing was attempted, so only the scalar speed mattered.  The Cray
is a whizbang processor, but it wasn't designed for Lisp.  The same
problem will undoubtedly apply to many other parallel architectures as
well.  Who's designing a multiprocessor Lisp machine? -- BD]

------------------------------

Date: 7 Nov 1985 17:34:18-EST
From: linus!brando@mitre-bedford.ARPA

With all due respect to those involved, isn't the whole Actors
paradigm essentially at one with object-oriented paradigms a la
Smalltalk, which have been around for quite some time now?  It seems
as if little more has been done than to change the names to protect
the innocent.  I don't mean to be provocative in any way -- I'm very
interested in finding out what differentiates these 2 ways of doing
business, other than using different terms to describe essentially the
same concepts.

Thom

[Objects -- from Smalltalk or even earlier from Simula -- were
certainly a motivation for Actors (see Carl Hewitt's early papers).
Unlike Smalltalk or Simula, Actors have additional features to support
distributed multiprocessing, including a buffered communication
network and distributed resource management.  In addition, the Actor
model is supported by an impressive foundation of philosophy and
theory, demonstrating in particular that the Actor model is strictly
more powerful than (some) other competing models of parallel
computing.

Would any Actor folks care to comment further?  -- BD]

------------------------------

Date: 8 Nov 1985 00:34:18-PST
From: Byron Davies <DAVIES@SUMEX-AIM.ARPA>
Re:   Seminars at IBM San Jose

Tues., Nov. 5  Computer Science Seminar
2:00 P.M.      SPUR - SYMBOLIC PROCESSING USING RISCS
Aud. A         We describe the architecture of a multiprocessor
               risc workstation under development at
               University of California, Berkeley.  Each
               processor consists of a custom VLSI CPU, an
               integrated cache controller/memory management
               unit, and a floating point co-processor.  The
               processor includes architectural support for
               Common LISP, the cache controller implements a
               hardware multiprocessor cache coherency scheme,
               the memory management unit performs high
               performance virtual memory translation without a
               TLB, and the floating point unit implements the
               I.E.E.E. standard.  Six faculty and twenty
               students are involved in the project, which
               spans from i.c. design to programming
               environments research.  The project is being
               supported by DARPA under the Strategic Computing
               Initiative.
 
               R. Katz, Computer Science Division,
                 University of California, Berkeley
               Host:  L. Cabrera
 
Thurs., Nov. 7 Computer Science Seminar
2:00 P.M.      THE SPRITE NETWORK OPERATING SYSTEM
Aud. B         Sprite is a new network operating system built
               by a team of graduate students and myself as
               part of the SPUR workstation project.  The talk
               will focus on three parts of Sprite:  the
               filesystem, process offloading, and the virtual
               memory system.  The filesystem will provide a
               single shared Unix-like file hierarchy
               distributed across several servers.  Clients
               will use dynamically-constructed prefix tables
               to associate file names with servers.  Sprite
               will include a process offloading mechanism that
               allows jobs to be run on idle workstations in
               the same way that jobs may be placed in
               background in Unix.  The virtual memory system
               will be Unix-like with simple extensions to
               permit shared data segments and synchronization.
               I'll talk about how changes in technology have
               influenced the design of Sprite and try to
               convince you that a) a simple network filesystem
               eliminates the need for other network protocols,
               RPC, name servers, and the like; b) magnetic
               disks will soon be obsolete; and c) paging is
               also about to be obsolete (sort of).
 
 
               J. Ousterhout, Computer Science Division,
                 University of California, Berkeley
               Host:  L. Cabrera

------------------------------

End of PARSYM Digest
********************

∂08-Nov-85  1826	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 8 Nov 85  18:26:15 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 8 Nov 85 18:24:50-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Fri 8 Nov 85 18:24:43-PST
Received: by rsch.wisc.edu; Fri, 8 Nov 85 20:00:59 CST
Message-Id: <8511090018.AA28827@rsch.wisc.edu>
Received: from IBM-SJ.ARPA (ibm-sj.csnet) by rsch.wisc.edu; Fri, 8 Nov 85 18:18:21 CST
Date: 8 Nov 85 16:19:07 PST
From: FAGIN@IBM-SJ.ARPA
To: theory@rsch.wisc.edu
Status: R
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 08 Nov 85 20:00:36 CST (Fri)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

  IBM San Jose Research has a new home, about 5 miles from our
  old Research Lab.  Our new name  is the IBM Almaden Research
  Center.  Our  new lab  has been  built on  690 acres  in the
  Almaden Valley of San Jose, only 40 acres of which are being
  developed.  Since  our Computer  Science Department  will be
  moving November 22-25, 1985, the purpose  of this note is to
  give  you  the  new  phone  numbers  and  addresses  of  the
  theoreticians in  our department.  You  can start  using the
  new  address right  away  (the mail  will  be forwarded,  if
  needed), and you  should use the new  phone numbers starting
  Monday, November 25.

  Ajtai, Miklos       (408) 927-1784      Department K53/801
  Backus, John        (408) 927-1848      Department K53/803
  Dwork, Cynthia      (408) 927-1752      Department K55/801
  Fagin, Ron          (408) 927-1726      Department K53/801
  Friedman, Joel      (408) 927-1807      Department K53/801
  Halpern, Joe        (408) 927-1787      Department K53/801
  Klawe, Maria        (408) 927-1714      Department K53/801
  Lucas, Peter        (408) 927-1847      Department K52/802
  Malachi, Yoni       (408) 927-1885      Department K52/803
  Megiddo, Nimrod     (408) 927-1786      Department K53/801
  Munshi, Ash         (408) 927-1808      Department K54/802
  Peleg, David        (408) 927-1790      Department K53/801
  Pippenger, Nick     (408) 927-1727      Department K51/801
  Rissanen, Jorma     (408) 927-1813      Department K54/802
  Shedler, Gerry      (408) 927-1783      Department K53/801
  Simons, Barbara     (408) 927-1785      Department K53/801
  Skeen, Dale         (408) 927-1755      Department K55/801
  Stockmeyer, Larry   (408) 927-1789      Department K53/801
  Strong, Ray         (408) 927-1758      Department K55/801
  Upfal, Eli          (408) 927-1788      Department K53/801
  Vardi, Moshe        (408) 927-1781      Department K55/802
  Wilber, Robert      (408) 927-1791      Department K53/801
  Williams, John      (408) 927-1888      Department K53/803
  Wimmers, Ed         (408) 927-1882      Department K53/803

  (Note: Wolfgang  Paul, who  is now in  physics, will  not be
  moving to the new lab for months).

  The  address  (where  you  should,  of  course,  supply  the
  appropriate department number) is:

       Department K53/801
       IBM Almaden Research Center
       650 Harry Road
       San Jose, California 95120-6099

  The CSNET/ARPANET  address is  usually lastname@ibm-sj  (for
  example,  Fagin@ibm-sj).  There  are a  few exceptions:  for
  example, Wolfgang Paul is Wolfgang@ibm-sj, Nick Pippenger is
  Nicholas@ibm-sj, Larry Stockmeyer is Stock@ibm-sj, Eli Upfal
  is Ely@ibm-sj.  We are now directly on Arpanet, so you don't
  need to  use Csnet-Relay.  The  BITNET/VNET address  is, for
  example, FAGIN at ALMVMA (with the same exceptions).


-------------
TN Message #9
-------------

∂08-Nov-85  1829	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 12 (R. Wilensky, UCB) 
Received: from UCBVAX.BERKELEY.EDU by SU-AI.ARPA with TCP; 8 Nov 85  18:28:49 PST
Received: by ucbvax.berkeley.edu (5.31/1.2)
	id AA16621; Thu, 7 Nov 85 17:27:35 PST
Received: by cogsci (5.31/5.13)
	id AA04733; Thu, 7 Nov 85 17:29:20 PST
Date: Thu, 7 Nov 85 17:29:20 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8511080129.AA04733@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 12 (R. Wilensky, UCB)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                     Cognitive Science Seminar - IDS 237A

                      Tuesday, November 12, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

             ``Knowledge Representation and a Theory of Meaning''
                                Robert Wilensky
                       Computer Science Division, U.C.B.

        Knowledge representation is central to most Artificial Intelli-
        gence   endeavors.    However,  most  knowledge  representation
        schemes are incomplete in a number  of  ways.   In  particular,
        their  coverage is inadequate, and they do not capture signifi-
        cant aspects of meanings.  Many do not  even  adhere  to  basic
        criteria of well-formedness for a meaning representation.

        KODIAK is a theory of  knowledge  representation  developed  at
        Berkeley  that  attempts to address some of these deficiencies.
        KODIAK incorporates representational ideas  that  have  emerged
        from  different  schools of thought, in particular from work in
        semantic networks, frames,  Conceptual  Dependency,  and  frame
        semantics.   In  particular,  KODIAK  eliminates the frame/slot
        distinction  found  in  frame-based  languages  (alternatively,
        case/slot distinction found in semantic network-based systems).
        In  its  place  KOKIAK  introduces  a  new  notion  called  the
        absolute/aspectual  distinction.   In addition, the theory sup-
        ports ``non-literal'' representations, namely, those  motivated
        by  metaphoric  and metonymic considerations.  Using these dev-
        ices, the theory allows for the representation  of  some  ideas
        that  in  the  past  have  only  been represented procedurally,
        informally, or not at all.

        KODIAK is being used to represent both linguistic  and  concep-
        tual   structures.   When  applied  to  the  representation  of
        linguistic knowledge, a new framework for talking about meaning
        emerges.   Five aspects of meaning have been identified.  These
        appear to  be  useful  in  describing  processing  theories  of
        natural language use.
        ----------------------------------------------------------------
        UPCOMING TALKS
        November 19: Richard Alterman, Computer Science, UCB
        November 26: Eve Clark, Linguistics, Stanford
        December  3: Bernard Baars, Langley Porter, UCSF
        ----------------------------------------------------------------
        ELSEWHERE ON CAMPUS
        Peter Pirolli will speak on ``Intelligent Tutoring Systems'' at
        the SESAME Colloquium on November 18, 2515 Tolman Hall, 4:00pm.
 

∂09-Nov-85  0513	VONHENKE@SRI-CSL.ARPA 	RISKS-1.22  
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 9 Nov 85  05:13:24 PST
Date: Fri 8 Nov 85 23:58:10-PST
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSL.ARPA>
Subject: RISKS-1.22
Sender: VONHENKE@SRI-CSL.ARPA
To: RISKS-LIST@SRI-CSL.ARPA

RISKS-LIST: RISKS-FORUM Digest  Wednesday, 16 Oct 1985  Volume 1 : Issue 22

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator
             Friedrich von Henke, guest moderator

Contents:
  Administratrivia (Friedrich von Henke)
  Medical software incidents (Nancy Leveson)
  European activities  (Udo Voges)
  Robots are different (Jerry Saltzer)
  Automobile computer control systems (Bennett Smith)
  Police computers (Dave Dyer)
  Electronic Surveillance (Geoffrey S. Goodfellow / Bill Keefe)
  Network Mailer Woes (Lynne Moore)
  Databases, grades, etc. (Karl Kluge, Andy Mondore, Mark Sienkiew)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Subject: Administratrivia
From: Friedrich von Henke (vonHenke@SRI-CSL)

I am temporarily acting as guest moderator of the Risks Forum.
Peter Neumann had to go rather abruptly on an overseas trip, and
the transition happened a bit disorderly.  My apologies to all of
you who were eagerly awaiting their twice-weekly cost of RISKS readings
but had to go without.  I hope to have things under control now.

Apparently the hiatus has had the effect of slowing down the stream of
contributions to a merely trickle; please don't hesitate to get the flow
going again!

			Friedrich von Henke

----------------------------------------------------------------------

Date: 11 Oct 85 19:39:27 PDT (Fri)
From: Nancy Leveson <nancy@UCI-ICSD.ARPA>
To: RISKS@sri-csl.ARPA
Subject: medical software incidents

I was just on a panel concerned with Software Safety at an IEEE
conference on Computers in Medicine and heard about some more
incidents involving software faults.  

The first was cited in a recent RISKS forum message (about the
programmable implanted pacemaker which was inadvertently reprogrammed by
emitted magnetic fields from an anti-theft device in a retail store),
but the patient
was cited as having survived.  Unfortunately, his weakened heart was
unable to stand the increased pace, and he died.

Other recalls by the FDA involve:

1)  An infusion-pump (used for insulin) had a software problem which
caused the infusion of insulin or dextrose to be delivered at the
maximum rather than the lower intended rate.  This occurred when certain
valid data was entered according to user instructions.

2)  A programmable pacemaker "locked-up" when being reset by an
external programming device.  Luckily this occurred in a doctor's office,
and the doctor was able to revive the patient.

3)  A multiple-patient monitoring system was recalled because the software
got patients' names mixed up with the wrong data.

4)  An algorithm was incorrectly programmed in a diagnostic lab instrument
which caused certain patient data to be reported erroneously as all zeros.

The reference for these incidents (for those who want to quote them) is:
  H. Bassen, J. Silberberg, F. Houston, W. Knight, C. Christman, and
  M. Greberman.  "Computerized Medical Devices:  Usage Trends, Problems,
  and Safety Technology," in Proc. 7th Annual Conference of IEEE 
  Engineering in Medicine and Biology Society, Sept. 27-30, 1985, 
  Chicago, Illinois, pp. 180-185.

Nancy Leveson
University of Calif. Irvine

------------------------------

Date:           Fri, 11 Oct 85 11:38:57 PDT
From:           Udo Voges <voges@LOCUS.UCLA.EDU>
To:             RISKS@SRI-CSL.ARPA
Subject:        European activities

I would like to bring some activities to your attention which are
going on in Europe, especially within and triggered by EWICS TC 7.

   The European Workshop on Industrial Computer Systems (EWICS), TC on
   Systems Reliability, Safety and Security (TC 7) is working since about
   10 years in this area, having some 100 members from industry, research
   and university. Previous work resulted in Position Papers on

       Development of safety related software
       Hardware of safe computer systems
       Guidelines for verification and validation of safety related software
       Guidelines for documentation of safety related computer systems
       Techniques for verification and validation of safety related software
       System requirements specification for safety related systems

   Current working areas include:

       System integrity
       Software quality assurance and metrics
       Design for system safety
       Reliability and safety assessment

   Besides conducting about four working meetings a year the TC is organizing
   the IFAC/IFIP Workshop on Achieving Safe Real-Time Computer Systems
   (Safecomp'79, '82, '83, '85, '86).

   The results of the work of TC 7 are introduced into the standardisation
   process (IEC, ISO, and national bodies) as well as used by companies
   and licensing authorities.

   Those interested in more information can either contact me or the
   current Chairman of TC 7: Mr. J.M.A. Rata, Electricite de France,
   1 Avenue du General de Gaulle,  F-92141 CLAMART   FRANCE.

   There exists an American counterpart to EWICS TC 7, but it was not
   possible to attract enough interested persons to keep it alive.
   The Japanese counterpart is also active, but due to the language
   barrier communication is minimal.

Udo

------------------------------

Date:  Sat, 12 Oct 85 01:30 EDT
From:  Saltzer@MIT-MULTICS.ARPA
Subject:  Robots are different
To:  risks@SRI-CSL.ARPA

When someone gets pinned to the wall by a robot, something different is going
on as compared to when someone gets gunned down by an FBI agent operating
under incorrect information retrieved from the NCIC.  Both cases may lead to
specific tragedies, yet the example of risks from robots seems to me to be
qualitatively different from many other computer-use risks.

The difference is that robots are used primarily in environments where
mechanically-oriented people are accustomed to balancing the risks of new
machinery against the benefits.  These people have, over the years, learned to
deal with risks from gear trains and drive belts, from swinging tailends on
steamshovels, from runaway elevators, from inadequately supported cranes.
They watch out, they learn from accidents, their insurers offer advice, they
make mistakes and take risks, and they learn.  To a first approximation, an
industrial robot presents a risk similar in kind to other new machinery, and
there is a moderately well-working system in place that is accustomed to
watching for the risks.  If anything, the average mechanic is suspicious of a
new piece of machinery in direct proportion to its complexity, newfangledness,
and gadgetry level, so is probably expecting the robot to screw up in
marvelous ways.  One might wish to argue with the particular balance that an
industry has struck between risks and benefits, but it is unusual to find one
in which mechanical risks are not understood at least moderately well.

The mechanic's suspicion of the new gadget is the essence of what seems to be
missing in many other applications of computers, and why it is so important to
raise awareness of the need to assess risks.  I'm not convinced we need to
harass our colleagues in the robot business with risk-consciousness-raising.
We should be instead talking to their installers to find out what we can
learn.

                                                  Jerry Saltzer

------------------------------

Date: Wed, 23 Oct 85 11:14:29 -0100
From: ircam!bks@seismo.CSS.GOV (Bennett Smith)
To: NEUMANN@SRI-CSL.ARPA
Subject: Automobile computer control systems susceptible to interference

By chance I saw an article in an issue of the
"Journal of Environmental Engineers" (published in England, date of
issue about 10 months ago, I believe) about the sensitivity of
a microprocessor-controlled automobile control system to external
electromagnetic radiation.  As I recall, a CB transmitter near the car
could, at the right frequency, make the engine slow down or speed up.
Perhaps this article would interest some of your contributors.

Bennett Smith					
IRCAM
31, Rue Saint Merri
75004 Paris, France		{seismo,philabs,decvax}!mcvax!ircam

------------------------------

Date: 15 Oct 1985 23:42:01 PDT
Subject: The human element
From: Dave Dyer       <DDYER@USC-ISIB.ARPA>
To: risks@SRI-CSL.ARPA


 The human element really is where the action is, and it is
a completely two-edged sword;  Human actions which have the
power to "fix" something almost inherently also give the power
to "break" things equally severely.  Conversely, weighty
check and balance systems intended to prevent abuse end up
preserving the status quo, however good or bad that may be.

 The "police computer horror story" I'm most familiar
with is illustrative.  This is a well documented case
I've been reading about in ACLU publications.   

 It seems some poor soul had his wallet stolen, and some criminal
adopted his identity and later was involved in a robbery/murder.  
Through some circumstances peculiar to the case, the stolen
identity, but not the culprit, were known to the LAPD.  The
detectives working on the case put the stolen identity into
a national police computer.   Our hero was stopped for
a routine traffic citation, the computer coughed his 
name up, and he ended up on ice for a few days as a
murder suspect.  

  So far, this is pretty harmless and understandable.  Eventually
the guy's identity and and non-involvement were establishd and
he was turned loose.   Then it happened again.  And Again.
The guy began carrying a letter from the local chief of police,
saying he wasn't the guy the computer said was wanted, but
that didn't cut it when he traveled.

  The problem was that the LAPD detectives who put in the
original "want" refused to remove it.   Eventually the
guy (and the ACLU) got the courts to mandate expunging 
the computer.   I think the detectives involved and the 
LAPD are being sued.   Quite rightly.

  The point is, it is <<hard>> to design a system that
can do its intended job, permit discovery and correction
of errors, and resist unautherized or inappropriate use.
I can't imagine a system that can do all three.

------------------------------

Date: 24 Oct 1985 11:17-PDT
Subject: Electronic Surveillance.
From: the tty of Geoffrey S. Goodfellow <Geoff@SRI-CSL.ARPA>
 
	[forwarded to RISKS by  Bill Keefe <keefe%milrat.DEC@decwrl.DEC.COM>
	 from TELECOM Digest Volume 5, Issue 55, October 24, 1985]


Americans' Privacy Exposed by New Technology, Congress Told
 
By LEE BYRD - Associated Press Writer
 
    WASHINGTON (AP) - The explosion in communications technology has so
outpaced privacy laws that Americans have little or no protection
against a plethora of new ways for government or private adversaries
to pry into their lives, a congressional agency reported today.
    The non-partisan Office of Technology Assessment found that 35 out
of 142 domestic federal agencies use or plan to use various
electronic surveillance methods, including modern devices not
governed by a landmark 1968 law that circumscribed the use of
wiretaps and bugs - concealed microphones.
    The agency said 36 agencies, not counting those in foreign
intelligence, already use a total of 85 computerized record systems
for investigative or intelligence purposes, and maintain 288 million
files on 114 million people. The report raised the ''technically
feasible'' specter of these being linked into a single data base
network that could track untold numbers of citizens without due
cause.
    The report, requested by House and Senate committees, noted that
many new and uncontrolled methods of surveillance are made possible
by the very technologies of which more and more Americans are
availing themselves - electronic mail, computer conferencing,
cellular and cordless telephones, beepers and electronic pagers.
Intercepting such devices is easy, and ''the law has not kept pace,''
the agency said.
    But other devices, such as miniature television cameras and pen
registers - which monitor the numbers called on a given telephone
line - have enabled new ways to spy on people even if their own
communications habits are more old-fashioned, the agency noted.
    Rep. Robert W. Kastenmeier, D-Wis., chairman of the House Judiciary
subcommittee on courts and civil liberties, said the study ''shows
how the law in this area has broken down; it is up to Congress to fix
it. If we fail to act, the personal and business communications of
Americans will not have the privacy protection they deserve.''
    Sen. Charles McC. Mathias, R-Md., said the report ''documents how
new and more intrusive forms of snooping have followed in the wake of
the exciting advances in communications technology,'' and agreed
Congress must ''bring federal privacy laws up to date.'
    Rep. Don Edwards, D-Calif., chairman of the House Judiciary
subcommittee on civil and constitutional rights, said, ''While the
attorney general of the United States is claiming that the civil
liberties granted by the Constitution should be limited to the
'original intentions' of the framers, the technological possibilities
for government surveillance have exploded. The framers knew nothing
of closed-circuit television, wiretapping and computer data banks.''
    The report noted that the Fourth Amendment, which protects ''the
right of the people to be secure in their persons, houses, papers and
effects, against unreasonable searches and seizures,'' was written
''at a time when people conducted their affairs in a simple direct,
and personalized fashion.''
    Neither, said the report, has Title III of the Crime Control and
Safe Streets Act of 1968, which was designed to protect the privacy
of wire and oral communications, kept pace.
    ''At the time Congress passed this act,'' the report said,
''electronic surveillance was limited primarily to simple telephone
taps and concealed microphones. Since then, the basic communications
infrastructure in the United States has been in rapid technological
change.''
    The congressional agency said it could not estimate the extent of
electronic surveillance in the private sector, saying only ''it is
probable that many forms ... go undetected, and if detected, go
unreported.''
    But in its survey of the federal bureaucracy, OTA found 35 agencies,
mostly in the Justice, Treasury and Defense departments, used or
planned to use:
    -Closed circuit television, 29 agencies.
    -Night vision systems, 22.
    -Miniature transmitters, 21.
    -Electronic beepers and sensors, 15.
    -Telephone taps, recorders, and pen registers, 14.
    -Computer usage monitoring, 6.
    -Electronic mail monitoring, 6.
    -Cellular radio interception, 5.
    -Satellite interception, 4.
    As for the 85 computerized record systems that could be used for
surveillance purposes, none of the operators provided statistics
requested by the OTA on record completeness and accuracy.
    Under the 1968 law, wiretaps and bugs are prohibited without a court
order based on the affirmation of a high-ranking prosecutor that a
crime has occurred, that the target of the surveillance is involved,
and that other means of investigation would be ineffective.
    According to the Administrative Office of the U.S. Courts, federal
and state judges approved 801 out of 802 requests last year for
electronic surveillance, primarily wiretaps and hidden microphones,
at an average cost of $45,000.
    The agency said that while there is some promise in emerging
techniques for low-cost data encryption or other means to protect
communication systems from eavesdropping, ''there is no immediate
technological answer ... against electronic surveillance.''
    Foreign intelligence cases are governed by a separate law, so the
CIA, National Security Agency and Defense Intelligence Agency were
not included in the survey.
 
------------------------------

Date: 0  0 00:00:00 CDT	    [sic! ed.]
From: "UV2::MOOREL" <moorel@uv2.decnet>
Subject: Network Mailer Woes...
To: "risks" <risks@sri-csl>

	In a recent issue of one of the digests on the net, there was a problem
mentioned that seems to have a bearing on risks on computer systems, particu-
larly as use of computer networking increases in the future. At their request, 
the names have been changed to preserve anonymity.

	        Apparently for the past month, the people who reside on the
	bitnet have been unable to receive [this digest].  There is a long
	story behind this [...]. This story is also a *lot*
	of guesswork as to what happened.
	        At the beginning of September, [our system] changed its host 
	name to conform to the new domain name standards.  The site we
	were using to get to bitnet, did not recognize the new name and
	began rejecting all mail from [our system].  We did not become
	aware of this because we were not receiving any rejections or errors
	back from the gateway.  We were however, receiving mail *from* the 
	people on Bitnet who were asking what happened to their [...] digest.
	        We attempted to contact the people at [the gateway] but of 
	course the mail failed and they never did anything to correct the 
	problem which they, of course, were not aware of because nobody was
	complaining.  (Sounds like a Catch-22 situation if I ever heard
	one).
	        In any case, the problem has now been resolved.
	Unfortunately, these people have missed close to 50 digests.  There
	is no way I can tie up the mailer at [either the host or the gateway] 
	in order to remail the messages.  I also understand that there is no 
	way to use FTP from the bitnet. 

It appears that several incidents conspired together to cause the loss of this 
information, and although the loss was not critical, it will take much time and
effort for the people involved to catch up. If this had been a more critical 
information transfer, it might have been corrected faster; however, there is a 
good chance that it would have gone undetected anyway. It seems to be one more 
reason for information about any potential changes to be passed on to any sites
that may be affected and to be thoroughly checked on both ends to prevent this 
kind of thing from reoccurring.

	Lynne C. Moore (Moorel@Eglin-Vax.Arpa)

------------------------------

From: Karl.Kluge@G.CS.CMU.EDU
To: risks@sri-csl.arpa
Subject: Grade changing.

Some doubt has been expressed in this forum recently about people
changing grades and living happily ever after. I can't talk about
college systems, but in high school all the grades, attendence, etc.
for my high school and several other high schools were kept on a
mainframe in a centralized location. There is a system in Michigan
called MOIS for vocational data, and on the back of my MOIScript on
computer science was the transcript of a terminal session between the
attendance people and the computer. The login message gave the phone
number of the mainframe. The password was overprinted, but that is
useless -- you can learn to read through almost any overprinting.
Access to the grading, course scheduling, and attendance programs was by
providing a social security number which was echoed and not overprinted.
I thus found myself able to do really miraculous things. Being sixth in
my high-school class, I had no real motivation to use my knowledge
maliciously, and informed the administration. Some safeguards were added
(old social security numbers retired, certain social security numbers
only giving access to certain programs, restricting access to certain
programs to certain accounts), but I'm sure those could have been
circumvented with minimal effort -- I was a fairly good systems hack on
the operating system, and there were people who could hack rings around
me. The operating system was a simple three-tier ring system, and a lot
of the features put in for the sake of usability made it very insecure.

I send this to give first-hand evidence to those who have been posting
doubts that such things happen.

------------------------------

Date: Wed, 16 Oct 85 17:11:36 EDT
From: Andy←Mondore%RPI-MTS.Mailnet@MIT-MULTICS.ARPA
To: RISKS@sri-csl.arpa
Subject: Databases, grades, etc.

One of the systems programmers here at RPI made a point about
administrators and students sharing the same computer:  You
really aren't that much more secure when you have administrators
and students on separate computers if there is a network or dial-up
connection to the administrative computer than you are when
administrators and students are on the same computer.  If you have
network or dial-up connections to an administrative computer, it
isn't too difficult for a student with an autodial modem to write
a program on a PC that simply tries all possible phone numbers
for certain telephone exchanges and record the numbers that
produce a carrier tone.  Then the student could have another
program that tries passwords unless the system disconnects the
line after a certain number of unsuccessful tries.  The major
advantage of separating administrators and students is that
it might be more difficult for a student to access an administrative
file from a student account assuming the administrative file has the
proper protection set.

------------------------------

Date:     Mon, 21 Oct 85 0:44:24 EDT
From:     Sienkiew@louie.udel.EDU
To:       risks@louie.udel.EDU
Subject:  Database, Grades, etc...


You can create an extremely effective audit trail if you think about it for a
while.  That just makes the problem more "challenging".

Suppose you do your auditing one day and find that there were unauthorized
grade changes made for every student in the CS department and 1/2 of them
requested (incorrect) printed transcripts.  It seems unlikely that everybody
independently broke in on the same day.  So who do you penalize?  How many
transcripts have been mailed out already?

Suppose no grades were changed but there is a trojan horse waiting to raise
the grade only under certain circumstances?

My point is that the data doesn't have to stay changed forever.  And you can't
check the auditing records for every transaction, if you expect to gain by
using the computer.  You need to do as much as you can, of course, but beware
of the false sense of security...

			Mark.

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂09-Nov-85  1516	ACUFF@SUMEX-AIM.ARPA 	More Explorer info
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Nov 85  15:16:15 PST
Date: Sat 9 Nov 85 15:17:03-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: More Explorer info
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12157968758.74.ACUFF@SUMEX-AIM.ARPA>

   I've produced a short document "Explorers for KSL D-machine Users",
filed in Sumex:<LispM>Exp-for-D.*, and placed copies by all the pool
Explorers and X11.  It briefly comments on what a D-machine user
trying to use and Explorer should know to transfer expertise.

   Also, the first thing anyone should do when using an Explorer for
the first time is to run the function (NEW-USER) to set up a directory
and initializatoin file.

	-- Rich
-------

∂09-Nov-85  1822	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #29  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Nov 85  18:22:05 PST
Date:  9 Nov 85 1757-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #29
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Sunday, 10 Nov 1985       Volume 1 : Issue 29

Today's Topics:

                           Actors -= Smalltalk
                               AI on Crays

----------------------------------------------------------------------

Subject: Actors -= Smalltalk
Date: 08 Nov 85 07:29:15 EST (Fri)
From: schwamb@mitre.ARPA

This is in response to linus!brando@mitre-bedford (Thom)
in PARSYM V1 #28.

I think there IS a key distinction between object-oriented paradigms
(a la Smalltalk, Flavors, and the like) and the Actors paradigm.  The
Actors approach is designed for the *asynchronous* operation of a set
of computational objects.  The object-oriented approaches on the other
hand are inherently synchronous.  In fact, implementations such as
Flavors are little more than function application with a level of
indirection (indeed Lisp programmers implemented early versions of
this approach by just attaching a function onto some property of an symbol).

Of course, parallel activity can be simulated in the object-oriented
approach quite nicely, but that's what it is, a simulation.  The
Actor model (as Byron Davies pointed out) was designed from the start
with distributed processing in mind.  As far as I know, no object-oriented
system is implemented this way.  (However, efforts are under way to
look at this, including Lisa Sokol's work at MITRE (Washington).)

This, of course, leads to different uses for message passing in the
two approaches.  In object-oriented programming you have a value returned
which is usually another object.  You could look at this as the "message
reply".  The Actor system has no automatic reply mechanism, and so the
communication is more easily implemented on parallel architectures.

I think an examination of the MIT Memos on Actors, and the ACT series
of languages will make this distinction clearer.

...Karl

------------------------------

Date: Fri, 8 Nov 85 10:04:38 pst
From: eugene@AMES-NAS.ARPA (Eugene Miya)
Subject: AI on Crays

The contact at LANL [Los Alamos National Laboratory] for PSL is Wayne
Anderson.  The object of LISP is to run symbolic algebra packages, I
think you can guess which ones.  One of our people moved XLISP to the
Cray-2.  No one has done any work on vectorizing LISP: how would you
do it?  LISP on the Cray is still faster than any LISP out there from
what I understand Anderson and others have told me.

--eugene miya

[Lisp on a Cray is 5-10x a Symbolics 3600, as I remember, while
Fortran on a Cray is 50-100x a VAX 780 on benchmarks of interest to
number hackers.  Since a 3600 and a VAX are approximately the same
"power", give or take a factor of 2, we're missing an order of
magnitude somewhere.  Is the order of magnitude due to Cray hardware
being inappropriate for Lisp or to the lack of a good Lisp compiler
for the Cray, or some combination?  Would a functional subset of Lisp
be appropriate for vectorization?  -- BD]


------------------------------

End of PARSYM Digest
********************

∂10-Nov-85  1804	@SU-SCORE.ARPA:JMC@SU-AI.ARPA 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Nov 85  18:03:58 PST
Received: from SU-AI.ARPA by SU-SCORE.ARPA with TCP; Sun 10 Nov 85 18:00:44-PST
Date: 10 Nov 85  1757 PST
From: John McCarthy <JMC@SU-AI.ARPA>
To:   faculty@SU-SCORE.ARPA 

baby
Carolyn Talcott (McCarthy) and John McCarthy announce the birth
of Timothy Talcott McCarthy on November 9, 1985 at 0022.  He
weighed five pounds nine ounces.  All are well.

∂10-Nov-85  1916	LANSKY@SRI-AI.ARPA 	PLANLUNCH reminder -- 11AM  Moshe Vardi 
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 10 Nov 85  19:16:14 PST
Date: Sun 10 Nov 85 19:17:07-PST
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH reminder -- 11AM  Moshe Vardi
To: planlunch.dis: ;

		MODEL THEORY FOR KNOWLEDGE AND BELIEF

			   Moshe Vardi
			   IBM San Jose
				
	 	    11:00 AM, MONDAY, November 11
       SRI International, Building E, Room EJ228 (new conference room)


Recently, there has  been  a surge of interest in the modal logic of
knowledge and belief, which has applications in many area of computer
science.  The  standard semantics for modal logic is Kripke semantics.
In this semantics, possible worlds and the possibility relation are both
primitive notions. This has both technical and conceptual shortcomings.
From a technical point of view, the mathematics associated with Kripke
semantics is often quite complicated.  From a conceptual point of view, 
it is not clear how to use Kripke structures to model knowledge and belief,
where one wants a clearer understanding of the notions that are taken as
primitive in Kripke semantics.

We introduce modal structures as models for modal logic. We use the idea
of possible worlds, but by directly describing the internal semantics of
each possible world.  It is much easier to study the standard logical
questions, such as completeness, decidability, and compactness, using
modal structures.  Furthermore, modal structures offer a much more
intuitive approach to modelling knowledge and belief.

As an application, we present a semantic model for knowledge
with the following properties: 

   (1) Knowledge is necessarily correct 

   (2) agents are logically omniscient, i.e., they know all 
       the consequences of their knowledge

   (3) agents are positively introspective, i.e., they are aware of
       their knowledge, but not negatively introspective, 
       i.e., they may not be aware of their ignorance.

We argue that this is the appropriate model for implicit knowledge.
We investigate the properties of the model, and use it to formalize
notions such as "to know more" and "all that is known is".


-------

∂11-Nov-85  0140	RESTIVO@SU-SCORE.ARPA 	PROLOG Digest   V3 #44
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  01:40:16 PST
Date: Saturday, November 9, 1985 1:39PM
From: Chuck Restivo (The Moderator) <PROLOG-REQUEST@SU-SCORE.ARPA>
Reply-to: PROLOG@SU-SCORE.ARPA
US-Mail: P.O. Box 4584, Stanford CA  94305
Phone: (415) 326-5550
Subject: PROLOG Digest   V3 #44
To: PROLOG@SU-SCORE.ARPA


PROLOG Digest            Monday, 11 Nov 1985       Volume 3 : Issue 44

Today's Topics:
                      Query - Profiling & Crays,
                   LP Philosophy - Lisp in Prolog,
                Implementation - DCG + Left Recursion
----------------------------------------------------------------------

Date: Sun, 20-Oct-85 20:52:26 PDT
From: (Jonathan Pincus) ernie!pincus@UCB-Vax
Subject: Profiling Programs?

Has anybody out there ever tried to profile a Prolog program?
Specifically, how do I find out how much time various procedures
require?  Obviously, I want to do this in order to speed up the
program in the important places (for that matter, does the old
chestnut about 10% of the code requiring 90% of the execution
time hold for Prolog?).

My specific environment is C-Prolog (version 1.2) on a VAX.  One
thought that has occurred to me is that timing information needs
to be recorded at the same time that "spy"ing information gets
printed out, but I'm not sure how to do that.  Any comments/advice
/solutions will be gratefully accepted, and if anything exciting
turns up I'll post a summary. Please don't tell me "Oh, that's just
an efficiency question, you shouldn't worry about that!"

-- Jon

------------------------------

Date: Wed, 23 Oct 85 10:20:44 edt
From: Carl Landwehr <landwehr@nrl-css.ARPA>
Subject: Prolog for Cray?

I am new to this list, so bear with me if this is an old query.
We have a Cray X-MP that might be available for us to run Prolog,
if we could find a Prolog for it.  Does anyone out there have one,
or know of plans to develop one?

-- Carl Landwehr

------------------------------

Date: 23 Oct 85 17:48:09 PDT (Wed)
From: Sanjai Narain <narain@rand-unix.ARPA>

We are about to begin a project which intends to use both
Franzlisp and Prolog IN THE SAME ENVIRONMENT. Are there any
suitable Prologs with efficiency >= 1KLips? We use SUNs and
Vaxes running UNIX. Any leads would be greatly appreciated.

-- Sanjai Narain

------------------------------

Date: Thu, 24 Oct 85 22:14 EDT
From: Hewitt@MIT-MC.ARPA
Subject: Lisp in Prolog?

I would like to reply to the message from William Clocksin which said

 ... it cannot have escaped the attention of most people that writing
 a compiler (in Prolog) to COMPILE Lisp into some target (say LAP)
 would be as easy as pie.  For those of us who have written
 compilers in both Lisp and Prolog, I daresay it would be a lot
 easier in Prolog.  I might further suggest that the compiler
 test is more useful than the interpreter test, and that the
 interpreter test has no especial theoretical advantage, since
 the compiler can be seen as a meaning-preserving transformation.
 (Ahem. Which in fact it isn't for Lisp if you're not careful
 with bindings).

I argue that the above compiler is not a very good test of Prolog
because the code produced by the proposed Prolog compiler for
Common Lisp WILL NOT RUN on a standalone Prolog system.  Thus the
proposed compiler does not address a fundamental problem which is
the LACK OF EXPRESSIVE CAPABILITY in the Prolog language:  there
is large and growing amount of software written in Common Lisp which
will NEVER execute efficiently on standalone Prolog systems.  On the
 other hand Prolog programs will ALREADY execute efficiently on Lisp
systems.  Thus the compiler which Clocksin proposes does not address
the fundamental problems of Prolog.

------------------------------

Date: Thu, 24 Oct 85 20:32:10 PDT
From: Adolfo Di-Mare <dimare@LOCUS.UCLA.EDU>
Subject: DCG + Left Recursion :- Succeed!

After looking into the dragon book, I found the trick I needed
to do avoid left recursion but still have left associative
operators.

The problem is how to evaluate arithmetic expressions with left
associative operators. The naive way is to say:

expr(Z) --> expr(Y), "-", term(X), {Z is X-Y}.
expr(Z) --> term(Z).
term(X) --> [C], {48=<C, C=<57, X is C-48}.

If we call:
        | ?- expr(Z,"2-3-5-1",[]).
this program loops.

The trick is very simple: inherit from the left (above) the
left operand, and use it right away when building the expression
tree. The modified grammar:

expr(Tree) --> term(Left), rest(Left,Tree).
rest(Left,Tree) --> "-",term(Right),{Temp is Left-Right},rest
                                                        (Temp,Tree).
rest(Tree,Tree) --> [].
term(X) --> [C], {48=<C, C=<57, X is C-48}.

The call
        | ?- expr(Z,"2-3-5-1",[]).

now gives the desired result: Z = -7.

Thanks to all those that helped me.

-- Adolfo

------------------------------

End of PROLOG Digest
********************

∂11-Nov-85  0945	NILSSON@SU-SCORE.ARPA 	New Keys    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  09:44:26 PST
Date: Mon 11 Nov 85 09:40:01-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: New Keys
To: csd@SU-SCORE.ARPA
Message-ID: <12158431689.15.NILSSON@SU-SCORE.ARPA>

For a number of reasons we decided that it was time to "re-key"
the CSD rooms in MJH.  I've been told that there are probably a
number of unauthorized "master keys" floating around.  These,
obviously, present security problems and place the department
at risk in case things of value disappear.

I realize quite well that there are no technical difficulties whatsoever
preventing the reappearance of unauthorized master keys.  One purpose of
this msg is to let everyone know that as of today, there are no
unauthorized master keys to the new locks--because installation of the
new locks won't be completed until tomorrow.  

Another purpose of the msg is to inform people that knowingly "becoming
involved with" (using, copying, creating, possessing, etc.) an
unauthorized master key will be regarded by the department
administration as a serious breach of trust.  (I take it that all of us,
perhaps grudgingly, acknowledge the necessity of a system of locks and
keys.  If we are serious about that, we must also acknowledge that
unauthorized master keys work to defeat that system.)  I will regard
evidence of being involved with unauthorized master keys as a matter to
be referred to the President's office as a possible violation of the
Fundamental Standard.

I hope the rekeying today and tomorrow goes smoothly and causes no
one any undue hardships.  Thanks,  -Nils
-------

∂11-Nov-85  1043	INGRID@SU-CSLI.ARPA 	Symposia on AI
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  10:42:31 PST
Date: Mon 11 Nov 85 10:36:42-PST
From: Ingrid Deiwiks <INGRID@SU-CSLI.ARPA>
Subject: Symposia on AI
To: SU-Bboards@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA


	 	          ANNOUNCEMENT


             Soto Symposia on Artificial Intelligence
             ----------------------------------------

A series of three symposia on AI will be held in the Soto Lounge,
Wilbur Hall, the first three Mondays in November, at 7 p.m.  These
symposia are intended to inform Stanford students and others who are
interested about the nature of AI, its prospects, and the intellectual
and social issues to which it gives rise.

The first symposium took place on Monday, November 4.  The other two
are scheduled as follows:

Monday, November 11, 7 p.m.

	"Can we make computers that think and talk?"

	Terry Winograd    Professor of Computer Science, Stanford

	Brian Smith       Computer Scientist, Xerox PARC
		          Consulting Professor of Philosophy,
	                  Stanford			  

	Nils Nilsson      Professor of Computer Science and
		          Chair of Department of Computer Science, 
			  Stanford


Monday, November 18, 7 p.m.

	"Star Wars and Computation"

	John McCarthy     Professor of Computer Science, Stanford

	David Redell      DEC Research Laboratory

	and possibly others

Each symposium will begin with short talks by the speakers, followed
by plenty of time for vigorous interchanges among the speakers and
with members of the audience.

Soto is the first dormitory in the Wilbur complex that one comes to
when walking out Escondido from the Undergraduate Library.

	
----------------------------------------------------------------------
<--UGLI		              Escondido  Rd.
----------------------------------------------------------------------
                                       
xxxxxxxxxxxxxx                              X XXX     XXX X 
		                   Soto --> X   Wilbur    X		
Stern Hall                                  X             X

xxxxxxxxxxxxxx                              X    Hall     X
                                            X             X
                                            X XXX     XXX X




-------

∂11-Nov-85  1040	NILSSON@SU-SCORE.ARPA 	More "Keys" 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  10:39:55 PST
Date: Mon 11 Nov 85 10:37:12-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: More "Keys"
To: csd@SU-SCORE.ARPA
Message-ID: <12158442100.15.NILSSON@SU-SCORE.ARPA>

Some people have pointed out to me that "unauthorized keys" in
the possession of one or two graduate students serves a very
useful purpose.  If that's so, why don't we give out one or
two such keys only we should call them "authorized" instead.
I'll talk more with LaDonna and Betty about this.  In the meantime,
the student representatives might be thinking about this idea.
-Nils
-------

∂11-Nov-85  1201	INGRID@SU-CSLI.ARPA 	HOUSEMATE WANTED   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  12:00:59 PST
Date: Mon 11 Nov 85 11:57:02-PST
From: Ingrid Deiwiks <INGRID@SU-CSLI.ARPA>
Subject: HOUSEMATE WANTED
To: SU-bboard@SU-CSLI.ARPA, folks@SU-CSLI.ARPA


Housemate wanted, from now until mid-June 1986.  Share house with
three professional journalists.  Beautiful house in a beautiful
location in Palo Alto, biking distance from Stanford.  Fully furnished
room, private bath, in fully furnished home; washer/dryer, dishwasher,
fireplace, backyard.  $485 per month.  Call 325-6737 evenings.

-------

∂11-Nov-85  1402	@SU-CSLI.ARPA:VAL@SU-AI.ARPA 	Non-Monotonic Reasoning Seminar    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  14:02:17 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 11 Nov 85 14:00:21-PST
Date: 11 Nov 85  1346 PST
From: Vladimir Lifschitz <VAL@SU-AI.ARPA>
Subject: Non-Monotonic Reasoning Seminar   
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    


		SOME RESULTS ON AUTOEPISTEMIC LOGIC	

			   Wiktor Marek
		      University of Kentucky
				
	 	  2:00 PM, WEDNESDAY, November 20
 			     MJH 252


	We discuss some properties of so-called stable theories in
autoepistemic logic (cf Moore, AIJ 25 (1985)), that is, sets of beliefs of
a fully rational agent. We show an operator constructing these theories out
of their objective parts and investigate the complexity of the construction.
We attempt to extend Moore's approach to the case of predicate logic.
Finally, we discuss the notion of inessential modal extension of a first
order theory.

∂11-Nov-85  1455	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #30  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  14:55:06 PST
Date: 11 Nov 85 1421-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #30
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 11 Nov 1985       Volume 1 : Issue 30

Today's Topics:

             XLISP on a Cray is Interpreted (2 messages)
                          Actors and Objects
               Vectorized Lisp and Multiprocessor Lisp
   Seminar: BFCP and GHC - Alternatives to Concurrent Prolog (MIT)

----------------------------------------------------------------------

Date: Sun, 10 Nov 85 11:44:17 EST
From: "William J. Bogstad" <bogstad@hopkins-eecs-bravo.ARPA>
Subject: Speed improvement of a Cray

	It should be noted that XLISP, the version of LISP ported to
the Cray, is strictly an interpreter.  Given this fact, the order of
magnitude loss which appears to occur may not be there.

				Bill Bogstad
				bogstad@hopkins.arpa

------------------------------

Date: Sun, 10 Nov 85 17:13:18 EST
From: Frank Ritter <ritter@BBN-LABS-B.ARPA>
Subject: XLISP on a Cray

XLISP runs only interpreted; there is no compiler for it.  You can
easily see about an order of magnitude in speed between compiled and
interpreted in other lisps.  The 3600 makes it very easy to always
only run the compiled version.  So that's where the X10 went: compiled
(less filling) vs. interpreted (debugs great).

Frank

------------------------------

Date:     10 Nov 85 (Sun) 13:26:02 EST
From:     Jim Hendler <jah%brown.csnet@CSNET-RELAY.ARPA>
Subject:  PARSYM Digest   V1 #29

Recent mailings re. Smalltalk and etc. lead me to believe that the
readers of this net share a certain confusion with most of the rest of
the world.  The problem is that there has been an amazing obfuscation
that has resulted from the linking of the terms "object-oriented" and
"message-passing."  Flavors, Smalltalk, and LOOPS are definitely
object-oriented languages, however their control structure is still
the basic stack structure of FUNCALL.  I believe that there is a clear
distinction between this and the "message passing" work that seems a
better model for parallelism.  Consider the following example:

   Byron asks you what time it is, you know that I have a watch
   (you don't).  When you ask me for the time, why should I reply
   to you (causing you to be in a wait state).  You'd rather send
   out a message to me saying "What time is it (oh, by the way,
   give the answer to Byron, not to me)?"

I haven't used actors in nearly 5 years, but I recall it had something
resembling true message-passing.  Brian Phillips and I, then at Texas
Instruments, hacked up a system in Interlisp that did actual
message-passing of this type [the implementation used the spaghetti
stacks of Interlisp.  It was quite slow and one could only leave a
small number of unanswered messages around before things broke.]  We
found some interesting properties to the algorithms we were able to
design using the new control (this was briefly described in an old
technical report called "A flexible control structure for the
conceptual analysis of natural language using message-passing" -- the
Interlisp code for the message-passing system is in an appendix).

This research was shelved while I pursued my doctoral research.  Now
that that is nearly done I will be starting some work at the
University of Maryland (as of Jan 1) to continue some present work on
object-oriented computing and to work some new ideas on
message-passing into a flavors-like object-oriented system.  I'm open
for advice, comments, and etc.

  -Jim Hendler
   Brown U.
 [best e-mail address is Hendler@tove]

------------------------------

Date: Sun, 10 Nov 85 16:24:09 cst
From: harrison@uicsrd.CSRD.UIUC.EDU (Luddy Harrison)
Subject: PARSYM Digest V1 #29

[Eugene Miya asked how Lisp could be vectorized.  Byron Davies asked
whether a functional subset of Lisp would be appropriate for
vectorization. -- BD]

	Some time ago, I mentioned that we are working on the
compilation of lisp for multiprocessors.  In fact, I mentioned that I
would soon have a report available for the asking concerning the
details of the compilation process.  Let me first apologize to those
who have waited for this paper: what started out to be a brief
overview of the compiler has become several hundred pages of detailed
description, with lots of figures and so on.  I have kept a list of
all who have requested this document, and will send it out as soon as
it is complete.

	Vectorization refers ordinarily to the compilation of a
program (usually in FORTRAN) for a vector machine, such as the Cray,
or perhaps an array processor.  There are several reasons why
"vectorizing" lisp is difficult:

1) Vector machines perform best when executing a single operation over
many elements of a vector.  Usually this is a simple operation such as
addition or multiplication; at any rate, the operation itself can
usually not be defined in software.  That is, only those operations
directly supported by the hardware may be performed upon the vector
register.  This means that even if we could overcome the problem of
gathering the elements of a list into a vector register, the types of
operations that we could perform upon them would not allow the
generality of say, a mapcar function.  And even if we could somehow
make mapcar perform well, more complex functions (most of real lisp
programs) would not be easily pressed into such a tight mold.

2) It is well known that frequent branching degrades the performance
of a pipelined architecture.  The way that this is ordinarily overcome
is by the translation of conditional expressions into mode vectors,
which serve as additional vector inputs to the pipe.  For instance,
the statement

if p(x[i]) then a[i] := b[i] + c[i]

might be compiled as

dovector q[i] := p(x[i])
dovector a[i] := b[i] + c[i] where q[i].

The vector q is called a mode vector.  An analagous translation is
possible in the case of lisp programs, but I don't think that anyone
(other than us) has described the way it can be done.  (But as I
pointed out, unless we want to perform floating point operations on
lists, which seems rather un-lispish to me, this is not likely to get
us very far.)  Without such a translation, we would be limited to very
simple operations indeed, upon the lists that we somehow got into the
vector registers.  For instance, we might want to perform floating
point divide (!) on every atomic element of a list; to perform this
test would require translating the test (atom x) into a mode vector.

	The good news is that none of these problems exists for a
multiprocessors like CEDAR or RP3, and we think that compiling lisp
for these architectures will be quite feasible.  We are not
restricting ourselves to a functional subset; it seems to be
unnecessary.

-Luddy Harrison

Center for Supercomputer Research and Development
University of Illinois
302D Talbot Laboratory
104 South Wright Street
Urbana IL 61801-2932

(217) 333 4168

------------------------------

Date: Sun, 10 Nov 85 00:56:01 EST
From: "Steven A. Swernofsky" <SASW@MIT-MC.ARPA>
Subject: Forwarded seminar announcement

To:   cog-sci-calendar
Thursday  7, November  2: 15pm  Room: NE43- 7th floor playroom

         BFCP and GHC - Alternatives to Concurrent Prolog

                 Jacob Levy
                 Department of Applied Mathematics
                 Weizmann Institute of Science

This talk will discuss some of the alternatives to Concurrent Prolog
recently proposed.  Each of these languages is designed to cover a
large subset of Concurrent prolog, but to be much easier to implement.
Flat Concurrent prolog (FCP) and Guarded Horn Clauses (GHC) will be
described in detail.

FCP, which has only And-parallelism, was developed at the Weizmann
Institute as a viable subset of Concurrent Prolog.  Its current
implementation, in terms of a Warren Abstract Machine, will be
described.

The GHC language, designed by K. Ueda of ICOT, Japan, has
OR-parallelism as well as And-parallelism, but instead has more
limited synchronization primitives than Concureent Prolog.  The second
part of this talk will briefly describe my implementation of GHC.

After the talk, a demo of FCP and Logix, its programming environment,
will be given.

HOSTS: Professors Gerald Jay Sussman and Henryk Jan Komorowski
                                            (Harvard)

------------------------------

End of PARSYM Digest
********************

∂11-Nov-85  1539	CAROL@SU-CSLI.ARPA 	Demo postponed 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  15:38:46 PST
Date: Mon 11 Nov 85 15:33:39-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Demo postponed
To: Folks@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA


*********************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DL
						
				Nov 11		    												
The  demo of the Gloss package, originally scheduled for Tues at 
4 p.m., has been postponed until after Thanksgiving. I am sorry
for any  inconvenience this may cause.  

On Nov. 19 Richard Burton will demo the new Sketch.  Further 
information forthcoming. 

                             -Carol     (321-2136,  Ventura 28)

ION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION*
*********************************************************************          
-------

∂11-Nov-85  1541	TAJNAI@SU-SCORE.ARPA 	IBM/Forum Party 11/14/85    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  15:41:09 PST
Date: Mon 11 Nov 85 15:37:37-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: IBM/Forum Party 11/14/85
To: su-bboards@SU-SCORE.ARPA
cc: csd@SU-SCORE.ARPA, csl-everyone@SU-SIERRA.ARPA
Message-ID: <12158496788.36.TAJNAI@SU-SCORE.ARPA>

                                 IBM Research

                               Cordially invites

                the Faculty, Staff and Students of CSD and CSL

                               to attend a party

                          Thursday, November 14, 1985

                               4:00 to 6:00 p.m.

                           Faculty Club, Gold Lounge


                        Sponsored by the Computer Forum

  p.s.  Lots of good food!!
-------

∂11-Nov-85  1644	PATASHNIK@SU-SUSHI.ARPA 	nonAFLB talk   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  16:43:52 PST
Date: Mon 11 Nov 85 16:44:18-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: nonAFLB talk
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12158508929.10.PATASHNIK@SU-SUSHI.ARPA>

There's a talk this Wednesday that will interest some of us:

-------
Speaker : Margaret Wright

	On projected Newton Barrier Methods for LP and
	an equivalence to Karmarkar's projective method

Wed, Nov 13, at 4:30 in Terman 453
-------

I know nothing about this besides what's in this message.

	--Oren
-------

∂11-Nov-85  2307	DAVIES@SUMEX-AIM.ARPA 	No meeting this week  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 11 Nov 85  23:07:33 PST
Date: Mon, 11 Nov 1985  23:09 PST
Message-ID: <DAVIES.12158579027.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No meeting this week
cc:   Davies@SUMEX-AIM.ARPA

There will be no AAP meeting this Wednesday, November 13.

If you're looking for something to do instead, the TI Satellite
Seminar on Knowledge-Based Systems will be going on all day Wednesday.
The seminar starts at 8:15 am in Law School 290.

	-- Bron

∂12-Nov-85  0812	RICHARDSON@SU-SCORE.ARPA 	CSD Lunch
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  08:12:46 PST
Date: Tue 12 Nov 85 08:13:03-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
cc: library@SU-SCORE.ARPA, maslin@SUMEX-AIM.ARPA
Message-ID: <12158678001.13.RICHARDSON@SU-SCORE.ARPA>


The Tuesday CSD Lunch Series will continue today with Elliot Levinthal on
the topic of CIMA.

-------

∂12-Nov-85  0835	EMMA@SU-CSLI.ARPA 	TINLunch   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  08:34:59 PST
Date: Tue 12 Nov 85 08:33:57-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: TINLunch
To: friends@SU-CSLI.ARPA
Tel:  497-3479


For complicated reasons, Jon Barwise's TINLunch has been moved from
December to this Thursday.  His abstract follows:


                     ``Machines and the Mental''
                             Fred Dretske

The paper argues that current computers do not exhibit anything that
deserves to be called rational cognitive activity.  Dretske even
claims that they can't add!  He then goes on to discuss how one might
build something that deserves to be called a rational machine.

This short, well-written paper is Dretske's Presidential Address to
the APA.  Read it can come prepared for a lively session.
-------

∂12-Nov-85  1117	BERGMAN@SU-SCORE.ARPA 	RA support  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  11:17:35 PST
Date: Tue 12 Nov 85 09:32:34-PST
From: Sharon Bergman <BERGMAN@SU-SCORE.ARPA>
Subject: RA support
To: faculty@SU-SCORE.ARPA
cc: bergman@SU-SCORE.ARPA
Message-ID: <12158692479.30.BERGMAN@SU-SCORE.ARPA>

If you need to make any changes in your support of research assistants
for Winter quarter (i.e. adding new students, stopping any appoint-
ments, changing account numbers, etc.), please let me know as soon as
possible so that I can send the paperwork through in a timely manner.
If you would like to see a list of the students my records show you are
currently supporting, please let me know and I'll send you the list.
Thanks.			Sharon Bergman
-------

∂12-Nov-85  1118	NILSSON@SU-SCORE.ARPA 	Student Names    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  11:18:39 PST
Date: Tue 12 Nov 85 09:34:53-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Student Names
To: faculty@SU-SCORE.ARPA
Message-ID: <12158692900.31.NILSSON@SU-SCORE.ARPA>

After receiving some faculty and student responses on the matter
of sending out lists of names of graduating students (and seeing
no evidence of a faculty-adopted policy to the contrary), I suggest
the following policy to take effect immediately (unless there
is a great outpouring of sentiment against it):

Victoria Cheadle (for the Ph.D. students) and Jutta McCormick (for the
M.S. students) will maintain lists of names of CSD students graduating
within the next few months.  They will do this by inviting students to
have their names on these lists. But no name will be included on the
lists unless students specifically ask to have them included.  (Once
someone asks to have his/her name included, it will stay on the list
until that person asks to have it deleted--or until the student leaves
us.)  Thus, the system will have a certain fail-safe feature.  You have
to respond positively to Victoria's/Jutta's request in order to have
your name on the list.

These lists will be sent free of charge to anyone who asks for
them.  I see no conflict here with what the Computer Forum
provides.  

I hope it won't occur to people to ask if several different kinds
of lists could be maintained.  (One for inquiring universities, one
for inquiring venture capital firms, etc.)  That could become
an administrative nightmare.

-Nils
-------

∂12-Nov-85  1119	NILSSON@SU-SCORE.ARPA 	SOE Committees   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  11:19:51 PST
Date: Tue 12 Nov 85 09:49:33-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: SOE Committees
To: faculty@SU-SCORE.ARPA
Message-ID: <12158695569.31.NILSSON@SU-SCORE.ARPA>

The School of Engineering has a number of standing committees.
They have asked me for suggestions about CSD faculty who would
be appropriate to serve on these committees.  These are
the committees and current membership:

Engineering Minority Program:
Ashley, Bates, Eisenhardt, Hillier, Kiremidjian, Bradford, Shevell,
Wiggins, Patterson, Leckie (chair).

Committee on Engineering in Biology and Medicine
Chang, Angell, Zajac, Robertson, Hesselink, Steele (chair)

School of Engr. Library Committee
Van Dyke, Mason, Cottle, Cantwell, Fondahl, Fraser-Smith, Freyberg, Gray,
Herrman, Howard (chair), Moffat, Plummer, Sullivan, Roberts

Committee on Computers
Homsy, Huggins, Hughes, Powell, Jucker, Gill, Kychakoff, Koseff,
Ferzinger (Chair)

SITN Advisory Committee
Kruger, White, Hausman, Knuth, Springer, Allen (Chair)

I don't know very much about what exactly these committees do.
If anyone would like to have his name suggested, please let
me know.  -Nils
-------

∂12-Nov-85  1236	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  12:35:57 PST
Date: Tue 12 Nov 85 08:58:13-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12158686224.13.RICHARDSON@SU-SCORE.ARPA>


My apologies....that should read SIMA.
-------

∂12-Nov-85  1236	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  12:36:22 PST
Date: Tue 12 Nov 85 08:59:14-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12158686411.27.LIBRARY@SU-SCORE.ARPA>

Computer-Aided Database Design The DATAID Project. ed. by A. Albano, 
V. de Antonellis and A. di Leva. QA76.9.D3C66 1985

PostScript Language Reference Manual. Adobe Systems Incorporated. 
QA76.73.P67P67 1985.

Foundations of Programming. by Jacques Arsac. Translated by Fraser Duncan.
APIC Studies in Datat Processing no. 23. QA76.6.A75813 1985

Fundamentals of Computation Theory. FCT '85. Cottbus, GDR, Sept. 1985.
ed. by Lothar Budach. Lecture Notes In Computer Science 199.
QA267.I57 1985.

Advances In Cryptology. Proceedings of Cryto 84.  ed. by G. R. Blakley and
David Chaum. Lecture Notes In Computer Science. QA76.9.A25C79.

Pattern Recognition Human and Mechanical. by Satosi Watanabe. Q327.W365
1985. c.2

The Shape Of Space How To Visualize Surfaces and Three-Dimensional
Manifolds.  Pure and Appplied Mathematics Series.  by Jeffrey Weeks.
QA612.2.W44 1985

Lancozos Algorithms for Large Symmetric Eigenvalue Computations. 
Volume 1  Theory Volume 2 Programs.  Progress in Scientific
Computing.  by Jane K. Cullum and Ralph A. Willougby
QA193.C84 1985 V. 1 and V. 2

Estimating Eigenvalues With a posteriori/a priori Inequalitities.
Research Notes In Mathematics.  QA374K95 1985.

Combinatorial Algorithms On Words.  NATO ASI Series. ed by Alberto
Apostolico and Zvi Galil.  QA164.N35 1984.

Logics and Models of Concurrent Systems. NATO ASI Series. ed by
Krzysztof R. Apt. QA76.5N16 1984.

Computational Mathematical Programming. NATO ASI Series.  ed by Klaus
Schittkowski.  QA402.5N365 1984.

Control Flow And Data Flow: Convepts Of Distributed Programming.
International Summer School Directed by F. L. Bauer, E. W. Dijkstra,
and C. A. R. Hoare. ed by Mangred Broy. NATO ASI Series.
QA76.9.D5N375 1984

Health Information Systems A Bibliography. by Donald Ziegenfuss and
James Ziegenfuss, Jr.  Z6675.E4Z53 1984.

H. Llull
-------

∂12-Nov-85  1237	RICHARDSON@SU-SCORE.ARPA 	CSD 500 Colloquium Series    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  12:37:30 PST
Date: Tue 12 Nov 85 09:03:17-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD 500 Colloquium Series
To: faculty@SU-SCORE.ARPA
Message-ID: <12158687147.13.RICHARDSON@SU-SCORE.ARPA>


Would someone like to volunteer to introduce Pradeep Sindhu today at the
CSD 500 talk which he is giving today in Skilling at 4:15? Perhaps someone
who is already planning on attending?
-------

∂12-Nov-85  1239	NILSSON@SU-SCORE.ARPA 	Master Keys 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  12:39:47 PST
Date: Tue 12 Nov 85 09:15:01-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Master Keys
To: csd@SU-SCORE.ARPA
Message-ID: <12158689283.31.NILSSON@SU-SCORE.ARPA>

People have persuaded me that it is very important for a 24-hour-a-day
operation like the CSD in MJH not to be hampered by the 8-hour-a-day
policy of locking things up after hours.  Things are greatly unhampered
by always having a few trusties around with a master key so that folks
can get into the rooms they would ordinarily be able to get into
during the day.  A fine idea!  My earlier msgs should be interpreted
as only suggesting that we might not necessarily want the trusties
to be those persons who are skilled at locksmithery.  So, people
who want to volunteer to be trusties should send a msg to LaDonna,
eppley@score, and we'll select a few.  -Nils
-------

∂12-Nov-85  1254	ACUFF@SUMEX-AIM.ARPA 	Explorer 1.0.2: anyone need it?  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  12:54:16 PST
Date: Tue 12 Nov 85 12:38:39-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorer 1.0.2: anyone need it?
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12158726352.18.ACUFF@SUMEX-AIM.ARPA>

   If anyone wants to use Explorer release 1.0.2, please let me know,
and I'll try to set it up.

	-- Rich
-------

∂12-Nov-85  1346	WASOW@SU-CSLI.ARPA 	consulting offer    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  13:46:35 PST
Date: Tue 12 Nov 85 13:37:48-PST
From: Tom Wasow <WASOW@SU-CSLI.ARPA>
Subject: consulting offer
To: Folks@SU-CSLI.ARPA

I have been contacted by a representative of a company called Qualogy,
which is looking for someone to evaluate some natural language translation
software that is being considered for product development.  He seemed 
to think that it would be a job of only a few hours, and indicated that
they would pay competitive consulting fees (though he didn't mention
a figure).  Anyone interested in pursuing this should contact me.

Tom Wasow
-------

∂12-Nov-85  1448	ullman@su-aimvax.arpa 	paper received   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  14:48:03 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 12 Nov 85 14:43:15 pst
Date: Tue, 12 Nov 85 14:43:15 pst
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

"Equivalence and Optimization of Relational Transactions"
by Abiteboul and Vianu.

This doesn't seem to have much to do with logic and databases,
but rather it is a study of optimization for insert/delete
operations.  Any interest?

∂12-Nov-85  1519	ullman@su-aimvax.arpa 	tomorrow's meeting    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  15:18:13 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 12 Nov 85 15:08:34 pst
Date: Tue, 12 Nov 85 15:08:34 pst
From: Jeff Ullman <ullman@diablo>
Subject: tomorrow's meeting
To: nail@diablo

I have (a) no agenda and (b) a 10AM meeting with the dean
that is likely to run over the normal NAIL time.
Thus, I suggest that we either cancel the meeting or
agree that those of us who are free will gather at 11AM
and grump about something.

∂12-Nov-85  1522	MODICA@SU-SCORE.ARPA 	info for TV students   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  15:22:44 PST
Date: Tue 12 Nov 85 15:20:39-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: info for TV students
To: faculty@SU-SCORE.ARPA, instructors@SU-SCORE.ARPA
Message-ID: <12158755846.21.MODICA@SU-SCORE.ARPA>

PLEASE SEND ME THE FOLLOWING INFORMATION AS SOON AS POSSIBLE :

1)IF YOU ARE TEACHING A COURSE WINTER QUARTER, ARE THE STUDENTS GOING TO NEED
  TO HAVE ACCESS TO ONE OF THE COMPUTING SYSTEMS (SUCH AS LOTS)?.....

2)..AND IF SO WILL TV STUDENTS HAVE THE OPTION OF USING A DIFFERENT COMPUTER?

IN YOU REPLY PLEASE INDICATE WHAT COURSE YOU WILL BE TEACHING. THE REASON I
NEED THIS INFORMATION IS THAT THOSE ARE VERY COMMON QUESTIONS FROM TV STUDENTS,
(ALONG WITH WHAT BOOKS ARE GOING TO BE USED, BUT I CAN GET THAT INFO FROM
KATHY BERG), AND I WOULD LIKE TO BE ABLE TO ANSWER THEM SO THAT THEY WON'T TRY
TO GET IN TOUCH WITH YOU. BECAUSE THEY CAN AND WILL CALL YOU SINCE THEY CAN
GET YOUR PHONE NUMBERS FROM CAMPUS INFORMATION.
THANK YOU.
GINA
-------

∂12-Nov-85  1522	MODICA@SU-SCORE.ARPA 	info for TV students   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  15:22:44 PST
Date: Tue 12 Nov 85 15:20:39-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: info for TV students
To: faculty@SU-SCORE.ARPA, instructors@SU-SCORE.ARPA
Message-ID: <12158755846.21.MODICA@SU-SCORE.ARPA>

PLEASE SEND ME THE FOLLOWING INFORMATION AS SOON AS POSSIBLE :

1)IF YOU ARE TEACHING A COURSE WINTER QUARTER, ARE THE STUDENTS GOING TO NEED
  TO HAVE ACCESS TO ONE OF THE COMPUTING SYSTEMS (SUCH AS LOTS)?.....

2)..AND IF SO WILL TV STUDENTS HAVE THE OPTION OF USING A DIFFERENT COMPUTER?

IN YOU REPLY PLEASE INDICATE WHAT COURSE YOU WILL BE TEACHING. THE REASON I
NEED THIS INFORMATION IS THAT THOSE ARE VERY COMMON QUESTIONS FROM TV STUDENTS,
(ALONG WITH WHAT BOOKS ARE GOING TO BE USED, BUT I CAN GET THAT INFO FROM
KATHY BERG), AND I WOULD LIKE TO BE ABLE TO ANSWER THEM SO THAT THEY WON'T TRY
TO GET IN TOUCH WITH YOU. BECAUSE THEY CAN AND WILL CALL YOU SINCE THEY CAN
GET YOUR PHONE NUMBERS FROM CAMPUS INFORMATION.
THANK YOU.
GINA
-------

∂12-Nov-85  1519	ullman@su-aimvax.arpa 	another paper    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  15:15:23 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 12 Nov 85 15:07:10 pst
Date: Tue, 12 Nov 85 15:07:10 pst
From: Jeff Ullman <ullman@diablo>
Subject: another paper
To: nail@diablo

"Recursive Axioms in Deductive databases: various solutions"
by L. Vieille.
Doesn't seem to say much, but surveys the difference between
approaches.  Besides, how many people have three consecutive
vowels in their name?

∂12-Nov-85  1747	PAPA@SU-SCORE.ARPA 	Re: another paper   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  17:47:41 PST
Received: from SU-SCORE.ARPA by su-aimvax.arpa with Sendmail; Tue, 12 Nov 85 17:40:57 pst
Date: Tue 12 Nov 85 17:41:18-PST
From: C. Papadimitriou <PAPA@SU-SCORE.ARPA>
Subject: Re: another paper
To: ullman@SU-AIMVAX.ARPA
Cc: nail@SU-AIMVAX.ARPA
In-Reply-To: Message from "Jeff Ullman <ullman@diablo>" of Tue 12 Nov 85 15:10:56-PST
Message-Id: <12158781449.27.PAPA@SU-SCORE.ARPA>

Gee, I guess not many people have three consecutive vowels in their name.
But then again, some names are long enough to make any pattern overwhelmingly
probable...
---Christos.
-------

∂12-Nov-85  1818	ullman@su-aimvax.arpa 	Re: another paper
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Nov 85  18:09:29 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 12 Nov 85 18:01:18 pst
Date: Tue, 12 Nov 85 18:01:18 pst
From: Jeff Ullman <ullman@diablo>
Subject: Re: another paper
To: PAPA@SU-Score
Cc: nail@diablo

Well I meant in the MIDDLE of their name.
Sorry, Christos.

∂13-Nov-85  0612	PATASHNIK@SU-SUSHI.ARPA 	AFLB abstracts 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  06:12:40 PST
Date: Wed 13 Nov 85 06:12:16-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: AFLB abstracts
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12158918159.8.PATASHNIK@SU-SUSHI.ARPA>

These are the abstracts for the next two AFLBs:
		*************************************

14-Nov-85  :  Richard Anderson (Stanford & MSRI)

	     A Parallel Algorithm for Depth First Search

A new parallel algorithm for constructing a depth first search tree in
an undirected graph will be described.  The algorithm is a P-RAM
algorithm and uses several probabilistic algorithms as subroutines.
The run time of the algorithm is 2↑{\sqrt{\log n}}.  This makes it an
almost RNC algorithm, since the run time is less than n↑e for any e>0.

The standard sequential algorithm for depth first search can be shown
to be ``inherently sequential'', so this shows that substantial speed
up for depth first search is possible when a different approach is
taken.

***** Time and place: November 14, 12:30 pm in MJ352 (Bldg. 460) ******

21-Nov-85  :  Ketan Mulmuley (MSRI)

		Parallel algorithms for rank and matching

A parallel algorithm will be given to find the rank of a matrix. It
was not previously known how to solve this problem in NC.  A new
parallel algorithm for matching will also be pressented.  The
algorithm runs in log↑2 parallel time and uses randomness.

***** Time and place: November 21, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.

If you have a topic you would like to talk about in the AFLB seminar
please let me know.  (Electronic mail: patashnik@su-sushi.arpa, phone:
(415) 497-1787). Contributions are wanted and welcome.  Not all time
slots for this academic year have been filled.

More information about future and past AFLB meetings and topics are in
the file [SUSHI]<patashnik.aflb>aflb.bboard .
						- Oren Patashnik
-------

∂13-Nov-85  1026	CAROL@SU-CSLI.ARPA 	Sketch demo    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  10:25:50 PST
Date: Wed 13 Nov 85 10:16:20-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Sketch demo
To: Folks@SU-CSLI.ARPA, LFG@SU-CSLI.ARPA, Bboard@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA

*********************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DL
						9 Nov 85

						 
						    												
Richard Burton of Xerox Parc will demonstrate the Sketch package, a
program which enables you to produce drawings or combinations of 
graphics and text.  These can either be saved as Sketch files, or
included in a TEdit file.  Many people use Sketch to produce slides 
for presentations.

Some of you are already familiar with Sketch, and may wish to explore
the subtleties of the program, or raise questions about problems you
have encountered.  Since the documentation is quite clear and comp-
rehensive, it would be easy for those who are interested to spend a 
bit of time  familiarizing themselves with the basics of the program 
so the demo can also be used not only as an introduction, but also as 
an opportunity to learn more advanced features of Sketch, and raise 
questions about bugs or special problems.  


The demo will cover recent updates.

WHEN:   Tues. 19 Nov.
WHERE:  Trailer F

Call if you need the document, or if you have any questions.


                              -Carol     (321-2136,  Ventura 28)

ION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION*
*********************************************************************

-------

∂13-Nov-85  1040	CAROL@SU-CSLI.ARPA 	Sketch demo, correction  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  10:40:23 PST
Date: Wed 13 Nov 85 10:31:56-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Sketch demo, correction
To: Folks@SU-CSLI.ARPA, Bboard@SU-CSLI.ARPA

*********************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DL
                   13 Nov 85						    	
			 **CORRECTION**						 
						    												
Richard Burton of Xerox Parc will demonstrate the Sketch package, a
program which enables you to produce drawings or combinations of 
graphics and text.  These can either be saved as Sketch files, or
included in a TEdit file.  Many people use Sketch to produce slides 
for presentations.

Some of you are already familiar with Sketch, and may wish to explore
the subtleties of the program, or raise questions about problems you
have encountered.  Since the documentation is quite clear and comp-
rehensive, it would be easy for those who are interested to spend a 
bit of time  familiarizing themselves with the basics of the program 
so the demo can also be used not only as an introduction, but also as 
an opportunity to learn more advanced features of Sketch, and raise 
questions about bugs or special problems.  


The demo will cover recent updates.

WHEN:   4 p.m.     
        Tues. 19 Nov.

WHERE:  Trailer F

Call if you need the document, or if you have any questions.


                              -Carol     (321-2136,  Ventura 28)

ION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEW**DLION**NEWS**DLION**
*********************************************************************
-------

∂13-Nov-85  1141	NILSSON@SU-SCORE.ARPA 	Herb Grosch 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  11:41:52 PST
Date: Wed 13 Nov 85 11:41:34-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Herb Grosch
To: senior-faculty@SU-SCORE.ARPA
Message-ID: <12158978105.13.NILSSON@SU-SCORE.ARPA>

Herb Grosch has written to President Kennedy volunteering to come
to Stanford to be a "living monument."  Should I be interested in
this?  -Nils
-------

∂13-Nov-85  1147	DALRYMPLE@SU-CSLI.ARPA 	party 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  11:47:29 PST
Date: Wed 13 Nov 85 11:38:42-PST
From: Mary Dalrymple <DALRYMPLE@SU-CSLI.ARPA>
Subject: party
To: folks@SU-CSLI.ARPA, linguists@SU-CSLI.ARPA

Don't miss the second weekly Linguistics happy hour, this
Friday at 4:45 in the Greenberg Room, Linguistics Department!

-------

∂13-Nov-85  1353	ACUFF@SUMEX-AIM.ARPA 	Bug reports for Explorers   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  13:53:04 PST
Date: Wed 13 Nov 85 13:51:11-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Bug reports for Explorers
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12159001702.21.ACUFF@SUMEX-AIM.ARPA>

Folks,
   Since we are running pre-release software on our Explorers, please
send all of your bug reports to me.  In particular, please don't sent
them to the BUG-TI-EXPLORER mailing list unless you are certain that
the problem also occurs under released software.  This way I can report
the problems to TI so that they can be fixed before the released
version comes out, but without misrepresenting TI's software quality
control to people who don't know that we are using pre-release software.

	Thanks,
	-- Rich
-------

∂13-Nov-85  1403	REGES@SU-SCORE.ARPA 	TAs 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  14:03:39 PST
Date: Wed 13 Nov 85 14:02:23-PST
From: Stuart Reges <REGES@SU-SCORE.ARPA>
Subject: TAs
To: faculty@SU-SCORE.ARPA
Office: Margaret Jacks 210, 497-9798
Message-ID: <12159003742.53.REGES@SU-SCORE.ARPA>

I will be appointing Winter quarter TAs very soon now.  I would like to do 
things a bit differently this time and appoint some people for Spring at the
same time.  My basic deal will be that if they are currently a TA and are
performing satisfactorily, they will be allowed to sign up for both Winter and
Spring.  Therefore, I need to know if there are any TAs who are not performing
satisfactorily.  Generally our TAs do a very fine job, so I will assume that is
true unless I hear otherwise.  I have already had mail from one instructor about
a TA, but I would appreciate hearing from more of you if there are more
problems.
-------

∂13-Nov-85  1606	NILSSON@SU-SCORE.ARPA 	Fac. Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  16:06:00 PST
Date: Wed 13 Nov 85 15:43:01-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Fac. Lunch
To: faculty@SU-SCORE.ARPA
Message-ID: <12159022059.13.NILSSON@SU-SCORE.ARPA>

I will not be able to organize a faculty lunch next Tuesday, Nov. 19
because of another commitment.  If anyone would like to suggest and
lead a discussion topic in my absence please contact Anne Richardson
(Richardson@score) before tomorrow late pm so she can maintain the
usual lunch order with The New Leaf.  (Otherwise, we'll just cancel
the order for next week.)  -Nils
-------

∂13-Nov-85  1704	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	BATS program
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  17:03:58 PST
Received: from decwrl.DEC.COM by SU-SUSHI.ARPA with TCP; Wed 13 Nov 85 17:02:01-PST
Received: from magic.ARPA (magic) by decwrl.DEC.COM (4.22.01/4.7.34)
	id AA03278; Wed, 13 Nov 85 17:00:36 pst
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA21044; Wed, 13 Nov 85 16:58:54 pst
From: broder@decwrl.DEC.COM (Andrei Broder)
Message-Id: <8511140058.AA21044@magic.ARPA>
Date: 13 Nov 1985 1658-PST (Wednesday)
To: aflb.all@sushi.ARPA
Cc: src@decwrl.DEC.COM, allen%ucsc.CSNET@CSNET-RELAY.ARPA
Fcc: sent
Subject: BATS program


The next Bay Area Theory Seminar will be held at DEC-SRC, Friday,
November 22, starting at 10 a.m. Below is the program.  Abstracts  and
driving directions will be mailed separately.   

Coordinators: please let me know how many people are coming from your
site in order to enable us to order an appropriate amount of food.  

Uncoordinated individuals: please let me know that you are coming.  (My
phone is (415) 853-2118.)

- Andrei 


10:00 Lyle Ramshaw (DEC - SRC) -  Beziers and B-Splines as multiaffine maps

11:00 Anna Karlin (Stanford) - An optimal algorithm for PRAM simulation using
universal hashing

12:00 Lunch 

1:00 Open problems - subject to stock on hand and prior solving.

1:30 Ketan Mulmuley (U. C. Berkeley)  - Parallel algorithms for rank and
matching.

2:30 Maria Klawe (IBM - Almaden) - Algorithms for embedding graphs in
multi-layer grids

∂13-Nov-85  1749	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #31  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  17:49:44 PST
Date: 13 Nov 85 1737-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #31
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 13 Nov 1985     Volume 1 : Issue 31

Today's Topics:

                          Lisp Vectorizable?
                         Smalltalk vs Actors

----------------------------------------------------------------------

Date: Tue, 12 Nov 85 13:47:55 est
From: eldridge@EDN-VAX.ARPA (Charles Eldridge)

In Re:


  [Lisp on a Cray is 5-10x a Symbolics 3600, as I remember, while
  Fortran on a Cray is 50-100x a VAX 780 on benchmarks of interest to
  number hackers.  Since a 3600 and a VAX are approximately the same
  "power", give or take a factor of 2, we're missing an order of
  magnitude somewhere.  Is the order of magnitude due to Cray hardware
  being inappropriate for Lisp or to the lack of a good Lisp compiler
  for the Cray, or some combination?  Would a functional subset of Lisp
  be appropriate for vectorization?  -- BD]


The speed of LISP is an interesting issue.  I would
conjecture that these odd relationships are a consequence
of the irregularity of instructions and memory references
in the execution of LISP applications.  In contrast,
vectorized problems are very regularized.  It may also be
that vectorizing LISP applications would be contrary to the
high flexibility offered by that language.  Vectorizing
numerical applications requires careful hand tailoring of
code in most cases.  Where might LISP be hand tailored to
support vectorization?  We must look for repeated application
of the same function, for example as in MAPCAR or in a recursive
function.  But, in the case of MAPCAR and its relatives, we
may have a complex function to evaluate with each application.
Vectorization thrives on simpler functions that can be implemented
in hardware.  In the case of recursion, the flow of control
is unpredictable.  We can't see in advance where the recursion
will stop.

LISP processing has been optimized at the microscopic or local
level using specialized instruction set architectures to
manipulate list structures.  A more global analysis, as would
be obtained by running LISP applications in a simulated mode
in order to view instruction and memory reference patterns, will be
necessary to ascertain whether a subset of LISP is appropriate
for vectorizing.

--Charles Eldridge

------------------------------

Date: 12 Nov 1985 20:34:26-EST
From: linus!brando@mitre-bedford.ARPA
Date: Tue, 12 Nov 85 15:31:10 est
From: T. J. Brando <linus!brando>
Subject: Smalltalk vs Actors

The remarks that have been made by schwamb@mitre.ARPA and Hendler@tove
describing the *key* or *clear distinctions* between object-oriented
paradigms and message-passing paradigms have only served to strengthen
my belief that there is no essential difference between Actors and the
other object-oriented paradigms that have been mentioned.  A careful
reading of the postings of Karl and Jim points out that they are, in
fact, comparing object-oriented *paradigms* with message-passing
*implementations*.  *Of course* the object-oriented implementations to
date have been sequential, because they were done on sequential machines!
On the other hand, Actors is being *designed* to execute in a multiple
processor environment!  How do the two *paradigms* compare, however, if
we *implement* message passing in a Smalltalk system such that each
message contains the identity of the object to which the response should
be sent (or nil, if no response is needed), and a sending object can
wait for a response or continue executing immediately after sending a
message?

Thom
linus!brando@mitre-bedford.arpa

------------------------------

End of PARSYM Digest
********************

∂13-Nov-85  1758	EMMA@SU-CSLI.ARPA 	Newsletter November 14, No. 2  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  17:57:58 PST
Date: Wed 13 Nov 85 17:05:26-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter November 14, No. 2
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
November 14, 1985               Stanford                        Vol. 3, No. 2
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, November 14, 1985

   12 noon		TINLunch
     Ventura Hall       Machines and the Mental
     Conference Room    by Fred Dretske
			Discussion led by Jon Barwise  (Barwise@su-csli.arpa)
			(Abstract on page 2)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	A Morphological Recognizer with Syntactic and
     Room G-19		Phonological Rules
			John Bear (Bear@sri-ai.arpa)
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Redwood Hall	Partial Truth Conditions and Their Logics
     Room G-19		Hans Kamp, University of Texas
                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *NEXT* THURSDAY, November 21, 1985

   12 noon		TINLunch
     Ventura Hall       Parsing as Deduction?
     Conference Room    by Fernando Pereira and David Warren
			Discussion led by Mark Johnson (Johnson@su-csli.arpa)
			(Abstract will appear next week)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	Interactive Modularity
     Room G-19		Ron Kaplan, Xerox PARC (Kaplan.pa@xerox.arpa)
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Redwood Hall	An Introduction to Information-based Complexity
     Room G-19		J. F. Traub, Computer Science Department, Columbia
			(Abstract on page 3)
                              ←←←←←←←←←←←←
!
Page 2                     CSLI Newsletter                   November 14, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                          THIS WEEK'S TINLUNCH
                         Machines and the Mental
                             by Fred Dretske

      The paper argues that current computers do not exhibit anything
   that deserves to be called rational cognitive activity.  Dretske even
   claims that they can't add!  He then goes on to discuss how one might
   build something that deserves to be called a rational machine.
      This short, well-written paper is Dretske's Presidential Address to
   the APA.  Read it and come prepared for a lively session.  --Jon Barwise
                              ←←←←←←←←←←←←
                        THIS WEEK'S CSLI SEMINAR
                       A Morphological Recognizer
                  with Syntactic and Phonological Rules

      In many natural language processing systems currently in use the
   morphological phenomena are handled by programs which do not interpret
   any sort of rules, but rather contain references to particular
   morphemes, graphemes, and grammatical categories.  Recently
   Koskenniemi, Karttunen, Kaplan and Kay have showed how to build
   morphological analyzers in which the descriptions of the phonological
   (or orthographic) and syntactic phenomena are separable from the code.
   A system will be described which is based on the work of the people
   mentioned above.  There are two main differences between the system to
   be described here and other existing systems of its kind.  Firstly,
   the spelling rules are not translated into finite state transducers,
   but are interpreted directly, thereby yielding a system more amenable
   to grammar development than one in which considerable time is
   necessary to compile the rules into transducers.  Secondly, the
   syntactic component has more flexibility than other current systems.
   Instead of encoding the syntax entirely in the lexicon by stipulating
   about each morpheme what category it is and what category may come
   next, this system contains a file of patr-type rules with the power to
   unify dags containing disjunctions.			--John Bear
                               ----------
                           NEXT WEEK'S SEMINAR
                         Interactive Modularity
                         Ron Kaplan, Xerox PARC

      Comprehensible scientific explanations for most complex natural
   phenomena are modular in character.  Phenomena are explained in terms
   of the operation of separate and independent components, with
   relatively minor interactions.  Modular accounts of complex cognitive
   phenomena, such as language processing, have also been proposed, with
   distinctions between phonological, syntactic, semantic, and pragmatic
   modules, for example, and with distinctions among various rules within
   modules.  But these modular accounts seem incompatible with the
   commonplace observations of substantial interactions across component
   boundaries: semantic and pragmatic factors, for instance, can be shown
   to operate even before the first couple of phonemes in an utterance
   have been identified.  In this talk I consider several methods of
   reconciling modular descriptions in service of scientific explanation
   with the apparent interactivity of on-line behavior.  Run-time methods
   utilize interpreters that allow on-line interleaving of operations
   from different modules, perhaps including additional ``scheduling''
!
Page 3                     CSLI Newsletter                  November 14, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

   components for controlling the cross-module flow of information.  But
   depending on their mathematical properties, modular specifications may
   also be transformed by off-line, compile-time operations into new
   specifications that directly represent all possible cross-module
   interactions.  Such compilation techniques allow for run-time
   elimination of module boundaries and of intermediate levels of
   representation.
                               ----------
                         NEXT WEEK'S COLLOQUIUM
             An Introduction to Information-based Complexity
                               J. F. Traub
           Computer Science Department,	Columbia University

      In information-based complexity ``information'' is, informally,
   what we know about a problem which we wish to solve.
      The goal of information-based complexity is to create a general
   theory about problems with partial and contaminated information and to
   apply the results to solving specific problems in varied disciplines.
   Problems with partial and contaminated information occur in areas such
   as vision, medical imaging, prediction, geophysical exploration,
   signal processing, control, and scientific and engineering
   calculation.
      For problems with partial and contaminated information, very general
   results can be obtained at the ``information level.''  Among the
   general results to be discussed is the power of parallel
   (non-adaptive) information and the application of such information to
   the solution of problems on distributed systems.
      The methodology and results of information-based complexity will be
   contrasted with the study of NP-complete problems where the
   information is assumed to be complete, exact, and free.
                               ----------
                          PIXELS AND PREDICATES
               Setting Tables and Illustrations with Style
              Rick Beach, Xerox PARC, (Beach.pa@xerox.arpa)
         CSLI trailers, 1:00 p.m., Wednesday, November 20, 1985

      Two difficult examples of incorporating complex information within
   electronic documents are illustrations and tables.  The notion of
   style, a way of maintaining consistency, helps manage the complexities
   of formatting both tables and illustrations.  The concept of graphical
   style extends document style to illustrations.  Observing that
   graphical style does not adequately deal with the layout of
   information leads to the study of formatting tabular material.  A grid
   system for describing the arrangement of information in a table, and a
   constraint solver for determining the layout of the table are key
   components of this research.  These ideas appear to extend to
   formatting other complex material, including mathematical typesetting
   and page layout.  Several typographic issues for illustrations and
   tables will be highlighted during the talk.
!
Page 4                     CSLI Newsletter                  November 14, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                         LEXICAL PROJECT MEETING
                  Lori Levin, University of Pittsburgh
          Monday, November 18, 10 a.m., Ventura Conference Room

      I will describe a theory of relation changing rules which is
   compatible with LFG.  The theory accounts for the interaction between
   semantically conditioned and syntactically productive relation
   changing rules, ability of relation changing rules to distinguish
   between subjects of unaccusative verbs and subjects of unergative
   verbs, and the apparent directionality of object-to-subject relation
   changes.  In order to handle these properties of relation changing
   rules, I introduce a new mechanism which I call Argument
   Classification which mediates between thematic roles and grammatical
   functions in the lexicon.  I will illustrate the formulation of
   various relation changing rules in English and Dutch using argument
   classification.

   (This talk is part of the Lexical Project meetings but open to
   everybody.)
                               ----------
                       ENVIRONMENTS GROUP MEETING
                     An Environment for VLSI Design
           Mike Spreitzer, Xerox PARC, (Spreitzer@xerox.arpa)
             Monday, November 18, noon, Ventura Seminar Room

     We in the PARC CSL VLSI CAD group are working on making our tools
   more integrated than they have been.  We are defining an in-memory
   data structure for representing the basic structure of a VLSI design.
   Other information is hung on this "skeleton" via property lists.
   Various tools communicate with each other through this decorated
   structure.  We think this will make it easier for the tools to
   cooperate more closely than in the past.
                               ----------
            COMMON SENSE AND NON-MONOTONIC REASONING SEMINAR
                   Some Results on Autoepistemic Logic
                  Wiktor Marek, University of Kentucky
                2:00 PM, Wednesday, November 20, MJH 252

      We discuss some properties of so-called stable theories in
   autoepistemic logic (cf. Moore, AIJ 25 (1985)), that is, sets of
   beliefs of a fully rational agent. We show an operator constructing
   these theories out of their objective parts and investigate the
   complexity of the construction.  We attempt to extend Moore's approach
   to the case of predicate logic.  Finally, we discuss the notion of
   inessential modal extension of a first order theory.
                               ----------
                             NEW CSLI REPORT

      Report No. CSLI-85-36, ``Limits of Correctness in Computers'' by
   Brian Cantwell Smith, has just been published.  This report may be
   obtained by writing to David Brown, CSLI, Ventura Hall, Stanford, CA
   94305 or Brown@SU-CSLI.
-------

∂13-Nov-85  1809	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	Driving to Bats  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  18:09:50 PST
Received: from decwrl.DEC.COM by SU-SUSHI.ARPA with TCP; Wed 13 Nov 85 18:05:46-PST
Received: from magic.ARPA (magic) by decwrl.DEC.COM (4.22.01/4.7.34)
	id AA03893; Wed, 13 Nov 85 18:05:01 pst
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA23937; Wed, 13 Nov 85 18:04:27 pst
From: broder@decwrl.DEC.COM (Andrei Broder)
Message-Id: <8511140204.AA23937@magic.ARPA>
Date: 13 Nov 1985 1804-PST (Wednesday)
To: aflb.all@sushi.ARPA
Cc: 
Fcc: sent
Subject: Driving to Bats


Where is DEC located?

DEC-SRC is located at 130 Lytton Ave., in Palo Alto.  That is the red
brick building on the corner of Lytton and Alma, opposite Bagel Works.
Lytton is parallel to University and is the next street north.  Alma is
parallel to El Camino and is the next street east (towards 101).
University is the street that continues Palm Drive.  The other end of
Palm Drive is the oval in front of Stanford CSD.  The total distance
from CSD to DEC is about 1 mile.

Public transportation:

DEC is one minute away from the Southern Pacific train station and the Palo 
Alto bus depot. 

Driving:

The entrance into our parking structure is from High street.  High is
parallel is one block east of Alma and is a one-way street going from
Lytton towards University.  Hence you need to get first on Lytton, turn
on High, and then immediately turn right into the parking structure.

For out-of-towners:  Exit 101 at University towards Palo Alto.
Continue on University until Middlefield (the first larger
intersection).  Turn right on Middlefield and after just one block turn
left on Lytton.  Continue on Lytton till High (the penultimate stop
light on Lytton - Lytton ends in Alma)  Turn left on High and
immediately right into the parking.

IN THE PARKING:  By the time you arrive probably all ground level
parking will be full.  On the left side, at the end, there is a ramp
that takes you to more parking spaces on the roof.  Register your car
at the front desk.

Biking:  There is some parking area for bikes near the front entrance.
It is generally closed, so that you might have to ask the person at the
front desk to open it for you.

See you next Friday!  - Andrei

∂13-Nov-85  1829	@SU-SCORE.ARPA:coraki!pratt@su-navajo.arpa 	Herb Grosch
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  18:29:35 PST
Received: from su-navajo.arpa by SU-SCORE.ARPA with TCP; Wed 13 Nov 85 18:10:41-PST
Received: by su-navajo.arpa with Sendmail; Wed, 13 Nov 85 18:10:35 pst
Received: by coraki.uucp (1.1/SMI-1.2)
	id AA25633; Wed, 13 Nov 85 13:07:05 pst
Date: Wed, 13 Nov 85 13:07:05 pst
From: coraki!pratt@su-navajo.arpa (Vaughan Pratt)
Message-Id: <8511132107.AA25633@coraki.uucp>
To: senior-faculty@su-score.ARPA
Subject: Herb Grosch

Did Herb say where he'd like to be erected?  The front of the Reagan library
springs to mind.

On many occasions, most visibly during his tenure as ACM president, Herb
played the role of the shrill anti-intellectual.  I found it hard to
identify with ACM's objectives during this time.  A vote for Herb is
my mind a vote counter to what academia in general and this department
in particular stands for - the unfettered pursuit of knowledge.
-v

∂13-Nov-85  2319	@SU-SCORE.ARPA:coraki!pratt@su-navajo.arpa 	Fac. Lunch 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Nov 85  23:19:24 PST
Received: from su-navajo.arpa by SU-SCORE.ARPA with TCP; Wed 13 Nov 85 23:18:55-PST
Received: by su-navajo.arpa with Sendmail; Wed, 13 Nov 85 23:18:50 pst
Received: by coraki.uucp (1.1/SMI-1.2)
	id AA26168; Wed, 13 Nov 85 23:12:53 pst
Date: Wed, 13 Nov 85 23:12:53 pst
From: coraki!pratt@su-navajo.arpa (Vaughan Pratt)
Message-Id: <8511140712.AA26168@coraki.uucp>
To: faculty@su-score.ARPA, richardson@su-score.ARPA
Subject: Fac. Lunch

I will be happy to lead a discussion on either Dolce Far Niente, or on
what happened at the EE retreat, as people prefer.  (The former topic
was intended to be discussed frequently at these lunches when I first
proposed to Gene that we have them.  Not having a pressing topic for
discussion was not meant to be a reason for the faculty not to be
sociable.)
-v

∂14-Nov-85  0002	Operator@SU-SUSHI.ARPA 	bats at dec, 11/22   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  00:02:29 PST
Date: Thursday, 14-Nov-85 00:00:36-PST
From: JF@SU-SUSHI.ARPA
Sender: The Remind Daemon <Operator>
Subject: bats at dec, 11/22
To: aflb.su
Please let me know whether you would like to attend bats at decsrc on
friday, nov. 22.  the talk titles are in 

SUSHI:<JF>bats.announce

I will send you a copy of the file if you don't have a sushi account

-Joan Feigenbaum
(jf@sushi)


-------

∂14-Nov-85  0830	EMMA@SU-CSLI.ARPA 	Newsletter addition  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  08:30:05 PST
Date: Thu 14 Nov 85 08:25:39-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter addition
To: friends@SU-CSLI.ARPA
cc: newsletter@SU-CSLI.ARPA
Tel:  497-3479


The following was omitted from the newsletter.

                            LOGIC SEMINAR
          Truth, the Liar, and Circular Propositions, Cont.
                   Jon Barwise and John Etchemendy
                      Friday, November 15, 1985

   Last time John Etchemendy gave an informal introduction to Peter
Aczel's set theory ZF/AFA, and showed how to use it to model the
Austinian conception of proposition.  He then discussed how the Liar,
the truth-teller, and other paradoxes and puzzles come out in this
model.  This week I will review ZF/AFA very briefly and then use it to
model the Russellian conception of proposition, and discuss how the
same puzzles come out in this model.		         --Jon Barwise


Noon, Math Faculty Lounge, building 380.

-------
-------

∂14-Nov-85  1109	@SU-SCORE.ARPA:guibas@decwrl.DEC.COM 	Re: info for TV students   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  11:06:42 PST
Received: from decwrl.DEC.COM by SU-SCORE.ARPA with TCP; Thu 14 Nov 85 10:06:51-PST
Received: from magic.ARPA (magic) by decwrl.DEC.COM (4.22.01/4.7.34)
	id AA15192; Thu, 14 Nov 85 10:06:27 pst
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA03480; Thu, 14 Nov 85 10:05:52 pst
From: guibas@decwrl.DEC.COM (Leo Guibas)
Message-Id: <8511141805.AA03480@magic.ARPA>
Date: 14 Nov 1985 1005-PST (Thursday)
To: Gina Modica <MODICA@SU-SCORE.ARPA>
Cc: faculty@SU-SCORE.ARPA, instructors@SU-SCORE.ARPA
Subject: Re: info for TV students
In-Reply-To: Your message of Tue 12 Nov 85 15:20:39-PST.
             <12158755846.21.MODICA@SU-SCORE.ARPA>

1) CS248: the students will need to use a 512k Macintosh (there are two
clusters of such machines in Terman).

2) no, but many may have access to 512k Macs on their own.

	L.

∂14-Nov-85  1109	@SU-SCORE.ARPA:guibas@decwrl.DEC.COM 	Re: info for TV students   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  11:06:42 PST
Received: from decwrl.DEC.COM by SU-SCORE.ARPA with TCP; Thu 14 Nov 85 10:06:51-PST
Received: from magic.ARPA (magic) by decwrl.DEC.COM (4.22.01/4.7.34)
	id AA15192; Thu, 14 Nov 85 10:06:27 pst
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA03480; Thu, 14 Nov 85 10:05:52 pst
From: guibas@decwrl.DEC.COM (Leo Guibas)
Message-Id: <8511141805.AA03480@magic.ARPA>
Date: 14 Nov 1985 1005-PST (Thursday)
To: Gina Modica <MODICA@SU-SCORE.ARPA>
Cc: faculty@SU-SCORE.ARPA, instructors@SU-SCORE.ARPA
Subject: Re: info for TV students
In-Reply-To: Your message of Tue 12 Nov 85 15:20:39-PST.
             <12158755846.21.MODICA@SU-SCORE.ARPA>

1) CS248: the students will need to use a 512k Macintosh (there are two
clusters of such machines in Terman).

2) no, but many may have access to 512k Macs on their own.

	L.

∂14-Nov-85  1343	CAROL@SU-CSLI.ARPA 	Meet the Author
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  13:43:08 PST
Date: Thu 14 Nov 85 13:34:10-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Meet the Author
To: Folks@SU-CSLI.ARPA, Bboard@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA

*********************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DL
                   14 Nov 85
						    	
		   **MEET THE AUTHOR**						 
						    												
Richard Burton of Xerox Parc will demonstrate SKETCH, and entertain
questions (and you).

WHEN:   4 p.m.     
        Tues. 19 Nov.

WHERE:  Trailer F

Contact me if you need the document, a quick introductory lesson, 
or if you have any questions.  There is a brief version of the 
documentation on {CSLI}<DANDELION.DOC>SKETCH, and in the 
yellow binders.


                              -Carol     (321-2136,  Ventura 28)

ION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION*
*********************************************************************
-------

∂14-Nov-85  1405	BSCOTT@SU-SCORE.ARPA 	Proposals    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  14:05:43 PST
Date: Thu 14 Nov 85 14:04:20-PST
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: Proposals
To: Academic-Council@SU-SCORE.ARPA
cc: Donoghue@SU-SCORE.ARPA, Mathy@SU-SCORE.ARPA, Anderson@SU-NAVAJO.ARPA
Message-ID: <12159266239.39.BSCOTT@SU-SCORE.ARPA>


It has come to my attention that because of a rush deadline to be met a
proposal was sent to the agency this week without appropriate signatures.
I think that most of you know that is an absolute no-no.  And, the agency
has called and said that signed copies must be submitted, so all the work
has to be duplicated.  Plus, though I don't know this first hand, I understand
that the proposal which was sent contained errors which had to be corrected.

Anyway, this note is to remind you all of the University/SPO requirements.
I also want to remind you that we do need a little turnaround time to pro-
cess proposals, and that it is extremely inconsiderate to ask that budgets
be prepared, proposals checked, etc., on a overnight basis.  Mary Donoghue
and other staff concerned with proposals simply cannot provide this kind of
service.  In the future, I will much appreciate your exerting every effort
to give advance notice of a proposal submission and then to allow a reasonable
time period in which to have it prepared, processed through SPO and mailed.

Thanks in advance.

Betty
-------

∂14-Nov-85  1434	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH plus other business...    
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  14:33:40 PST
Date: Thu 14 Nov 85 14:31:42-PST
From: LANSKY@SRI-AI.ARPA
Subject: Next Monday's PLANLUNCH plus other business...
To: planlunch.dis: ;

PLANLUNCHers:

I have recently received a request not to send out PLANLUNCH reminders 
(i.e. the message I send out the day previous to the talk).   My only fear is
that this reminder message is the main reason PLANLUNCH has had such good
attendance!  I am, however, willing to form two mailing lists -- 
planlunch + planlunch-reminder.
If you DO NOT wish to be on the reminder list, please send me a message.
(Unfortunately, though, if you happen to be on MUGS or JMC-LISTS I will have
a hard time deleting you unless someone sends me the contents of those
lists...).  

Also note that I will henceforth try to include the computer net
address of the speaker in planlunch announcements.

-Amy Lansky
-------------------------------------------------------------------------

	EXPLOITATION OF CONSTRAINTS IN DEDUCTIVE DESIGN SYNTHESIS

			    Jeff Finger
		        Stanford University
				
	 	    11:00 AM, MONDAY, November 18
       SRI International, Building E, Room EJ228 (new conference room)

The talk will cover two related topics in deductive design synthesis: 
(1) efficiency gained by reasoning forward from subgoals, and 
(2) advantages and disadvantages of using a declarative representation 
    for partially completed designs.

The first part of the talk gives the deductive framework for capturing the
following intuition:

	Suppose I have decided that X and Y and to be true of my
	design.  Perhaps I should think about what else X and Y imply
	about the design, say Z.  Otherwise, I might waste time
	trying to complete the design process by making decisions
	that have *already* been ruled out by X and Y, for example,
	NOT(Z).

The conditions that X and Y imply (called "necessary constraints" or "NC's")
are found via reasoning forward from subgoals. We show how NC's of a
subgoal can be used to prune the design space either by preventing some
impossible possibilities from ever being generated or by providing a quick
means of filtering bad choices.  In terms of resolution, the above use of
NC's corresponds to the rather counterintuitive notion of allowing
OR-INTRODUCTION on clauses in the set of support. We will also discuss
inheritance of NC's from goal to subgoal and the relation of finding NC's to
that of checking consistency of partially completed designs.

The second part of the talk deals with declarative representation of
partially completed designs. Deductive design systems such as QA3 or Manna
and Waldinger's reify the design as a single term in the logic.
However, it is difficult to express many sorts of constraints on partially
completed designs as a single term. Examples include two actions in an
unspecified order, or the constraint that Action A takes place less than 3
seconds or more than 8 seconds after Action B. We present a system called
RESIDUE in which we build up the design as a set of facts we are willing to
assume about of the design.  Using facts rather than a single term, we can
make finer-grained decisions, avoiding unwitting commitments that might
result in unnecessary backtracking.  In addition, forward reasoning on
subgoals (as in the first portion of the talk) may be done directly on the
set of facts. 
-------
-------

∂14-Nov-85  1459	PHILOSOPHY@SU-CSLI.ARPA 	Colloquium
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  14:59:11 PST
Date: Thu 14 Nov 85 14:44:41-PST
From: Eve Wasmer <PHILOSOPHY@SU-CSLI.ARPA>
Subject: Colloquium
To: bboard@SU-CSLI.ARPA
cc: folks@SU-CSLI.ARPA


                PHILOSOPHY DEPARTMENT COLLOQUIUM


                        David Hilbert

                           Stanford



                    COLORS AND SCIENCE


             Friday, November 22  3:15 p.m.

             Philosophy Seminar Room 92Q
-------

∂14-Nov-85  1515	LANSKY@SRI-AI.ARPA 	oops!
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  15:12:35 PST
Date: Thu 14 Nov 85 14:37:03-PST
From: LANSKY@SRI-AI.ARPA
Subject: oops!
To: planlunch.dis: ;

I guess my temporal logic was wrong -- ``henceforth'' does include NOW!
Jeff Finger's net address is:  JFINGER@SUSHI.

-Amy
-------

∂14-Nov-85  2006	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 19 (R. Alterman, UCB) 
Received: from UCBVAX.BERKELEY.EDU by SU-AI.ARPA with TCP; 14 Nov 85  20:06:46 PST
Received: by ucbvax.berkeley.edu (5.31/1.2)
	id AA04624; Thu, 14 Nov 85 16:55:28 PST
Received: by cogsci (5.31/5.13)
	id AA05204; Thu, 14 Nov 85 16:57:53 PST
Date: Thu, 14 Nov 85 16:57:53 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8511150057.AA05204@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 19 (R. Alterman, UCB)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
$9                                      ←λF←λa←λl←λl ←λ1←λ9←λ8←λ5
$9                        Cognitive Science Seminar - IDS 237A

$9                         Tuesday, November 19, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

                 ``Adaptive Planning is Commonsense Planning''
                               Richard Alterman
                       Computer Science Division, U.C.B.

        A  characteristic  of  commonsense  planning  is  that  it   is
        knowledge  intensive.   For  most  mundane  sorts of situations
        human planners have access to, and are capable  of  exploiting,
        large quantities of knowledge.  Commonsense planners re-use old
        plans under their normal circumstances.  Moreover,  commonsense
        planners  are  capable  of  refitting  old  plans to novel cir-
        cumstances.  A commonsense planner can plan about a wide  range
        of phenomena, not so much because his/her depth of knowledge is
        consistent throughout that range, but because s/he  can  re-fit
        old plans to novel contexts.

             This talk is about an  approach  to  commonsense  planning
        called ←λa←λd←λa←λp←λt←λi←λv←λe ←λp←λl←λa←λn←λn←λi←λn←λg. An adaptive planner plans by exploit-
        ing planning knowledge in a manner that delays the reduction of
        commonsense planning to problem-solving.    Key elements in the
        theory of adaptive planning are  its  treatment  of  background
        knowledge  and  the  introduction  of  a  notion of planning by
        situation matching.  This talk will describe adaptive  planning
        as  it  applies to a number of commonsense planning situations,
        including: riding the NYC subway, trading  books,  transferring
        planes at JFK airport, and driving a rented car.
        ----------------------------------------------------------------
        UPCOMING TALKS
$9           November 26:Eve Clark, Linguistics, Stanford
        December   3:Bernard Baars, Langley Porter, UCSF
        ----------------------------------------------------------------
        ELSEWHERE ON CAMPUS
        Jeff Shrager of Xerox  PARC  will  speak  on  ``Instructionless
        Learning'' at the SESAME Colloquium on November 18, 2515 Tolman
        Hall, 4:00pm.

        The Physics Department is sponsoring a talk by J.  J.  Hopfield
        of  CALTECH  on Wednesday, November 20 at 4:30pm in 1 Le Conte.
        Dr. Hopfield will be speaking on ``Neural  Networks.''   A  tea
        precedes the talk.

∂14-Nov-85  2017	SF@SU-CSLI.ARPA 	Talk by Rachel Feferman
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  20:17:30 PST
Date: Thu 14 Nov 85 15:17:47-PST
From: Sol Feferman <SF@SU-CSLI.ARPA>
Subject: Talk by Rachel Feferman
To: bboard@SU-CSLI.ARPA, folks@SU-CSLI.ARPA

		Saturday Nov. 16 and Sunday Nov. 17 at 8:00

	EYES  HEART  HAND:  A translation in images
	
			Rachel Feferman

	A slide show/talk about her work and the creative process

		Animated films, drawings, paintings, doll-figures

			     at

	Watercourse Way Movement Center     855 High Street at Channing

			doors open at 7:30

	Seating limited   Tickets $5.00     to reserve call:493-4540

		ART WORK WILL BE ON DISPLAY AND FOR SALE

			SUNDAY RECEPTION    2-5pm
-------

∂14-Nov-85  2023	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 19 (R. Alterman, UCB)  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  20:23:20 PST
Received: from ucbvax.berkeley.edu by SU-CSLI.ARPA with TCP; Thu 14 Nov 85 19:58:52-PST
Received: by ucbvax.berkeley.edu (5.31/1.2)
	id AA04624; Thu, 14 Nov 85 16:55:28 PST
Received: by cogsci (5.31/5.13)
	id AA05204; Thu, 14 Nov 85 16:57:53 PST
Date: Thu, 14 Nov 85 16:57:53 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8511150057.AA05204@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 19 (R. Alterman, UCB)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
$9                                      ←λF←λa←λl←λl ←λ1←λ9←λ8←λ5
$9                        Cognitive Science Seminar - IDS 237A

$9                         Tuesday, November 19, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

                 ``Adaptive Planning is Commonsense Planning''
                               Richard Alterman
                       Computer Science Division, U.C.B.

        A  characteristic  of  commonsense  planning  is  that  it   is
        knowledge  intensive.   For  most  mundane  sorts of situations
        human planners have access to, and are capable  of  exploiting,
        large quantities of knowledge.  Commonsense planners re-use old
        plans under their normal circumstances.  Moreover,  commonsense
        planners  are  capable  of  refitting  old  plans to novel cir-
        cumstances.  A commonsense planner can plan about a wide  range
        of phenomena, not so much because his/her depth of knowledge is
        consistent throughout that range, but because s/he  can  re-fit
        old plans to novel contexts.

             This talk is about an  approach  to  commonsense  planning
        called ←λa←λd←λa←λp←λt←λi←λv←λe ←λp←λl←λa←λn←λn←λi←λn←λg. An adaptive planner plans by exploit-
        ing planning knowledge in a manner that delays the reduction of
        commonsense planning to problem-solving.    Key elements in the
        theory of adaptive planning are  its  treatment  of  background
        knowledge  and  the  introduction  of  a  notion of planning by
        situation matching.  This talk will describe adaptive  planning
        as  it  applies to a number of commonsense planning situations,
        including: riding the NYC subway, trading  books,  transferring
        planes at JFK airport, and driving a rented car.
        ----------------------------------------------------------------
        UPCOMING TALKS
$9           November 26:Eve Clark, Linguistics, Stanford
        December   3:Bernard Baars, Langley Porter, UCSF
        ----------------------------------------------------------------
        ELSEWHERE ON CAMPUS
        Jeff Shrager of Xerox  PARC  will  speak  on  ``Instructionless
        Learning'' at the SESAME Colloquium on November 18, 2515 Tolman
        Hall, 4:00pm.

        The Physics Department is sponsoring a talk by J.  J.  Hopfield
        of  CALTECH  on Wednesday, November 20 at 4:30pm in 1 Le Conte.
        Dr. Hopfield will be speaking on ``Neural  Networks.''   A  tea
        precedes the talk.

∂14-Nov-85  2045	@SU-SCORE.ARPA:cbosgd!packard!ihesa!olaf@seismo.CSS.GOV
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  20:45:08 PST
Received: from seismo.CSS.GOV by SU-SCORE.ARPA with TCP; Thu 14 Nov 85 14:52:09-PST
Return-Path: <cbosgd!packard!ihesa!olaf>
Received: from cbosgd.UUCP by seismo.CSS.GOV with UUCP; Thu, 14 Nov 85 17:40:02 EST
Received: by cbosgd.ATT.UUCP (4.12/UUCP-Project/11.09.85)
	id AA18330; Thu, 14 Nov 85 16:16:25 est
Date: Thu, 14 Nov 85 15:02:25 cst
From: packard!ihesa!olaf@seismo.CSS.GOV (O I Henjum)
Message-Id: <8511142102.AA16654@ihesa.UUCP>
Received: by ihnp4.ATT.UUCP id AA17591; 14 Nov 85 15:05:48 CST (Thu)
Received: by ihesa.UUCP (4.12/4.7)
	id AA16654; Thu, 14 Nov 85 15:02:25 cst
Received: from ihnp4.UUCP by py/garage/packard.DK; 8511142108
To: csd@su-score.arpa, olaf@seismo.CSS.GOV

I realize this letter will probably get mailed to many people who couldn't care less
about it and I apologize to them in advance.  However, I simply don't have the E-Mail
addresses for everybody who DOES care to see it.

My new mailing address is:

Olaf I. Henjum
4721 St. Joseph Creek Road #4B
Lisle, IL 60532

Home telephone number: (312) 968-2468

If you want to keep in touch with me, please reply to this msg and I'll collect the
return paths from your messages.

   -- Olaf Henjum ("ihnp4!ihesa!olaf"@berkeley)

∂14-Nov-85  2145	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.berkeley.edu  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 14 Nov 85  21:45:35 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 14 Nov 85 21:44:23-PST
Received: from ucbvax.berkeley.edu by SU-SCORE.ARPA with TCP; Thu 14 Nov 85 21:43:34-PST
Received: by ucbvax.berkeley.edu (5.31/1.2)
	id AA29649; Wed, 13 Nov 85 10:56:07 PST
Received: by ernie (5.31/5.13)
	id AA24751; Wed, 13 Nov 85 10:55:15 PST
Date: Wed, 13 Nov 85 10:55:15 PST
From: karp@ernie.berkeley.edu (Richard Karp)
Message-Id: <8511131855.AA24751@ernie>
To: aflb.local@su-score.arpa, allmsgs@ernie.berkeley.edu,
        csfac@ernie.berkeley.edu, csmsgs@ernie.berkeley.edu,
        theory-b@ernie.berkeley.edu, theory-f@ernie.berkeley.edu

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley   (415)642-0143

The following seminars will be held in the MSRI Lecture Hall

Tuesday, Nov. 19   2:00
Eva Tardos "How to Make Polynomial Algorithms Strongly Polynomial"

Tuesday, Nov. 19   4:00
Joe Traub  "Information-Based Complexity"

Thursday, Nov. 21  4:00
Jonathan Buss "Relativized Alternations"

∂15-Nov-85  0734	JF@SU-SUSHI.ARPA 	bats abstracts   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 15 Nov 85  07:34:16 PST
Date: Thu 14 Nov 85 23:12:48-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: bats abstracts
To: aflb.su@SU-SUSHI.ARPA
Message-ID: <12159366086.14.JF@SU-SUSHI.ARPA>

I now have the abstracts for next week's bats at DEC.  They are in

SUSHI:<JF>bats.abstracts

If you cannot conveniently read sushi files, let me know and I will send
you a copy.
Joan
-------

∂15-Nov-85  0733	@SU-SUSHI.ARPA:broder@decwrl.DEC.COM 	Abstracts for BATS talks at DEC 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 15 Nov 85  07:33:19 PST
Received: from decwrl.DEC.COM by SU-SUSHI.ARPA with TCP; Thu 14 Nov 85 22:31:23-PST
Received: from magic.ARPA (magic) by decwrl.DEC.COM (4.22.01/4.7.34)
	id AA26826; Thu, 14 Nov 85 22:27:57 pst
Received: by magic.ARPA (4.22.01/4.7.34)
	id AA18287; Thu, 14 Nov 85 22:27:22 pst
From: broder@decwrl.DEC.COM (Andrei Broder)
Message-Id: <8511150627.AA18287@magic.ARPA>
Date: 14 Nov 1985 2227-PST (Thursday)
To: aflb.all@sushi
Cc: pryde@decwrl.DEC.COM, allen%ucsc.CSNET@CSNET-RELAY.ARPA,
        terada@cad.berkeley.edu
Fcc: sent
Subject: Abstracts for BATS talks at DEC


Here are the abstracts for the  Bay Area Theory Seminar to be held at
DEC-SRC, Friday, November 22, starting at 10 a.m. - Andrei

----------------------------------------------------------------------

10:00 - Lyle Ramshaw (DEC) -  Beziers and B-Splines as multiaffine maps

Curves and surfaces that are built up as splines out of polynomial pieces
are fundamental in computer aided geometric design.  There is an established
body of theory, associated with the names Bezier, de Casteljau, and de Boor,
for conveniently specifying, evaluating, and manipulating such splines.  We
shall redevelop this standard theory in a novel way by exploiting a general
mathematical principle:  functions of degree n in one argument are
essentially equivalent to symmetric functions of n arguments that have
degree one in each argument separately.  Tackling splines from this new
perspective makes many things simpler.  In particular, it gives us a good
system of multivariate labels for the points that arise in spline
constructions.  These labels make the geometric relations between the points
so obvious that it becomes straightforward to reinvent the standard
algorithms in this area.


11:00 Anna Karlin (Stanford) - An optimal algorithm for PRAM simulation using
universal hashing

We present a new probabilistic algorithm which simulates T steps of an
arbitrary given PRAM program (where the number of shared variables is
polynomial in the number of processors) in O(T log n) steps on a butterfly,
with probability tending to 1. The algorithm utilizes a log n - universal
class of hash functions. 

The previously known best results for simulating T steps of an arbitrary
PRAM program on a bounded degree network are O(T (log n)↑2 log* n) steps in
the deterministic case (nonconstructive) and O(T (log n)↑2) steps, with
probability tending to 1, in the probabilistic case.

An application of the parallel hashing scheme developed will be presented.

(This work was done jointly with Eli Upfal.)

12:00 Lunch (catered by "Picnics in the Park")

1:00 Open problems

1:30 Ketan Mulmuley (U. C. Berkeley)  - Parallel algorithms for rank and
matching.

We show that the rank of a matrix over an arbitrary field can be computed in
deterministic 0(log↑2 n) time using a polynomial number of processors.  As a
result, many important problems descend in the complexity hierarchy from RNC
to NC.  These include basic problems in linear algebra, notably solving a
system of linear equations over an arbitrary field, and many other problems
related to group theory, graph isomorphism, factorization of polynomials.

We shall also give a very simple RNC↑2 algorithm to construct a maximal
matching in a graph.  This is an improvement over the previous RNC↑3
algorithm due to Karp and Widgerson.


2:30 - Maria Klawe (IBM - Almaden) - Algorithms for embedding graphs in
multi-layer grids

This talk presents a model for multi-layer printed circuit boards, and a
number of area-efficient algorithms for laying out circuits on such boards.
In some cases we can show that the area used by these algorithms cannot be
substantially reduced.  This work significantly improves previously known
results even for a single layer.

(joint work with Alok Aggarwal, David Lichtenstein, Nati Linial and Avi
Wigderson)


∂15-Nov-85  0848	ACUFF@SUMEX-AIM.ARPA 	Newest batch of Explorers   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 15 Nov 85  08:48:23 PST
Date: Fri 15 Nov 85 02:19:24-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Newest batch of Explorers
To: ksl-lispm@SUMEX-AIM.ARPA, DELANEY@SUMEX-AIM.ARPA, nii@SUMEX-AIM.ARPA,
    WASHINGTON@SUMEX-AIM.ARPA, UNRUH@SUMEX-AIM.ARPA
Message-ID: <12159400053.19.ACUFF@SUMEX-AIM.ARPA>

   The newest batch of Explorers (5 soon to be 6 and then 7, bringing
the total to 18) is pretty much ready for use.  4 machines will be
located in the lobby of Whelan Bldg. C as soon as their tables arrive
(hopefullly next week), and there are two pool machines in Whelan
A-1105.

   Please read the "Getting Started with Explorers at KSL" and related
documents before the one marked "READ FIRST" from TI.  In particular,
remember that we are running pre-release software, so please be sure
to let me know of problems.

	-- Rich
-------

∂15-Nov-85  0848	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #32  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 15 Nov 85  08:48:50 PST
Date: 15 Nov 85 0637-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #32
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Friday, 15 Nov 1985       Volume 1 : Issue 32

Today's Topics:

                         Smalltalk vs Actors
                          Lisp Vectorizable?

----------------------------------------------------------------------

Date: Wed, 13 Nov 85 23:02 EST
From: Henry Lieberman <Henry%OZ@MIT-REAGAN.ARPA>
Subject: Smalltalk vs Actors


I think the remarks made by Schwab and Hendler on Actors vs. Smalltalk
were well done, which is why I didn't feel the need to reply myself.
There are lots of inessential differences, but the major ones are that
actors are designed for parallelism, and permit non-stack control
structures through the use of continuations.

However, Brando seems still confused.  The fact that Smalltalk is
implemented on a sequential machine doesn't matter.  Actors have in
fact been implemented several times on sequential machines, too.  But
the languages handle parallelism very differently.  Smalltalk's
parallelism requires a scheduler [one is written in Smalltalk], and
Smalltalk processes must EXPLICITLY relinquish control.  This is a
coroutine model, not a parallel one.  Actors assume parallel
execution, create parallelism using the future mechanism, and must
explicitly use serializer actors for sequential synchronization.  An
actor implementation on a sequential machine must make all
time-sharing completely invisible to the user's program.

Another difference between Actors and Smalltalk is the mechanism for
sharing knowledge between objects. Smalltalk uses classes, subclasses
and instances, borrowing the model from Simula.  Actors make no
distinction between classes and instances -- every object can be
considered as a "prototype" for creating new objects.  If an object
receives a message that it doesn't understand, it can "delegate" or
forward the request to another object with more general knowledge.

------------------------------

Date: Thu 14 Nov 85 00:27:34-EST
From: Randy Haskins <rh%MIT-EECS@mit-eddie.MIT.EDU>
Subject: Re: PARSYM Digest   V1 #31

    Date: Tue, 12 Nov 85 13:47:55 est
    From: eldridge@EDN-VAX.ARPA (Charles Eldridge)

    The speed of LISP is an interesting issue.  I would
    conjecture that these odd relationships are a consequence
    of the irregularity of instructions and memory references
    in the execution of LISP applications.  In contrast,
    vectorized problems are very regularized.  It may also be
    that vectorizing LISP applications would be contrary to the
    high flexibility offered by that language.  Vectorizing
    numerical applications requires careful hand tailoring of
    code in most cases.  Where might LISP be hand tailored to
    support vectorization?  We must look for repeated application
    of the same function, for example as in MAPCAR or in a recursive
    function.  But, in the case of MAPCAR and its relatives, we
    may have a complex function to evaluate with each application.
    Vectorization thrives on simpler functions that can be implemented
    in hardware.  In the case of recursion, the flow of control
    is unpredictable.  We can't see in advance where the recursion
    will stop.

Page faulting in LISP is probably a major issue here.  It's very
difficult (probably impossible, unless you have a radical compiler) to
keep paging down by the way you write code.  If you use lots of
macros, you can achieve a large degree of inline-coding, with the
disadvantage that you have to recompile large amounts of code with
each change.  (You could develop with DEFUNs, and deliver with
DEFSUBSTs or DEFMACROs to get around this.  Of course, this is less
space-efficient...)  My understanding of vectorization isn't all that
great, but from what I know of it, it seems that MAPCAR-class
primitives would be the most obvious uses.  Perhaps the ZetaLisp LOOP
macro (maybe some invocations of LOOPing on lists can use MAPCAR) and
DO constructs could use vectorization, although one has to be careful
that a given iteration doesn't depend on ANY side-effects of previous
iterations.  For this sort of reason, I would expect that recursion
would be right out of there for vectorization.  I don't use much
recursion since the general features of ZetaLisp heavily discourage
it.  I suppose another good use would be things like BITBLT, which is
used for copying 2-D array portions around to other array portions.
(Lisp Machine displays depend heavily on this, so it is microcoded.)
Since strings are implemented as arrays of characters, copying them
around and capitalizing and things like that could be vectorized.
Hmm, maybe there are applications for vectorization.  Any other ideas?

Random

------------------------------

End of PARSYM Digest
********************

∂15-Nov-85  0918	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 15 Nov 85  09:18:50 PST
Date: Fri 15 Nov 85 09:07:44-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA
cc: library@SU-SCORE.ARPA, maslin@SUMEX-AIM.ARPA
Message-ID: <12159474389.20.RICHARDSON@SU-SCORE.ARPA>


There will NOT be a CSD lunch on Tuesday, November 19. The weekly lunch
will resume on Tuesday, November 26.
-------

∂15-Nov-85  0921	JJW  	MJH-LispM news
To:   MJH-LispM@SU-AI.ARPA  
I've created a mailing list for broadcasting information to users of the
Symbolics Lisp Machines in MJH.  To mail to this list, use the address
MJH-LispM@SAIL.  There is a similar list called KSL-LispM for users of
the machines owned by KSL at Welch Road.  Some people may want to be on
both of these lists.

Here's some recent news about our Lisp Machines, which most of you no
doubt already know.

1. We have a fifth machine in the building, a 3640, and are scheduled to
   get a sixth one soon.  The new machine is called "Frivolous".  It is
   currently in MJH 433, the 4th floor machine room, but will be moved to
   the Formal Reasoning group's area in the near future.  Because of its
   small disk, file space on Frivolous is fairly limited.

2. You can send hardcopy text output to Boise from any of the MJH 3600s by
   including the form (load "C:>weening>boise.bin") in your init file.  It
   is no longer necessary to set the default printer individually; it has
   been made a default for everyone in the namespace data for the site.

   Rich Acuff at KSL is working on Imagen support and we will be able to
   get a copy of that when it is available.  Dover printing is pretty much
   out of the question for now, because it uses the PUP protocol.

3. The "secure subnets" list in the namespace has been fixed so that you
   should now be able to open TCP/FTP connections from any Stanford machine
   to any of the MJH (or KSL) Lisp Machines.

∂16-Nov-85  0750	ullman@su-aimvax.arpa 	Re:  another paper    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  07:50:49 PST
Received: by su-aimvax.arpa with Sendmail; Fri, 15 Nov 85 20:00:48 pst
Date: Fri, 15 Nov 85 20:00:48 pst
From: Jeff Ullman <ullman@diablo>
Subject: Re:  another paper
To: maier%oregon-grad.csnet@CSNET-RELAY.ARPA, nail@diablo

Well I meant three vowels in a row NOT ALL DISTINCT.
Sorry you missed that point David.

∂16-Nov-85  0751	ark@SALLY.UTEXAS.EDU 	Re:  another paper
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  07:51:47 PST
Received: from sally.UTEXAS.EDU by su-aimvax.arpa with Sendmail; Fri, 15 Nov 85 21:41:57 pst
Date: Fri, 15 Nov 85 23:35:42 cst
From: ark@SALLY.UTEXAS.EDU (Arthur M. Keller)
Posted-Date: Fri, 15 Nov 85 23:35:42 cst
Message-Id: <8511160535.AA03224@sally.UTEXAS.EDU>
Received: by sally.UTEXAS.EDU (4.22/4.22)
	id AA03224; Fri, 15 Nov 85 23:35:42 cst
To: nail@SU-AIMVAX.ARPA
Subject: Re:  another paper

This exchange reminds me of the lesson Jeff Ullman once tried to teach me
about doing research.  Come up with a statement you think is true.  Try
to generate counterexamples while you try to prove it.  Ask your friends
for counterexamples.  When you find a counterexample, extend your hypothesis
to exclude the counterexamples.  It's not so often that you can actually
watch Jeff in action using this very technique, but those of us on this
mailing list are being treated to a view of Jeff in action.  I propose
to assist the progress in this problem by doing research.  That is, I will
use the best source of last names that I have handy, the Greater Austin
Residence White Pages a.k.a. the Phone Book.  Perhaps someone like Mike
Lesk could do this using some online version of the White Pages, after
all he's at Bellcore.  The first page has two examples (adjacent in fact)
of the case David Maier reported: Abijaoude and Abijaoudi.  A sequential
search through the phone book would be quite lengthy, so I used one
important techique, variation on an existing counterexample.  So I looked
up Meier.  There were 31 entries, plus one for Meiers.  Particularly
appropriate was the address for one Paul Meier: 11504 Ruffled Grouse.
On the other hand, reminding me of the sign I saw in Tel Aviv in 1982
("Ullman Shoes") was the address for J B Meier: 10410 Broken Shoe Trail.

If the next problem will be to find three adjacent vowels, with two
adjacent ones matching, let me report this entry on page 313 (a palindrome):
Massoud MINOOI.

To close, I present the following name for your inspection: COUIE Edward D.

Arthur

∂16-Nov-85  0752	maier%oregon-grad.csnet@CSNET-RELAY.ARPA 	Re:  another paper
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  07:52:35 PST
Received: from CSNET-RELAY.ARPA by su-aimvax.arpa with Sendmail; Fri, 15 Nov 85 18:17:40 pst
Received: from oregon-grad by csnet-relay.csnet id aa04973; 15 Nov 85 20:59 EST
Received: by ogcvax.ogc.uucp (4.12/OGC←1.2)
	id AA07083; Fri, 15 Nov 85 17:05:59 pst
Date: Fri, 15 Nov 85 17:05:59 pst
From: "Prof. David Maier" <maier%oregon-grad.csnet@CSNET-RELAY.ARPA>
Message-Id: <8511160105.AA07083@ogcvax.ogc.uucp>
To: nail@su-aimvax.ARPA
Subject: Re:  another paper


Ahem

david mAIEr

∂16-Nov-85  0757	ACUFF@SUMEX-AIM.ARPA 	Re: MJH-LispM news
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  07:57:32 PST
Date: Fri 15 Nov 85 12:41:49-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Re: MJH-LispM news
To: JJW@SU-AI.ARPA, MJH-LispM@SU-AI.ARPA
In-Reply-To: Message from "Joe Weening <JJW@SU-AI.ARPA>" of Fri 15 Nov 85 09:21:00-PST
Message-ID: <12159513361.70.ACUFF@SUMEX-AIM.ARPA>

   KSL-LispM is directed to ALL users of KSL Symbolics or TI lisp machines,
not just Welch Rd.  There is a list called Whelan-LispM-Users which does
that.

	-- Rich
-------

∂16-Nov-85  0759	NEALE@SU-CSLI.ARPA  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  07:59:04 PST
Date: Fri, 15 Nov 1985  14:33 PST
Message-ID: <NEALE.12159533770.BABYL@SU-CSLI.ARPA>
From: NEALE@SU-CSLI.ARPA
To:   folks@SU-CSLI.ARPA


On Tuesday Nov. 19, MENTAL LIMOUSINE (Stephen Neale, Sami Shaio,
Susana Wessling &c.) will be performing at

		The VIS club, 628 Divisadero, S.F.


For just $2 you get to see THE VISIT (Another Stanford all original
band), BURNING WITCHES (a feminist thrash-punk band---really wild), as
well as MENTAL LIMOUSINE playing their own brand of Indo-European
dirge jazz-punk. Come along, have a ball. (Performance starts at 10:00 pm).

∂16-Nov-85  0800	ACUFF@SUMEX-AIM.ARPA 	KSL-3640-7 down   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  08:00:36 PST
Date: Fri 15 Nov 85 16:14:37-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: KSL-3640-7 down
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12159552100.79.ACUFF@SUMEX-AIM.ARPA>

   47 has popped it's tube, and will probably not be worked on until
Monday.

	-- Rich
-------

∂16-Nov-85  0759	RICHARDSON@SU-SCORE.ARPA 	General Faculty Meeting 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  07:59:17 PST
Date: Fri 15 Nov 85 11:08:01-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: General Faculty Meeting
To: faculty@SU-SCORE.ARPA
Message-ID: <12159496287.20.RICHARDSON@SU-SCORE.ARPA>


There will be a general faculty meeting on Tuesday, December 10 at 1:30 pm.
Please send any items that you would like to be included on the agenda to
Richardson@score.

-------

∂16-Nov-85  0801	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #33  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  08:01:42 PST
Date: 15 Nov 85 2009-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #33
To: PARSYM@SUMEX-AIM.ARPA

PARSYM Digest           Saturday, 16 Nov 1985      Volume 1 : Issue 33

Today's Topics:

    Third PARSYM Survey: Hardware for Parallel Symbolic Computing

----------------------------------------------------------------------

Date: Fri, 15 Nov 1985  20:01 PST
From: DAVIES@SUMEX-AIM.ARPA
Subject: Third PARSYM Survey

               Hardware for Parallel Symbolic Computing

The Third PARSYM Survey is designed to inform the PARSYM readership
about computer hardware systems that are being used or developed for
parallel symbolic computing.

Some research groups are using multiprocessor hardware -- obtained
commercially, acquired under special arrangement with the manufacturer,
or developed locally.  Others are simulating multiprocessing on
uniprocessors, such as Lisp machines.  Still others are connecting
multiple uniprocessors using LANs or specialized communication hardware.
Let us know what category you fall into -- or if I need to add more
categories -- and what the general nature of your hardware setup is.

There is an intentional ambiguity in this survey between hardware to
support research and hardware as the end result of research.  Since
the field and the available hardware are so new, it's hard to draw the
line between "commercial" and "experimental" hardware.  We would like
to hear about both.

Please provide a brief summary of your project, in case others don't
know about it.  If you've already described your project to PARSYM,
you'll be forgiven if you don't repeat yourself.  If you wish to
describe your software environment in order to clarify the
capabilities of your hardware, please do.  You may, however, want to
save the software details for a later survey, which will specifically
address software environments.

When I began putting this survey together, I thought it would be easy
to answer, compared to the last one on data abstractions, which was
rather technical.  Now this one looks rather daunting, but I hope that
won't discourage potential respondents from telling us about their
hardware.  If you don't have time to answer in full, please spend a
few minutes to fill in the first part of the survey.

Thanks go to Jay Glicksman (AI&DS), who wrote much of the
questionnaire, and Bruce Delagi (Stanford), who offered useful
comments.
!

	       The Third PARSYM Survey: Computer Hardware

                (Reply to PARSYM-Survey@SUMEX-AIM.ARPA)

			      Part I (Easy)

What computer hardware are you using or developing?

What are its components (processors, memory) and how many; how do the
processors and memory communicate?

SIMD/MIMD?  Shared memory or distributed memory?

What model of concurrency are you investigating with this system?

What would you like to see in your next hardware system?




		  Part II (Optional, for extra credit)


0. Project summary:

What model of concurrency (e.g., dataflow, communicating sequential
processes, actors, futures, etc.) is your work based on?  What
language, if any, are you using in your work?


1. Architectural classification:

SIMD or MIMD?


2. Processing elements:

Describe each processor in the system in terms of type (e.g., 68000
equivalent, Lisp machine, custom), "power" (e.g., in MIPS or relative
to a Vax or 68000), and memory size.  If the processors are simulated
then also describe the uniprocessor and mechanism for simulation.  If
the processors in the system are not of uniform type then describe
each type and their relationship to each other.

How many processors are there in the minimum and maximum configurations?
How many are you currently using?


3. Memory system:

Describe how memory is distributed and accessed in the system.  Shared
memory or distributed memory?  Is there a single address space or does
each processor have its own address space?  What sort of caching is
employed, if any: snoopy, write-through, or something else?  What
sizes are the caches?

What are the minimum and maximum amounts of memory?  How much are you
currently working with?


4. Communication system:

How are processors connected to each other and to memory?  Describe in
terms of topology (e.g., tree, systolic array, pipe, grid), switching
type (e.g., LAN, bus, Butterfly switch), and throughput (e.g., bus
transmission rates, time to access memory on remote nodes, time to
broadcast a message to all nodes).


5. Synchronization:

What hardware mechanisms does the system supply to support
synchronization?


6. Pragmatics

Is your system geared towards either specific programming languages or
applications?

Is the hardware of your system generally available or, if custom-made,
likely to be available in future?

If you're now using a multiprocessor system, how does the architecture
of your development system correspond to your research target
architecture?

What are the advantages of your hardware setup, particularly for the
problems you are investigating?  What are the disadvantages?  Is there
a system that would better meet your needs?

------------------------------

End of PARSYM Digest
********************

∂16-Nov-85  0805	PATASHNIK@SU-SUSHI.ARPA 	Upcoming AFLB canceled   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  08:05:45 PST
Date: Fri 15 Nov 85 12:02:50-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Upcoming AFLB canceled
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12159506266.7.PATASHNIK@SU-SUSHI.ARPA>

It turns out that Ketan Mulmuley is giving the same talk at BATS on
Friday, Nov 22 that he was to give at the upcoming (Nov 21) AFLB.
Consequently his AFLB talk is canceled, and unless another speaker
volunteers there will be no AFLB on the 21st.
	--Oren
-------

∂16-Nov-85  0808	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	CS UG program update  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  08:08:22 PST
Received: from su-aimvax.arpa by SU-SCORE.ARPA with TCP; Fri 15 Nov 85 20:26:50-PST
Received: by su-aimvax.arpa with Sendmail; Fri, 15 Nov 85 20:26:26 pst
Date: Fri, 15 Nov 85 20:26:26 pst
From: Jeff Ullman <ullman@diablo>
Subject: CS UG program update
To: faculty@score

The Engineering undergraduate council approved the CSUG program,
except for the detail of the exit requirement, which we have
been informed must be considered by the Committee on Undergraduate
Studies.  In the process of negotiation, several changes have been
made, and I would like to outline these.

1. The UGC agreed to allow CS to trade 5 units of Science for 5
of Math.  Our new program has a required 12 units of science,
and 28 of math, 10 "engineering basics" (CS 106 + E 40 = electronics),
and 44 units of CS.  The total, 94 units, is 8 fewer than the
proposal of September.

2. After a hard struggle, they approved the use of Psych 102, 106, and 108
as science options.

3. We dropped the 6 units of "CS electives" from the previous proposal.
Some of the courses were transferred to "Math electives", below.

4. After a close vote, the committee decided to drop the Math 120
(algebra) requirement, but put it in a group of math electives designed
to beef up our math requirement, and thereby get the science + math
total up to >= 38.  Students are required to choose two from:
Math 44, Math 120 or 120S, Stat 110 or 116, and CS 135,237A, and 260.
The UGC suggested that it would approve OR 151 (which was formerly
a "CS elective" as a math elective, and I assume the CSD curriculum
committee will add that to the list.


The faculty senate has  decided that it should approve our
program, so we shall have to bring it before the CUS and then the
senate.

The UGC has also endorsed our intent to develop a "numerical linear
algebra course, and there was indication that if we offered it,
the course would be required by other departments as well.
				---Jeff

∂16-Nov-85  1209	BSCOTT@SU-SCORE.ARPA 	Locks   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 16 Nov 85  12:09:05 PST
Date: Sat 16 Nov 85 11:47:43-PST
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: Locks
To: CSD@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12159765658.17.BSCOTT@SU-SCORE.ARPA>

If you have an office in MJH, be sure your door is actually locked when you
leave.  When the locks were changed a few of the locking mechanisms may have
been inadvertently adjusted so that when the doors are closed they are not
locked.

Betty
-------

∂17-Nov-85  0623	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #34  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 17 Nov 85  06:23:01 PST
Date: 16 Nov 85 1340-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #34
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Saturday, 16 Nov 1985      Volume 1 : Issue 34

Today's Topics:

                         Actors ?= Smalltalk
                   A PARSYM challenge: MIMD vs SIMD

----------------------------------------------------------------------

Date: 15 Nov 85 (Fri) 11:57:08 EST
From: Keiji Kanazawa <kgk%brown.csnet@CSNET-RELAY.ARPA>
Subject: Actors ?= Smalltalk


> With all due respect to those involved, isn't the whole Actors
> paradigm essentially at one with object-oriented paradigms a la
> Smalltalk, which have been around for quite some time now?

Scheme and FP are both applicative programming languages (after a
fashion).  Nice...  But I don't see that you'd want to say that they're
essentially at one.

If there is a textbook 20 years from now, and it classified languages as
'declarative', 'applicative', 'object-oriented', yes, it may well put
Smalltalk together with Prelude/Act-1/2/3.  I think few people would
disagree; it's more or less a trivial point.

But for that very reason, it's hard for me to think of Actors any
longer as anything but a parallel computation model.  Who cares if
it's object oriented?

It seems implied ("which have been around for quite some time now")
that Actors came long after Smalltalk did.  This is not the case.  I
think original work on PLASMA came as early as 1975.

> A careful reading of the postings of Karl and Jim points out that they
> are, in fact, comparing object-oriented *paradigms* with
> message-passing *implementations*.

Importantly, parallelism has been embodied in the Actor model from
those very beginnings.  The two are inseparable; it's not a question of
implementation.  It had to do with the idea that a Society of Agents
could cooperate to solve problems.

Finally, the Actor model seems broader in scope than other object
oriented languages.  It's conceivable that a large Actor system,
encompassing, let's say, the whole state of California, would include
a Smalltalk workstation as one of its many actors.  The reverse
situation is not addressed in the vanilla Smalltalk model.
Furthermore, such a system illustrates another point - that the Actor
universe is one in which, potentially, at any particular point in
time, the whole state of the world is not completely known or
knowable.  This may not be a solved problem for Actors; but it is not
addressed at all by Smalltalk, and it isn't about implementation.


Keiji Kanazawa

------------------------------

Subject: A parsym challenge: MIMD vs SIMD
Date: 16 Nov 85 11:02:48 EST (Sat)
From: Bradley C.  Kuszmaul <bradley@THINK.COM>

This submission argues that SIMD is just as good as MIMD and concludes
with a challenge to the parsym readership.

I propose the following hypothesis about parallel computing:
 `All real applications for parallel computing can run as well on a
  SIMD machine as on a MIMD machine'.

Now before you say `like wow, this guy is crazy', let me finish...

The following discussion assumes that you know what I mean by MIMD,
SIMD, and how MIMD and SIMD machines are usually programmed.  I will
define my terms a little more carefully later in this document.

It is clear that in some sense MIMD machines are more powerful than
SIMD machines, since a MIMD machine can simulate a SIMD machine by
simply having every processor do the same thing.  However, if we
accept my hypothesis, then it might very well be the case that we
would rather build SIMD machines for a variety of reasons (such as
cost and the difficulty of designing the machines).

		      The Argument Against SIMD

There are several programming languages which, as far as I know, are
relatively difficult to implement on SIMD machines because the
semantics of the language requires that each processor be doing a
different thing at the same time.  One example of this would be a
functional programming language (e.g.  pure lisp), in which case every
processor might be responsible for evaluating one subexpression, and
depending on the implementation, you might see and-parallelism,
or-parallelism or both.  For example, when evaluating the expression

  (+ (F 1) (G 2))

it seems reasonable to evaluate (F 1) and (G 2) in parallel on
different processors since the addition operator is strict; it will
need both of its arguments in order to finish evaluating.  (When
exploiting parallelism in the IF statement we must be more careful
because one of the `arms' of the IF won't be used.)  The problem here
is that the evaluation of (F 1) and (G 2) may be totally different
(for example if (F 1) involves lots of floating point operations while
(G 2) involves lots of BIGNUM operations).  This situation clearly
calls for a MIMD machine, so that one processor can be doing a
floating point operation while another does a bignum operation.

		      The Argument Against MIMD

Now that I have presented what I view as the fundamental argument
against my hypothesis, let me explain why the hypothesis might still
be true.  The hypothesis calls for `real applications'.  A programming
language is not a real application.  If someone implements a simulator
for systolic arrays written in pure lisp, then the full power of the
pure lisp implementation is being wasted, since the systolic array
could have been implemented efficiently on a SIMD machine.  Thus the
hypothesis is that there are no actual applications that someone
really needs which can not be implemented efficiently on a SIMD
machine.  The standard application for MIMD machines is simulating
other MIMD machines, which is circular, and I don't buy.

	    Examples of Applications for which SIMD Wins

Example 1: Finite element analysis.  In this case there are a set of
elements connected in a graph.  The analysis involves each element
communicating with its neighbors, and then performing some state
change based on that communications.  The communication and state
change program which is run on each element is almost independent of
which element is being worked on.  In a SIMD machine every processor
would simulate one of the finite elements, and all the processors
would be kept busy doing useful work all the time (and we can thus
conclude that the implementation is efficient).

Example 2: Finding a minimal PLA for a given set of boolean equations.
The implemention for a SIMD machine is that every processor simulates
one possible PLA.  The program to simulate the PLA is the same for
every PLA (since the description of the PLA can be treated as data).
It is possible to object to this example on the grounds that the
description of the PLA is part of the program to simulate the PLA, and
that we are describing a MIMD machine because each processor simulates
one PLA.

		     Post Discussion Definitions

So now for the interesting question of what I mean by SIMD and MIMD.
There is a spectrum of possible machines.  The problem in defining our
terms is that I could argue that a network of M68000's clocked by a
common signal is SIMD:

   There are N processors, each of which has some local data, some of
   which tells the program how to run.  Each processor understands
   exactly one instruction, the EVAL instruction which interprets one
   opcode of the local data.  Since there is only one instruction we can
   represent the instruction with no bits, and we need only to distribute
   a common clock to all of the processors.

My claim is that the distinction between SIMD and MIMD is subjective,
and the difference is how you write programs for them.

You are in MIMD mode if you write a program which runs on a single
processor which communicates with other processors.

You are in SIMD mode if you write a program which is to be executed on
all processors at the same time.

These distinctions are still subjective...  oh well.

			     A Challenge

Can anyone propose real applications which can not be implemented
efficiently on a SIMD machine?

Please send entries to parsym to be posted and to me (see my address
below) so I can have a chance to argue that it really can be
implemented efficiently.

Please include a description of the application and a strategy for
efficient implemention on a MIMD machine, and if you want you can
include an explanation of why you believe that it can not be
efficiently implemented on a SIMD machine.

In order to set the ground rules a little more firmly, I allow you to
use any MIMD machine but require that your application have an
unbounded amount of parallelism (as a function of the size of the
problem).  I will use a machine like the connection machine as my
canonical example of a SIMD machine.  If I can keep my connection
machine processors busy `most of the time' doing useful computation
towards solving your problem than I will claim that the application is
an instance which supports my hypothesis.

Let's see those applications come pouring in...
 -Brad

 bradley@think.arpa
   or
 bradley@think.com.arpa
   or
 bradley@think.uucp

------------------------------

End of PARSYM Digest
********************

∂17-Nov-85  1850	ullman@su-aimvax.arpa 	Re:  another paper    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Nov 85  18:50:28 PST
Received: by su-aimvax.arpa with Sendmail; Sun, 17 Nov 85 14:51:17 pst
Date: Sun, 17 Nov 85 14:51:17 pst
From: Jeff Ullman <ullman@diablo>
Subject: Re:  another paper
To: ark@SALLY.UTEXAS.EDU, nail@diablo

1. I *think* Papadimitriou is an anomaly because in the original
Greek, it does not have three vowels in a row (I may be wrong sbout
this--Christos??) The same probably applies to Abijaoude, where
the ou is a transliteration of one vowel.
As for Couie, I meant *exactly* three vowels in a row, not
three or more.  Sorry you missed this point, Arthur.
2.

∂17-Nov-85  1851	LANSKY@SRI-AI.ARPA 	PLANLUNCH reminder -- Jeff Finger, 11am 
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 17 Nov 85  18:51:24 PST
Date: Sun 17 Nov 85 18:49:03-PST
From: LANSKY@SRI-AI.ARPA
Subject: PLANLUNCH reminder -- Jeff Finger, 11am
To: planlunch-reminder.dis: ;

	EXPLOITATION OF CONSTRAINTS IN DEDUCTIVE DESIGN SYNTHESIS

			    Jeff Finger (JFINGER@SUSHI)
		        Stanford University
				
	 	    11:00 AM, MONDAY, November 18
       SRI International, Building E, Room EJ228 (new conference room)

The talk will cover two related topics in deductive design synthesis: 
(1) efficiency gained by reasoning forward from subgoals, and 
(2) advantages and disadvantages of using a declarative representation 
    for partially completed designs.

The first part of the talk gives the deductive framework for capturing the
following intuition:

	Suppose I have decided that X and Y and to be true of my
	design.  Perhaps I should think about what else X and Y imply
	about the design, say Z.  Otherwise, I might waste time
	trying to complete the design process by making decisions
	that have *already* been ruled out by X and Y, for example,
	NOT(Z).

The conditions that X and Y imply (called "necessary constraints" or "NC's")
are found via reasoning forward from subgoals. We show how NC's of a
subgoal can be used to prune the design space either by preventing some
impossible possibilities from ever being generated or by providing a quick
means of filtering bad choices.  In terms of resolution, the above use of
NC's corresponds to the rather counterintuitive notion of allowing
OR-INTRODUCTION on clauses in the set of support. We will also discuss
inheritance of NC's from goal to subgoal and the relation of finding NC's to
that of checking consistency of partially completed designs.

The second part of the talk deals with declarative representation of
partially completed designs. Deductive design systems such as QA3 or Manna
and Waldinger's reify the design as a single term in the logic.
However, it is difficult to express many sorts of constraints on partially
completed designs as a single term. Examples include two actions in an
unspecified order, or the constraint that Action A takes place less than 3
seconds or more than 8 seconds after Action B. We present a system called
RESIDUE in which we build up the design as a set of facts we are willing to
assume about of the design.  Using facts rather than a single term, we can
make finer-grained decisions, avoiding unwitting commitments that might
result in unnecessary backtracking.  In addition, forward reasoning on
subgoals (as in the first portion of the talk) may be done directly on the
set of facts. 
-------
-------

∂17-Nov-85  1858	BRESNAN@SU-CSLI.ARPA 	Morphology, Syntax, Discourse Interactions 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Nov 85  18:58:05 PST
Date: Sun 17 Nov 85 17:26:44-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: Morphology, Syntax, Discourse Interactions
To: folks@SU-CSLI.ARPA

←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
This Thursday, November 21, Carolyn Coleman will present her work
on Reflexives in Kunparlang in the CSLI conference room at 10:00
a.m.  Everyone interested in lexical rules, lexical morphology,
and the semantics of reflexive pronouns is invited.

            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
             ``Morphological Structure of Kunparlang Verbs''

      Kunparlang verbs are extremely complex morphologically.  They
   cross- reference Subject and Object functions, incorporate nominal
   roots, use `applicative' derivational morphology, carry modal,
   directional and aspectual affixes, and inflect for Tense and Mood.
   There are two levels of hierarchical morphological structure,
      (i) The stem, which carries all morphology having compositional
          semantics.
     (ii) The lexical base, which caries all semantically idiosyncratic
          morphology.
      Kunparlang verbs undergo two types of reflexive operation which
   have a partially complementary distribution and which have different
   semantic effects on the verbs to which they apply.  With the first
   reflexive operation the reflexive subject is always an Actor; with the
   second the reflexive subject may be an Undergoer.  The second reflexive
   operation has a range of meanings which match those of 
   mediopassive constructions in other languages as well as the 
   reflexive reading.  Both reflexivizing operations are derivations that
   apply at the level of the lexical base; given that they have the same
   morphological status, there is a problem of how to semantically
   characterise them in a manner that will clearly show the semantic
   similarities and differences between them.		--Carolyn Coleman
-------

∂17-Nov-85  1859	ACUFF@SUMEX-AIM.ARPA 	Bug Reporting from Explorers
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 17 Nov 85  18:59:25 PST
Date: Sun 17 Nov 85 16:30:36-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Bug Reporting from Explorers
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12160079299.63.ACUFF@SUMEX-AIM.ARPA>

  If you'd like to report a bug on an Explorer and include a stack
trace in the report (generally a good idea if you can get one), you
can do the following: (1) use C-M from the debugger to get into Zmacs
with lots of debugging info already put into the buffer, (2) fill in
your comments, (3) write the file to your favorite mail host (eg. Sumex
or Diablo), (4) login on the host you wrote the bug report to and send
it to me.  This is necessary because the Explorers don't speak TCP
SMTP (as yet) or PUP MTP, and so can't send mail directly to any
Stanford hosts.

	-- Rich
-------

∂18-Nov-85  1012	WECHSLER@SU-CSLI.ARPA 	XenoneX
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Nov 85  10:12:12 PST
Date: Mon 18 Nov 85 10:03:10-PST
From: Stephen Wechsler <WECHSLER@SU-CSLI.ARPA>
Subject: XenoneX
To: folks@SU-CSLI.ARPA

The techno-formula music group XENOFILE (Steve Wechsler,
Linguistics Dept.; Elizabeth Dykstra, X-ist theorist;
Jim Shearer, Mathematics, UCB; and others)
will team up with the NYC-based post-modern 
dance troupe PALINDROME to present excerpts from
their new collaborative work "XENONEX".  Coordinates:

	Time = Saturday, Nov. 23rd at 10:00 pm

	Space = THE LAB (1805 Divisadero at Bush, S.F.)

A party till 3 am follows (w/ DJ Raoul!); a $5 cover will benefit
THE LAB, "a cooperative laboratory for the arts."




-------

∂18-Nov-85  1018	INGRID@SU-CSLI.ARPA 	Symposium on AI    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Nov 85  10:18:51 PST
Date: Mon 18 Nov 85 10:08:49-PST
From: Ingrid Deiwiks <INGRID@SU-CSLI.ARPA>
Subject: Symposium on AI
To: SU-Bboards@SU-CSLI.ARPA, Folks@SU-CSLI.ARPA


	 	          ANNOUNCEMENT


             Soto Symposia on Artificial Intelligence
             ----------------------------------------

A series of three symposia on AI will be held in the Soto Lounge,
Wilbur Hall, the first three Mondays in November, at 7 p.m.  These
symposia are intended to inform Stanford students and others who are
interested about the nature of AI, its prospects, and the intellectual
and social issues to which it gives rise.

The first two symposia were held during the last couple of Mondays.
The third one is scheduled for today:


Monday, November 18, 7 p.m.

	"Star Wars and Computation"

	John McCarthy     Professor of Computer Science, Stanford

	David Redell      DEC Research Laboratory

	and possibly others

This symposium will begin with short talks by the speakers, followed
by plenty of time for vigorous interchanges among the speakers and
with members of the audience.

Soto is the first dormitory in the Wilbur complex that one comes to
when walking out Escondido from the Undergraduate Library.

	
----------------------------------------------------------------------
<--UGLI		              Escondido  Rd.
----------------------------------------------------------------------
                                       
xxxxxxxxxxxxxx                              X XXX     XXX X 
		                   Soto --> X   Wilbur    X		
Stern Hall                                  X             X

xxxxxxxxxxxxxx                              X    Hall     X
                                            X             X
                                            X XXX     XXX X




-------

∂19-Nov-85  0827	@SU-SUSHI.ARPA:RUTENBURG@SU-SCORE.ARPA 	ICALP
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  08:27:19 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Mon 18 Nov 85 17:51:17-PST
Date: Mon 18 Nov 85 17:51:19-PST
From: Vladislav Rutenburg <RUTENBURG@SU-SCORE.ARPA>
Subject: ICALP
To: aflb.local@SU-SCORE.ARPA
Message-ID: <12160356138.22.RUTENBURG@SU-SCORE.ARPA>

Does anybody have the call for papers for the ICALP in France in July or
know who is on the organizing committee and the phone numbers to contact them?
we seem to have missed the deadline.
-------

∂19-Nov-85  0843	@SU-CSLI.ARPA:BrianSmith.pa@Xerox.ARPA 	A day's mail lost   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  08:43:17 PST
Received: from Xerox.ARPA by SU-CSLI.ARPA with TCP; Tue 19 Nov 85 08:33:24-PST
Received: from Cabernet.ms by ArpaGateway.ms ; 19 NOV 85 08:38:07 PST
Date: 19 Nov 85 08:26 PST
From: BrianSmith.pa@Xerox.ARPA
Subject: A day's mail lost
To: ISL↑.pa@Xerox.ARPA, Folks@SU-CSLI.ARPA, Wendy Hansen
 <P.PURPLE@[36.48.0.5]>, lamping@Su-sushi.ARPA, Berlin@MIT-XX.ARPA,
 Borning@Washington.ARPA, Estrin@MIT-XX.ARPA, Gould.pa@Xerox.ARPA,
 Ornstein.pa@Xerox.ARPA, Suchman.pa@Xerox.ARPA, TW@SU-AI.ARPA,
 Zilles.IBM-SJ@CSNet-Relay.ARPA, gnelson@DECWRL.DEC.COM,
 redell@DECWRL.DEC.COM, PARC-CSLI!chapman@parcvax.ARPA
cc: BrianSmith.pa@Xerox.ARPA
Reply-to: BrianSmith.pa@Xerox.ARPA
Message-ID: <851119-083807-1465@Xerox>

Due to a software foul-up, I have lost all mail received since about
11:00 yesterday morning.  If you sent anything important, could you send
along another copy?

Many thanks

Brian

∂19-Nov-85  0845	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #35  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  08:45:05 PST
Date: 18 Nov 85 0912-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #35
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 18 Nov 1985       Volume 1 : Issue 35

Today's Topics:

                      SIMD vs. MIMD (2 messages)
                         Smalltalk vs. Actors
                     Survey on Data Abstractions
                       Seminar: Multilisp (MIT)

[As you've probably noticed, PARSYM is very active at the moment.
There are three major threads: the recently initiated discussion of
SIMD vs. MIMD, the waning discussion of Smalltalk vs. Actors, and the
ongoing PARSYM surveys.  So as not to distract from the
reader-initiated discussions, the hardware survey responses are
temporarily being queued for a future issue.  One survey response has
been received so far (on Berkeley's Aquarius multiprocessor, from
fagin@dali.berkeley.edu) and others are eagerly awaited.  -- BD]

----------------------------------------------------------------------

Date: Saturday, 16 November 1985  20:45-PST
From: coraki!pratt at su-navajo.arpa (Vaughan Pratt)
Subject: SIMD vs. MIMD

I agree with Brad about SIMD vs. MIMD being a matter of level.  Abstracting
away to the n-th degree, as mathematicians are wont to do, a clean example
of this principle is the expression L**n, denoting the concatenation of
n copies of the language L with itself.  The SIMD interpretation of this
is that all n copies of L are the same language.  Going down a level and
looking at individual strings in L**n, we find that up to n different
strings have been chosen from the n copies of L, which looks more like
MIMD behavior.  (That may be *too* abstract for some: as executed
programs are fond of asking, who's in charge here?)

Another key observation Brad makes is that one machine's data is
another's instruction.  This too shows how dependent the SIMD/MIMD
distinction is on the level of abstraction.  MIMD can become SIMD when
the MI part turns into a piece of the MD part.

Finally I agree fully with Brad's claim that it is the programmer's
level that matters.  My own reason for taking this position is economic:
programs either written or executed (the distinction is not important
for this argument) by people cost a fortune compared to those written
or executed by machines.  If you have SIMD at one level and MIMD at
another, you will save a bundle by arranging things to put the
programmer and SIMD at the same level, on the premise that SIMD
software is cheaper than MIMD software.  (I'm happy to enter into a
debate on that premise if there's anyone out there that doubts it.)
-v

------------------------------

Date: Monday, 18 November 1985  07:36-PST
From: Sal Stolfo <sal at CS.COLUMBIA.EDU>
Subject: PARSYM Digest   V1 #34

SIMD machines do have certain severe limitations which make MIMD
processing the processing mode of choice.  SIMD machines (at the risk
of appearing overly general) require that the data set being processed
(and of course distributed to concurrent PE's) comprises data objects
of the same size, shape and extent.  Therefore, concurrent lock-step
instructions can be written with the assumption that the data being
manipulated at a PE is of the exact same type.

Here is a simple operation that blows this out of the water:

Distribute a set of arbitrary first order literals to a set of PE's,
one literal per PE.

Broadcast a single first order GOAL LITERAL to all PE's and execute
logical unification between the locally stored literal and the
broadcast literal.

 ------------

For a MIMD processor, this operation is straightforward and quite simple.

For an SIMD processor, it is much more problematic, not to mention an
horrific task to program in SIMD mode.

The problem arises when the first order terms in each locally stored
literal varies in type.  A logical variable appearing in the broadcast
literal may indeed have to be compared and bound to a constant,
another variable in the locally stored literal, or worse a first order
term which may include variables of course.

Think of the different tests and branches that would need to be issued
from the control processor to the SIMD processor to account for every
possible type of first order term encountered.

I believe it is quite possible to implement this operation in SIMD
mode, but I also strongly believe that the SIMD processor will not
outperform an MIMD processor.  Indeed, for certain anomolous cases we
can invent, the SIMD processor will probably be slower that a
sequential processor working on the same data set.

One other issue to be carefully studied here for this operation.  Most
of the SIMD processors I am aware of comprise a very limited memory at
each PE.  Logical unification has the nasty side effect of producing
resolvents which are "larger" than the two unified parent literals.

Consider unifying:

 P(x1,x2,...,xn)

 P(f(x0,x0),f(x1,x1),...,f(xn-1,xn-1))

where P is a first order predicate symbol,
      xi is a variable
      f is a binary skolem function

What result do you get if you use a "string or token" representation
for literals?  This is really nasty.

Of course, a list representation is necessary to unify and store the
result.  The SIMD processor therefore requires list representations,
and garbage collection.  It is safe to assume that the literals being
operated upon will of course span across several PE's.  This then
requires trading in concurrency (since multiple PE's are needed to
store a literal) to process the data set.

This operation is easier to implement on an MIMD processor where each
of the PE's has a considerably larger memory than the "typical"
proposed SIMD processor.  Implementation ease aside, an MIMD processor
is probably much faster at this game than an SIMD processor.  Don't
forget, in parallel processing SPEED IS THE NAME OF THE GAME.

Sal Stolfo

Stolfo@Columbia

------------------------------

Date: Sunday, 17 November 1985  08:00-PST
From: Steven H. Gutfreund <GUTFREUND@umass-cs.csnet>
Subject: Smalltalk vs. Actors

I have found that one should not look too closely at any of the
definitions that Language or System people provide.

I have a friend who when taking his SYSTEM qualifier, was asked what
was the definition of a PROCESS.  For his own amusement (and because he
is a very sharp system person) he proceeded to tell the qual committee
(it was an oral qual) that no such thing exists in reality.

Most people would say that a process is an independent thread of code
with its own state.  But he argued that this was so close to the idea
of a co-routine, or to a subroutine that maintains its state between
calls, that one can blur the lines and say that there really is not
clear distinct definition of what a process or task is.

BTW: the technique worked, he amused the committee, kept their minds
reeling and passed the quals, but then, they knew he knew his stuff.

						- steve

------------------------------

Date: Saturday, 16 November 1985  22:07-PST
From: coraki!pratt at su-navajo.arpa (Vaughan Pratt)
Subject: Data Abstractions (*very* late response)

(Sorry for the late response - things have been hectic of late.)

Yes, I have developed and am using abstractions specialized for
parallel computing.  The central data abstraction is the partially
ordered multiset (pomset), or partial string as some European authors
call it.  Here's the idea.

First consider the 5-symbol string LEVEL.  This may be considered to
be a 5 element multiset {E,E,L,L,V} together with a linear (total) order
putting the 5 elements in the order L-E-V-E-L.

Now relax the requirement that the order be total.  For example we may
choose to make one of the E's independent of V and the other E.  This
gives us (reading left to right)

				  E---V
				 /     \
				L-     -L
				  \   /
				   -E-

Terminology: in this *pomset* there are 5 *events* and 3 *actions* L,E,V.

Application: let the actions model any of:
	states
	state transitions
	messages
	programs
and let the events model instances of those actions (e.g. an instance
of a program is a job).  Then a pomset models a *computation* (or *trace*,
or *behavior*) of a system.

The role of the partial order is to specify *necessary temporal
precedence*.

Here's one nice application of this concept.  Suppose we want to model
a communication channel.  Consider the four events T0, T1, R0, R1
corresponding to transmitting a 0, transmitting a 1, receiving a 0, and
receiving a 1.  Suppose the 0 is transmitted before the 1, and the
channel is order-preserving.  Clearly we want the order

				T0---R0
				 |    |
				T1---R1

Where do we get it?  Here's an easy way.  Start with two strings 01 and
TR, corresponding respectively to a sequence of messages (0 followed by
1) and a sequence of transactions (transmission followed by receipt).
Form the *direct product* of these two strings, viewed as 2-element
partially ordered structures.  This is an algebraic construction
yielding a structure whose elements form the Cartesian product of the
elements of its arguments, and which orders (u,v)<=(u',v') just when
u<=u' and v<=v'.  A little thought shows that the above desired square
is exactly what you get when you form the direct product of 01 with TR.

Now this square is *not* a string, even though it was formed as the
(direct) product of two strings.  Yet it is just what we wanted.  The
situation is a little like that with complex numbers.  You take the
square root of a negative real and suddenly you're outside the domain
of real numbers.

How does all this relate to concurrency?  Well, it gives a trivial
definition for two events to be concurrent, namely when they are
incomparable.  Thus strings model purely sequential computation, whereas
other pomsets allow for concurrency.

Rather than get into more technical detail on this model (a bottomless
pit relative to the scope of this message), let me just point to some
papers on the subject.

Gischer, J., Partial Orders and the Axiomatic Theory of
Shuffle, Ph.D. Thesis, Computer Science Dept., Stanford University,
Dec. 1984.
	(contains many results about equational theories of pomsets
	under various combinations of axioms.  Of interest to people
	interested in logics of concurrency.)

Mazurkiewicz, Traces, Histories, Graphs: Instances of a Process
Monoid, Proc. Conf. on Mathematical Foundations of Computer Science,
Springer-Verlag Lecture Notes in Computer Science LNCS 176, 1984.
	(how several pomset-like models can be viewed as algebras
	with one binary associative operation)

Pratt, V.R., The Pomset Model of Parallel Processes: Unifying
the Temporal and the Spatial, Proc. CMU/SERC Workshop on Analysis
of Concurrency, Springer-Verlag Lecture Notes in Computer Science
LNCS 196, Pittsburgh, 1984.
	(older paper of mine, good mainly for background)

Pratt, V.R., Some Constructions for Order-Theoretic Models
of Concurrency, Proc. Conf. on Logics of Programs, Springer-Verlag
Lecture Notes in Computer Science LNCS 193, Brooklyn, 1985.
	(introduces orthocurrence, cleans up the utilization concept)
(Recently revised version: Modelling Concurrency with Partial
Orders, much more readable, I like it a lot, contact me for a copy.)

Winskel, G., Events in Computation, Ph.D. Thesis, CST-10-80,
Dept. of Computer Science, University of Edinburgh, 1980.
	(a heavily algebraic treatment of Petri Nets using a
	pomset-like model)

Winskel, G., A New Definition of Morphism on Petri Nets, Proc.
CMU/SERC Workshop on Analysis of Concurrency, Springer-Verlag Lecture
Notes in Computer Science LNCS 196, Pittsburgh, 1984.
	(gives one solution to the problem of what is a reasonable
	category of Petri nets)

Winskel, G., Categories of Models for Concurrency, Technical Report
no. 58, University of Cambridge, England, undated (rec'd Dec. 1984).
	(use of adjunctions between categories to capture
	"isomorphisms" between various models of concurrencies)

------------------------------

Date: Tue 12 Nov 85 12:45:02-EST
From: Brian C. Williams <WILLIAMS%MIT-OZ at MIT-MC.ARPA>
To:   AIList
Re:   Seminar - Multilisp (MIT)

[Forwarded by Steven A. Swernofsky <SASW at MIT-MC.ARPA>]

Thursday , October 14  4:00pm  Room: NE43- 8th floor Playroom

                    The Artificial Intelligence Lab
                        Revolving Seminar Series

         "Multilisp:  A Language for Parallel Symbolic Computing"

                               Burt Halstead

                                 MIT, LCS


Multilisp is an extension of Scheme with additional operators and
additional semantics for parallel execution.  These have been added
without removing side effects from the language.  The principal
parallelism construct in Multilisp is the "future," which exhibits
some features of both eager and lazy evaluation.  Current work focuses
on making Multilisp a more humane programming environment, and on
expanding the power of Multilisp to express task scheduling policies.

A skeletal Multilisp has been implemented, and has been run on the
shared-memory Concert multiprocessor, using as many as eight
processors, as well as on a BBN Butterfly machine with as many as 128
processors.  The implementation uses interesting techniques for task
scheduling and garbage collection.  The task scheduler helps control
excessive resource utilization by means of an unfair scheduling
policy; the garbage collector uses a multiprocessor algorithm modeled
after the incremental garbage collector of Baker.

The talk will describe Multilisp, discuss the areas of current
activity, and indicate the future direction of the project.

------------------------------

End of PARSYM Digest
********************

∂19-Nov-85  0852	cheriton@su-pescadero.ARPA 	Re:  another paper    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  08:51:27 PST
Received: from su-pescadero.arpa by su-aimvax.arpa with Sendmail; Mon, 18 Nov 85 22:28:31 pst
Received: by su-pescadero.arpa with Sendmail; Mon, 18 Nov 85 22:28:26 pst
Date: Mon, 18 Nov 85 22:28:26 pst
From: David Cheriton <cheriton@su-pescadero.ARPA>
Subject: Re:  another paper
To: ark@SALLY.UTEXAS.EDU, nail@diablo, ullman@diablo

How about a real oddity.  Surnames with 4 consonants occurring in consecutive
alphabetical ordering (except for duplicates) with the consonants consecutive
except for one vowel!

∂19-Nov-85  0856	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	PALMS     
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  08:53:45 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 18 Nov 85 16:09:59-PST
Date: 18 Nov 85  1439 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: PALMS    
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    



                PAcific Logic MeetingS
The first of three scheduled for this year will meet at CalTech on
Saturday, Dec. 7.  The speakers will be Harvey Friedman, Alain
Louveau, Tony Martin, and Robert Solovay.  For more information
contact Matt Foreman, (818) 792-4677.


∂19-Nov-85  0925	RICHARDSON@SU-SCORE.ARPA 	Tuesday CSD Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  09:24:55 PST
Date: Tue 19 Nov 85 09:10:34-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Tuesday CSD Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12160523482.11.RICHARDSON@SU-SCORE.ARPA>


Just a reminder that there will NOT be a lunch today.
-------

∂19-Nov-85  0938	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  09:38:21 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 19 Nov 85 09:32:04 pst
Date: Tue, 19 Nov 85 09:32:04 pst
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

TOmorrow, Kathy is going to fill us in on the interface for
capture rules.  I'd like then to start writing some capture
rules in Prolog and also get going on the ICODE->Prolog
translator.  Volunteers?

The meeting starts at 11AM in 301.
				---Jeff

∂19-Nov-85  0939	LIBRARY@SU-SCORE.ARPA 	Socrates:  Additional Terminal In Math/CS Library   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  09:39:30 PST
Date: Tue 19 Nov 85 09:37:47-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates:  Additional Terminal In Math/CS Library
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12160528437.33.LIBRARY@SU-SCORE.ARPA>

We now have two public access terminals in the Math/CS Library for searching
Socrates.  These terminals also access the technical reports file in
Socrates which is now the only up-to-date index to this material as it
is added to our collection.

Harry Llull
-------

∂19-Nov-85  0944	NILSSON@SU-SCORE.ARPA 	Visiting Professorship
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  09:44:31 PST
Date: Tue 19 Nov 85 09:41:01-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Visiting Professorship
To: ac@SU-SCORE.ARPA
Message-ID: <12160529023.22.NILSSON@SU-SCORE.ARPA>

John McCarthy has suggested that we institute a regular "visiting
professorship" every year.  This will complement our "visiting
industrial professorship." The new visiting professor would ordinarily
be someone who is a professor at some other institution.
It seems to me that this is a good idea for CSD because it would 
allow us to benefit from new ideas (perhaps in areas that we are not
covering thoroughly) and it would also allow us to get acquainted with
people who might ultimately want to move to Stanford.  The "visiting
professorship" would not replace other visitors that our faculty from
time-to-time invite and support out of their own research funds.  Betty
Scott and I will look into funding matters on this, but I wanted first
to assure myself that there are no reasons why this isn't a good idea.
(I can't imagine any such reasons.)  May I presume that hearing no
objections to the contrary we could go ahead with this--providing
we can find the money?  -Nils
-------

∂19-Nov-85  0945	NEUMANN@SRI-CSL.ARPA 	RISKS-1.23   
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  09:45:04 PST
Date: Tue 19 Nov 85 07:59:48-PST
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSL.ARPA>
Subject: RISKS-1.23
Sender: NEUMANN@SRI-CSL.ARPA
To: RISKS-LIST@SRI-CSL.ARPA

RISKS-LIST: RISKS-FORUM Digest  Tuesday, 19 Nov 1985  Volume 1 : Issue 23 

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Expecting the unexpected (Peter G. Neumann)
  Safety Group Activities in the U.S. (Nancy Leveson)
  Automobile computer control systems susceptible to interference (Bennett Smith)
  Irresponsible computer "game"; BBS Legislation (Ted Shapin)
  SDI Debate at MIT (John L. Mills)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

Sorry for the hiatus.  Circumstances were beyond my control.  Thanks to 
Frieder von Henke for keeping it from being longer.  PGN

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------


Date: Mon 18 Nov 85 12:57:03-PST
From: Peter G. Neumann <Neumann@SRI-CSL.ARPA>
Subject: Expecting the unexpected
To: RISKS@SRI-CSL.ARPA

We have been talking about risks to the public in the use of computer
systems in this forum since 1 August, and trying to keep the discussions
mostly technical -- but not political or polemical.  On the other hand, I
have noted on various occasions that it is sometimes hard to exclude
nontechnical arguments.  This message may initially seem off the mark, but I
think it has some subliminal relevance that might help put into perspective
the importance of the underlying assumptions that we make and the need to
anticipate essentially all possible unusual circumstances.

On 13 October 1985, in Florence, Italy, at about 10:05 PM, European Standard
Time, my apparently very healthy 19-year-old son Chris -- never having had a
medical problem in his life -- suddenly had his heart stop and died within
moments.  Resuscitation attempts failed.  The autopsy found no discernible
apparent cause of death.  A neurophysiologist friend who joined me in
Florence suggested ventricular fibrillation as the most likely actual cause.
The heart muscles receive signals from distributed sources via independent
paths, and normally all of the signals arrive sufficiently synchronously to
trigger heart contraction.  Under fibrillation, the signals arrive
incoherently, and cannot be integrated sufficiently to trigger contraction.
So, why do I mention this in a RISKS Forum?  Well, even in an apparently
completely healthy person there exists some risk of spontaneous malfunction.
It seems to me that in making arguments about how hardware and software will
operate correctly or will continue to operate correctly if they once did, we
make all sorts of implicit underlying assumptions that may be true most of
the time, but that may indeed break down -- particularly in times of stress.
The nastiest problems of all seem to involve unanticipated conditions in
timing and sequencing, in both synchronous and ansynchronous systems, and
particularly in distributed systems.  We have seen various problems in the
past -- the ARPANET collapse (due to an accidentally propagated virus, after
years of successful operation) and the first shuttle launch immediately come
to mind as specific well-documented examples.  In some cases there is also
signal inference -- as in the pacemaker problems (see Nancy Leveson's
message in RISKS-1.22).  I think that in our lives as in computer systems,
we tend to make unjustified or oversimplified assumptions.  In life, this
makes it possible for us to go on living without inordinate worrying.
However, in computer systems, greater concern is often warranted.  So, my
point in introducing this message into the RISKS Forum is to urge us to try
to make our assumptions both more explicit and more realistic.  Basing a
system design on assumptions that are almost but not quite always true may
seem like a close approximation, but may imply the presence of enormous
unanticipated risks.  In systems such as those involved in Strategic
Defense, for example, every one of the potentially critical assumptions must
be sufficiently correct.

                 Peter G. Neumann (back again on the net)

------------------------------


Date: 09 Nov 85 13:04:51 PST (Sat)
From: Nancy Leveson <nancy@ICSD.UCI.EDU>
To: risks@sri-csl.arpa
Subject: Safety Group Activities in the U.S.

Udo wrote about the EWICS TC7 in a previous RISKS forum and said that
such a group died through lack of interest in the U.S.  Actually, there
is a similar group which has been active in the U.S. for about three
years.  It is called the Software System Safety Working Group and was
started by the Air Force although it is now a tri-service group.  Although
sponsored by the DoD, it is not limited to military applications and has
participants from other branches of the government and industry.  The latest
meeting was held in conjunction with a conference on computers in medicine.

Meetings are held approximately twice a year and usually have from 50-200
participants.  One of the products of the group is a Software Safety Handbook
which is meant to accompany the recent MIL-STD-882b (System Safety) update.  
The main purpose of the group has been to meet and discuss new techniques,
share experiences, exchange ideas, etc.  There is tentatively a meeting
planned for January in Washington D.C.  Anybody interested in the group should
contact me (nancy@uci.edu) or Al Friend (friend@nrl-css) who is with the Navy 
and is currently chairing the group.  A future plan is to have an on-line
safety database which will reside at the SRI NIC.  

Other activities in which I have been asked to participate and which might
be of interest to readers of this forum are a conference on safety
which will be held in Washington D.C. next July and a workshop on safety and
security sponsored by the Center for Software Reliability in England next
September.  I am also considering organizing a workshop in California on safety
which would be held right before the next International Conference on Software
Engineering in Spring 1987.  Anyone interested in more information on any
of these activities can again contact me and I will direct you to the
right people.
 
Nancy Leveson
University of California, Irvine

------------------------------

Date: Wed, 23 Oct 85 11:14:29 -0100
From: ircam!bks@seismo.CSS.GOV (Bennett Smith)
Message-Id: <8510231014.AA02863@ircam.UUCP>
To: NEUMANN@SRI-CSL.ARPA
Subject: Automobile computer control systems susceptible to interference

By chance I saw an article in an issue of the
"Journal of Environmental Engineers" (published in England, date of
issue about 10 months ago, I believe) about the sensitivity of
a microprocessor-controlled automobile control system to external
electromagnetic radiation.  As I recall, a CB transmitter near the car
could, at the right frequency, make the engine slow down or speed up.
Perhaps this article would interest some of your contributors.

Bennett Smith					
IRCAM
31, Rue Saint Merri
75004 Paris, France		{seismo,philabs,decvax}!mcvax!ircam

------------------------------

Date: Mon 18 Nov 85 11:54:52-PST
From: Ted Shapin <BEC.SHAPIN@USC-ECL.ARPA>
Subject: Irresponsible computer "game"
To: risks@SRI-CSL.ARPA, human-nets@RED.RUTGERS.EDU
Phone: (714)961-3393; Mail:Beckman Instruments, Inc.
Mail-addr: 2500 Harbor Blvd., X-11, Fullerton CA 92634

From a Toys R Us ad in the L.A. Times, 11/17/85:

Activision
HACKER
Makes you feel like you've unlocked someone else's computer system!
For C-64, C-128. $24.97.

[And on the package:]

TEMPTATION
HACKER
To stumble into somebody else's computer system.  To be someplace
you're really not supposed to be.  And to get the strange feeling
that it really does matter.  "LOGON PLEASE" is all you get to
start with. That's it.  From there, it's up to you.  If you're
clever enough and smart enough, you could discover a world you've
never before experienced on your computer. Very temptimg.
- - -

This "product" is socially irresponsible!  It leads young people
to think breaking into unknown systems is OK. The "world" they
discover may be the world of the penal system!

Ted Shapin

------------------------------

Date: Thu 14 Nov 85 10:34:50-PST
From: Ted Shapin <BEC.SHAPIN@USC-ECL.ARPA>
Subject: BBS Legislation
To: human-nets@RED.RUTGERS.EDU, risks@SRI-CSL.ARPA

I have remailed several messages on pending BBS legislation to
INFO-LAW@sri-CSL.  One is a draft of a bill by Senator Leahy's aid Podesta
which is a good bill.  People interested in preserving open communcation via
BBS's may wish to read these items.  Ted Shapin.

------------------------------

From: Moderator <ARMS-D-Request%MIT-MC.ARPA@MIT-XX.ARPA>
Date: 29 Oct 85 09:33-EST
Subject: Arms-Discussion Digest V5 #8 [EXCERPT: SDI Debate]
To: ARMS-D%MIT-MC.ARPA@MIT-XX.ARPA
Reply-To: ARMS-D%MIT-MC.ARPA@MIT-XX.ARPA

EXCERPTED FROM Arms-Discussion Digest Tuesday, October 29, 1985 9:33AM
Volume 5, Issue 8 
                  [There is sufficient NONOVERLAP in our readerships to
                   warrant reproducing this.  Apologies to those who have
                   read it already.  PGN]

Date:  Mon, 28 Oct 85 10:58 EST
From:  Mills@CISL-SERVICE-MULTICS.ARPA
Subject:  SDI Debate

This is a summary of my impressions of the panel discussion/debate
entitile "Star Wars: Can the Computing Requirements be Met?" This
took place on Monday October 21 at MIT.  The panelists where Danny
Cohen, David L Parnas, Charles L Seitz, and Joseph Weizenbaum. The
moderator was Michael L Dertouzos.

I was basically disappointed in this panel discussion.  I was hoping
to hear a good counter to the the arguments Dr Parnas had put forth
in his papers.  Dr Cohen started what looked like an organized attack
on Dr Parnas' "Octet", refering to the series of eight papers Dr
Parnas presented his arguments in.  Dr Cohen correctly dismissed the
eighth paper,"Is SDIO An Efficient Way To Fund Worthwhile Research",
as being outside the bounds of the current discussion.  Unfortunately
Dr Cohen only further discussed one of the other papers.  The other
six where dealt with with some minor hand waving.  I have to admit I
don't remember which paper Dr Cohen "went into detail" on.  This is
because the detail amounted to a one slide outline of the major six
points of this paper.  This slide was up for no more than one minute
with some more hand waving that none of these points were true.

Back to the side claiming the software is not feasable, Dr Weizenbaum
didn't realy add much of anything to Dr Parnas' arguments.  He thought
that Dr Parnas had done a wonderful job and there wasn't much he
could add.  He gracefully didn't take up much time saying this either.
Dr Parnas basically presented the material in his papers.  He added
the new point that even if we build this thing and it "tested OK",
we could never realy trust it and would be forced not to rely on it.

Charles Seitz made no attempt to directly attack Dr Parnas' argument.
He focused his presentation on a simplistic hierarchal structure the
software for SDI could take.  Unfortuanately this looked like a highly
centralized form of controlling all the weapons and sensors resulting
in a high degree of complexity and size.

Both Dr Cohen and Seitz hit upon the point that the software for SDI
is not necessarily as large and complex as some people might think.
They claimed that it could be built of smaller fairly independant
parts.  To me this appeared contradictory to Dr Seitz's hierarchal
control structure.  It did come through that If you had enough
totally independant platforms shotting things down, you stood a good
change of hitting most the targets.  It is also clear that you would
need a very high level of overkill to make this work since the other
platforms don't know who else is shotting at what.

Back to Dr Parnas' points,  I did get the feeling that there is some
general agreement that there is a limit to the scale and complexity
our software engineering can handle.  Dr Parnas furthered this point
by saying  large advances in the mathematics of discreet functions
are going to be a major stumbling block in the furthering software
engineering.  He doesn't expect these large advances on the grounds
that if you simplify the equations to much you are loosing
information.  A discreet function can only represent so many bits.
I may not have this argument exactly right.  He also went thru his
standing arguments against AI or automatic programing helping very
much.

    [ I think the argument is that we need concise, manipulable
    discrete functions modelling software in order to achieve what
    other fields of engineering can do with concise, manipulable
    continuous functions.  However, such concise representations may
    not be possible due to information-theoretic constraints on the
    number of bits that can be represented by a certain number of
    symbols.  --MDAY@MIT-XX ]

    [I didn't get quite this impression, though I agree with it.  Rather, I
    thought Parnas was saying that the problem was in the fact that with
    software that is fundamentally digital, there is no such thing as a
    continuous function, and that therefore the usual engineering
    assumption valid in most of the world that small changes in input or
    in correctness necessarily mean small changes in output or result
    simply isn't valid in the software engineering world.  Until it is
    possible to analyze software in terms of approximately correct
    functions, graceful software degradation (in terms of an approximately
    correct program always doing approximately correct things) is not
    really possible. -- LIN@MIT-MC]

Both sides came up with a number of interesting and colorful
analogies.  The most relavent is the Space Shuttle.  Dr Cohen claims
that the Space Shuttle works.  This is obviously true in some sense.
However, it was also pointed out that there have been times when the
software on the shuttle has not worked within seconds of launch.  It
seems that it would be impractical to ask the Soviets to wait 15
minutes while we reboot the system.

    [Indeed, Seitz conceded that under certain circumstances plausible in
    the context of a nuclear missile attack, it might be necessary to
    re-boot the system.  He then proceeed to ignore the consequences of
    that; he did not even say that there are ways to eliminate the need
    for re-booting. -- LIN@MIT-MC]

In summary it seems that there are very real limits on what our
software engineering can handle reliably.  We are actually not that
far from those limits in some of our current efforts.  If SDI is to
work its architecture must be dictated by what is doable by the
software.  It is unclear that SDI is feasably from a material cost point
of view if the platforms are small and independant enough to be
reliable from the software standpoint.

In closing I would like to say that I don't think either side did a
particularly good job sticking to just the software feasibility
issue.  One other interesting thing happened.  Dr Parnas claimed to
have asked some person with authority over SDIO whether "No, we can't
do this" was an acceptable answer.  He did this for the first time at
this debate because he did not want to say this behind this person's
back.  Unfortunately, I don't remember this other person's name, but
he was in the audience.  Dr Parnas claims that the answer was, "No is
not an accepatble answer" and challenged the other person to deny
this.  The other person promptly stood up and did exactly
that.

    [If you mean that it was political, that's certainly true.  But
    politics is really the determinant of the software specifications at
    the top level.  That is how it should be, and people who want to
    ignore that are ignoring the most important part of the problem.

    However, in other instances, the Administration has noted that the SDI
    is central to future US defense policy.  In addition, it has never
    specified what evidence it would consider adequate or sufficient to
    turn off the SDI.

    -- LIN@MIT-MC]

------------------------------

End of RISKS-FORUM Digest
************************
-------

∂19-Nov-85  1014	DAVIES@SUMEX-AIM.ARPA 	No meeting tomorrow, November 20
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  10:13:55 PST
Date: Tue, 19 Nov 1985  10:14 PST
Message-ID: <DAVIES.12160535030.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No meeting tomorrow, November 20
cc:   Davies@SUMEX-AIM.ARPA

No speaker, no meeting.

A speaker is urgently requested to come forward for next week's
meeting.  Any volunteers or suggestions for topics?

	-- Byron

∂19-Nov-85  1014	CAROL@SU-CSLI.ARPA 	Sketch demo    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  10:14:03 PST
Date: Tue 19 Nov 85 10:01:10-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Sketch demo
To: Folks@SU-CSLI.ARPA, Bboard@SU-CSLI.ARPA

*********************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DL
                   19 Nov 85
						    	
		   **REMINDER**						 
						    												
Richard Burton of Xerox Parc will demonstrate SKETCH, and entertain
questions (and you).

WHEN:   4 p.m.     
        Tues. 19 Nov.  (Today)

WHERE:  Trailer F

Contact me if you need the document, a quick introductory lesson, 
or if you have any questions.  There is a brief version of the 
documentation on {CSLI}<DANDELION.DOC>SKETCH, and in the 
yellow binders.


                              -Carol     (321-2136,  Ventura 28)

ION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION*
*********************************************************************
-------

∂19-Nov-85  1013	LIBRARY@SU-SCORE.ARPA 	Math/CS Library:  New Books--CS 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  10:13:10 PST
Date: Tue 19 Nov 85 09:51:21-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library:  New Books--CS
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12160530905.33.LIBRARY@SU-SCORE.ARPA>

Nested Transactions: An Approach To Reliable Distributed Computing.
Research Reports and Notes. Information Systems Series.  MIT Press.
by J. Eliot B. Moss.  QA76.9.D5M67 1985.

The International Workshop On Language Generation.  July 8-10, 1984.
CSLI.  (8512399)

Prolog For Programmers. APIC Studies In Data Processing. No. 24.  by
Feliks Kluzniak and Stanislaw Szpakowicz.  QA76.73.P76K58 1985.

An Artificial Intelligence Approach To VLSI Design. by Thaddeus Kowalski.
Kluwer Academic Pub.   TK7874.K675 1985.

Macintosh Revealed. Programming With the Toolbox. Volume twe. by Stephen
Chernicoff.  QA76.8.M3C48 1985 V.2.

Current Perspectives In Health Computing. Birmingham University. March
1984.  ed by Barbara Kostrewski. British Computer Society.  R858.A2C87 1984.

H. Llull
-------

∂19-Nov-85  1140	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar in Logic and Foundations   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  11:40:01 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 19 Nov 85 11:32:37-PST
Date: 19 Nov 85  1125 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Seminar in Logic and Foundations  
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    


 
Speaker:  Ian Mason, Philosophy Dept., Stanford

Title: "Proving properties of destructive LISP programs"

Time:  Friday Nov. 22, 1985, 12:00-1:00

Place: Math. Faculty Lounge, Room 383-N

                                   S. Feferman


∂19-Nov-85  1145	JF@SU-SUSHI.ARPA 	p.s. on BATS 11/22/85 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  11:45:49 PST
Date: Tue 19 Nov 85 11:05:57-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: p.s. on BATS 11/22/85
To: aflb.su@SU-SUSHI.ARPA
Message-ID: <12160544486.29.JF@SU-SUSHI.ARPA>

If you want to read the abstracts before deciding, look at 
SUSHI:<JF>bats.abstracts

Joan
-------

∂19-Nov-85  1146	JF@SU-SUSHI.ARPA 	bats headcount   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  11:46:53 PST
Date: Tue 19 Nov 85 11:04:15-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: bats headcount
To: aflb.su@SU-SUSHI.ARPA
Message-ID: <12160544177.29.JF@SU-SUSHI.ARPA>

Is there anyone who plans to attend BATS at DEC this Friday (Nov. 22) who
has not told me?  If so, please tell me.
Joan (jf@sushi)
-------

∂19-Nov-85  1243	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	course description  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  12:41:24 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 19 Nov 85 12:30:05-PST
Date: 19 Nov 85  1218 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: course description 
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA, su-bboards@SU-AI.ARPA 


Course description  for CS350 (MTC)  Winter 1986:

    Programming and proving with functional and control abstractions.

			     Carolyn Talcott


This course combines ideas from semantics of programming languages,
mathematical theory of computation, and program transformations to develop
a semantic foundation for symbolic (Lisp-type) computation.  The goals
are:  to develop a context in which both intensional and extensinal
aspects of computation can be treated; to provide mathematical tools for
obtaining a better understanding of current practice in symbolic
computation and for representing and proving properties of programs; and
to provide operations on programs with meanings to transform and meanings
to preserve.

The main idea is to treat computation as a process of generating
computation structures such as trees and sequences and to treat more
abstract interpretations of programs as derived notions.  The language for
describing computations is that of the lambda calculus extended by
conditional, sequence formation, and context noting primitives.
Additional mathematical tools for reasoning about computation are provided
by a class of approximation and equivalence relations on programs.  These
relations are a means forgetting selected details of computations while
preserving evaluation and application relations.  There is a maximum
approximation and a maximum equivalence.  These maximum relations are
extensional.  The recursion operator (Y-combinator analog) computes the
least fixed point with respect to the maximum approximation.

Use of the mathematical tools developed will be illustrated by a variety
of applications.  The applications include:  proving properties of
programs that use streams, escape mechanisms and co-routines; a
correspondence between streams and co-routines; the use of functionals to
construct and prove properties of programs; program transformations
involving introduction of function and control abstractions; a general
method for converting intensional properties of programs into extensional
properties of ``derived'' programs; and using derived programs to analyze
the effects of program transformations.

Much of the material will be based on a model of computation developed by
Talcott.  As background, work in semantics and theory of computation will
be reviewed including work of McCarthy, Landin, Morris, and Wegner.  Work
based on alternate semantic models will also be discussed, including work
of Scott, Plotkin, Mosses, and Moschovakis.

Suggested prerequisites: CS306 or equivalent background in logic and Lisp



∂19-Nov-85  1315	EMMA@SU-CSLI.ARPA 	Newsletter 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  13:15:34 PST
Date: Tue 19 Nov 85 13:05:45-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter
To: folks@SU-CSLI.ARPA
Tel:  497-3479


  There will be no newsletter next week because of the Thanksgiving
Holiday.  I will be including in this week's newsletter announcements
of all events that occur before Friday, December 6, i.e., two weeks from
this Friday.

Please send the necessary information before noon, tomorrow (Wednesday).

Many thanks,

Emma
-------

∂19-Nov-85  1337	ACUFF@SUMEX-AIM.ARPA 	Explorer Hardware Watch
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  13:37:21 PST
Date: Tue 19 Nov 85 13:04:02-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Explorer Hardware Watch
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12160565981.31.ACUFF@SUMEX-AIM.ARPA>

   If you see either of the following two problems, I'd appreciate it if you'd
let me know so that the associated hardware problem can be addressed:

1. "Burbling" wherein the machine acts like a cat is walking on the keyboard.
   Powering the console on and off a few times will usually clear thi
   condition, but let me know if it happens.

2. Repeating the last character typed even though no key is down.  Typing
   another character (anything) will clear this, but, again, let me know.

	Thanks,
	-- Rich
-------

∂19-Nov-85  1618	RICHARDSON@SU-SCORE.ARPA 	Faculty Accomplishments 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  16:18:11 PST
Date: Tue 19 Nov 85 16:17:49-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Faculty Accomplishments
To: ac@SU-SCORE.ARPA
Message-ID: <12160601258.31.RICHARDSON@SU-SCORE.ARPA>

This year we will be reporting our "faculty accomplishments" through the
School of Engineering instead of through Humanities and Sciences. SOE has
a stanford form for "faculty accomplishments" which is attached to this
net message. (You will also receive a hard copy through ID mail.) 

Please return your completed forms to me either by e-mail or hard copy by
December 2 for forwarding to SOE.

Thanks,
Anne
--------------------------------------------------------------------------

Stanford University
School of Engineering
Annual Faculty Report for Academic Year 1984-85


November, 1985


It is time again for a Faculty Report.  This office finds it very useful to
have the information outlined below, and I appreciate you taking time to fill
out the form carefully.  I realize that this represents only a summary 
of your contributions to the School and misses completely your goodwill 
and spirit which are equally important to our mission.

Please give this completed form to your departmental secretary by 
December 2, 1985.  Thanks for your help with this chore and 
for your contributions to the School and the University.

Cordially,

Jim Gibbons,
Dean


(Please note:  Information requested pertains to the period 9/1/84 to 
8/31/85 only.)

NAME←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
Last, First, Middle

ACADEMIC RANK←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

DEPARTMENT←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

Publications:  (Please indicate nature of work, such as books,
monographs, journals, technical reports, etc., giving title, date, pages and
publisher or issuing agency.  Include papers submitted for publication and
designate as such.)





Teaching: (Please indicate by quarter, course title, number of units and
enrollment.  Also include course or curriculum development, computer education
software tutorials, specially prepared television presentations or other
relevant work.)





Advising:

Number of freshman advisees

Number of undergraduate advisees

Number of graduate advisees


Supervision of Ph.D. Candidates:

Number of students for which you
are principal dissertation advisor.

Number of students for which you
are on reading committee.


Research Projects:


Project title and       Names of Principal      Approx. annual dollar
Funding Source          and co-Principal        value of project for 
                        Investigators,          which you are 
                        if any                  responsible.




University Service Other Than Teaching and Research: (Include
administrative duties and committee work.)



Professional Activities Outside the University:  (Include offices in
professional organizations, services to government agencies or industry,
editorship of journals, invited presentations, and outside adminstrative or
public service.)



Honors and Awards:
  

Describe below any activities or make any comments that do not fit under
previous categories.






-------
-------

∂19-Nov-85  2039	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #36  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 19 Nov 85  20:32:30 PST
Date: 19 Nov 85 2017-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #36
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 20 Nov 1985     Volume 1 : Issue 36

Today's Topics:

                      SIMD vs MIMD (4 messages)
                     Connectionist Summer School

----------------------------------------------------------------------

Date: Monday, 18 November 1985  06:36-PST
From: Seth Steinberg <sas at BBN-VAX.ARPA>
Subject: PARSYM Digest   V1 #34

I have just read Alan Bawden's and Phil Agre's paper in which they
explore the programming of the Connections machine by implementing a
Scheme interpreter for it.  Since this is probably one of the hairiest
Connections Machine (SIMD) programs ever written it gives one a good
idea of what can be programmed on this type of architecture.

Suppose we want to run a program which contains a conditional on a
SIMD machine.  First we tell all the processors to evaluate the
predicate, which is rather straightforward.  All processors do the
same test and then they set some condition code.  Then we send out the
"then clause" with the instructions suitably modified so that they are
only performed if the condition code is set, then we send out the
"else clause" with the instructions modified so that they are only
performed if the condition code is not set.  In other words executing
a conditional on a SIMD machine we take time P+T+E
(predicate+then+else) as opposed to max(P+T,P+E).  As Danny Hillis
points out in his book (The Connection Machine), you tend to have to
accept the worst case rather than the typical case.

I recently read a paper by Alan Bawden and Phil Agre in which they
implemented a Scheme interpreter for the Connection Machine.  A Scheme
program is decomposed into abstract data objects (e.g. assignment
objects which know about the target and the expression, conditional
objects which know about the predicate, then and else clauses and so
on).  They implement this by using one processor to represent one data
object by placing it in a suitable state so that it only executes the
instructions for implementing that type of object.

It's kind of like those Avalon Hill war games where the enemy gets to
advance scouts, advance infantry, now you get to advance scouts and
infantry, then they shoot at each other, now the enemy advances his
tanks, then you advance your tanks, they shoot at each other and at
the infantry and the scouts.  Now the medics remove the wounded and
you tot up the damage done and you're ready for the next turn.  Some
Avalon Hill games were real good SIMD training.

To interpret Scheme the master controller has to send out all the
instructions for all of the conditional type objects (object and
processor will be used interchangeably here).  This means ALL the
instructions including those for each case (handling both the then and
else cases).  It also has to send out all the instructions for all the
assignment type objects, and the sequences, and the procedure
applications, and the closure creaters and so on.  (I haven't even
mentioned the code for +, -, eq?, null?).  In other words, the entire
Scheme interpreter (just about) has to be sent down the instruction
wires every so often.

I imagine the folks at MIT actually considered running Scheme this way
especially if they considered the projected enrollment in 6.001 in the
year 1990, but they decided just to get more detached work stations.
In any event, Scheme will not run particularly efficiently on a
Connection Machine because it has too many different types of objects
in it.

Programs which run efficiently on a CM are those in which all the
objects are of the same type.  As the theoretical physicists tell us,
it is not all that dificult to make two different typed objects seem
to be of the same type modulo some marker (spontaneous symmetry
breaking) but a SIMD machine must still check for those markers at
some time.  If we argue that all Motorola 68000 instructions are the
same except for their opcodes we can indeed emulate a MIMD machine on
a SIMD machine but we still have to send out the emulation sequence
for every instruction each emulated cycle.

In general, SIMD machines execute Fortran and APL very well.  (The
latter language is full of nifty symmetry hacks).  I am exaggerating
for effect but it is probably safe to say that any algorithm which
actually looks elegant when written out in Fortran can probably be
made to run efficiently on a SIMD machine.  Any APL one liner (and
there were a lot of graph theory one liners) can be made to run
efficiently as well.  This is a LOT of code, and more importantly,
this is a lot of code FULL of repeated iterations - it is that 10%
which eats up 90% of the typical conventional processor.

There are many algorithms which don't quite fall into this basket
though.  It is hard to do symbolic integration, make legal decisions,
parse a sentence, compile COMMON LISP, run a numerical backsolver, or
perform any operation which has to deal with lots of different kinds
of objects and lots of special cases.  UNIX, like any other piece of
system code, is FULL of special cases.

My feeling is that SIMD machines have their place but that they are
still poorly understood.  Perhaps they should be partitionable with
different partitions receiving different instruction streams (NIMD)
which would make BIBOP Scheme quite practical.  As far as software
goes we are back at core memory and Fortran I.

					Speaking for Myself,
					Seth Steinberg

P.S.  I realize that some of my opinions in here could be quite
offensive.  No offense has been intended, I have been trying to
understand the SIMD world and my understanding has been subject to
massive and sudden mood swings.

------------------------------

Date: Monday, 18 November 1985  19:18-PST
From: gary at THINK.COM
Subject: SIMD vs MIMD

In reply to Sal's letter:

>Think of the different tests and branches that would need to be issued
>from the control processor to the SIMD processor to account for every
>possible type of first order term encountered.

The tests and branches that a SIMD program would have to contain are,
syntactically, exactly the same as those in a program running on a
processor in a MIMD machine.  It is just as easy to write the SIMD
program as to write an ordinary serial program.  The difference comes at
run time: the SIMD host has to run through the code for all possible
outcomes of conditional branches, but a serial computer, or the
processors of a MIMD computer, can just skip over sections of code when
a condition evaluates to false.  However, this process is transparent to
the programmer.


>Implementation ease aside, an MIMD processor is probably much faster at
>this game than an SIMD processor.  Don't forget, in parallel processing
>SPEED IS THE NAME OF THE GAME.

Of course a MIMD computer will be faster than a SIMD computer at any
task, when they have an equivalent number of processors.  The comparison
that should be made is between a MIMD and a SIMD computer implemented on
equivalent amounts of silicon.  Because SIMD computers do not need to
devote silicon to control and program store at each processor, a SIMD
computer can have more processors than a MIMD computer for any given
amount of silicon.  For most "real problems" (as defined by Brad's
original letter) this more than compensates for the need to turn off
some of the processors when a conditionalized statement is executed.

------------------------------

Date: Tuesday, 19 November 1985  10:38-PST
From: eldridge at EDN-VAX.ARPA (Charles Eldridge)
Subject: SIMD versus MIMD

Expert systems should be an application area in which MIMD
has advantages over SIMD.  That is, a number of searches must
be performed in order to competitively evaluate chains of
reasoning.  SIMD would require that these searches be carbon
copies of one another, while MIMD would allow them to pursue
differing branches, having different lengths and evaluating
different (perhaps) mathematical functions along the way.

Have you seen "A Qualitative Assessment of Parallelism in
Expert Systems," by R. J. Douglass in IEEE Software, May 1985,

and

"Providing Architectural Support for Expert Systems," by
P. C. J. Graham, ACM SIGARCH Computer Architecture News,
December, 1984?

--Charles Eldridge

------------------------------

Date: Tuesday, 19 November 1985  09:02-PST
From: Guy Steele <gls at THINK-AQUINAS.ARPA>
Subject: PARSYM Digest   V1 #35

This comment is in response to the interesting and provocative remarks
of Sal Stolfo about SIMD versus MIMD.

Perhaps speed is the name of the game, but I would suggest that speed
of solving a problem is the name of the game, not speed of individual
processors.

I think that SIMD appears to many people compare unfavorably to MIMD
because it is tempting, in the mind's eye, to picture a single
processor, whether SIMD or MIMD, as being of about the same size.  I
think we tend to mentally compare a 1000-processor SIMD machine with a
1000-processor MIMD machine.  In such a comparison the MIMD machine
probably does come out ahead in speed and flexibility (though perhaps
not in ease of programming, yet again perhaps so).

But if one compares the two strategies by insisting that each systems
consist of about the same amount of hardware (or have about the same
dollar cost), then the SIMD machine might come out with (a) more
processors and/or (b) greater data bandwidth.  It might have more
processors because instruction-fetching and decoding circuitry can be
shared, and so the part of each processor that must be replicated can
be smaller.  It might have greater data bandwidth because a much
smaller proportion of memory accesses go toward instruction fetching.
These two effects may compensate for inefficiencies introduced by the
time-slicing of SIMD processors in emulation of MIMD programming mode.

As for the amount of memory per processor, I think that any computer
tends to have a certain fraction of hardware devoted to memory and the
rest to processor.  (Whether this is essential (I doubt it), the
result of tradition, or the result of economic pressures is unclear to
me.)  If a SIMD and a MIMD system have the same amount of hardware,
and each devotes the same fraction of hardware to memory, and if
conjecture (a) above holds (the SIMD machine has more processors),
then naturally there will be less memory per processor.  But the total
amount of memory is the same, and if one insists on assigning one
problem element per processor, then the SIMD machine can handle
problems with more elements, even though each one must be smaller.
Alternatively, one can afford to assign multiple processors to each
problem element; I don't call this "trading in concurrency" if this
allows handling a problem of the same size as the equivalent (in
hardware size) MIMD machine.

I certainly agree that performing unification on many expressions in
parallel is much easier to think about in MIMD terms, and a MIMD
programming interface is a useful one.  I do not think that this
necessarily implies that the underlying software and hardware should
be understood or constructed in MIMD terms at all levels.  In any case
we should be careful to define exactly and fairly what is being
compared.  A Cray will almost always beat out a PCJr--but at what
cost?  As another example, I don't think it is fair to say that
NON-VON is better than a Connection Machine because NON-VON has 8-bit
processors and CM has 1-bit processors.  Nor is it fair to say that CM
is better for having 64K processors to NON-VON's 8K processors.
(Indeed, I think these two systems are roughly comparable at an
architectural level, in that on each "cycle", whatever that means,
each machine can fetch and process 64K bits.  I am speaking here in
ignorance of clock speeds.  Differences in instruction sets will of
course result in differing performance on specific applications.)  One
can have a very good discussion about whether NON-VON's multiple
controllers, giving it more of a MIMD flavor at that level than CM,
constitute an advantage for certain applications.

--Guy

------------------------------

Date: Monday, 18 November 1985  18:36-PST
From: Brewster Kahle <KAHLE at THINK-AQUINAS.ARPA>
Subject: A good use of MIMD

MIMD machines are appropriate (at least) when the processors are
physically far from each other, such as process control in a factory.
This may seem like a trivial point, but it is enough of an issue to
keep some people working on how to control them.

-brewster

------------------------------

Date: Monday, 18 November 1985  20:29-PST
From: Dave.Touretzky at A.CS.CMU.EDU
Subject: summer school

                    CONNECTIONIST MODELS:  A SUMMER SCHOOL

                       Sponsored by the Sloan Foundation


ORGANIZERS:  Geoffrey Hinton (Carnegie-Mellon University)
             Terrence Sejnowski (The Johns Hopkins University)
             David Touretzky (Carnegie-Mellon University)

DATE:  June 20 through 29, 1986

PLACE:  Carnegie-Mellon University, Pittsburgh, Pennsylvania

PURPOSE OF  THE  PROGRAM: The  purpose  of  the summer  school  is  to
familiarize young researchers with current  techniques in the area  of
connectionist  models   of   intelligence.    This   includes   search
procedures,  learning   procedures,  and   methods  for   representing
knolwedge  in  massively  parallel  networks  of  neuron-like   units.
Application areas include vision, speech, associative memory,  natural
language and motor control.


FACULTY:  There will be six full time Tutors plus several Guest Lecturers.

    Tutors:                          	Guest Lecturers:
    James Anderson, Brown University 	Jerome Feldman, U. of Rochester
    Dana Ballard, U. of Rochester    	Christof Koch, MIT
    Andrew Barto, U. Mass. Amherst   	David Rumelhart, UCSD
    Geoffrey Hinton, CMU             	David Touretzky, CMU
    James McClelland, CMU            	others to be announced
    Terrence Sejnowski, Johns Hopkins


WHO MAY  ATTEND: Participation  is limited  to graduate  students  and
recent PhD's  who are  or  will be  working on  connectionist  models.
About  40  students  will  be  accepted.   Persons  who  have  already
completed a Ph.D.  degree must have  done so no  earlier than  January
1985 to be eligible to attend.

EXPENSES:  There  is  no  tuition  charge.   Funding  from  the  Sloan
Foundation will  provide dormitory  accommodations and  breakfast  and
lunch for each attendee, plus reimbursement for a substantial  portion
of travel expenses.

HOW TO APPLY: By March 1, 1986, send your curriculum vitae and a  copy
of one relevant paper, technical report, or research proposal to:  Dr.
David  Touretzky,   Computer   Science   Department,   Carnegie-Mellon
University, Pittsburgh,  PA, 15213.   Applicants will  be notified  of
acceptance by April 15, 1986.

------------------------------

End of PARSYM Digest
********************

∂20-Nov-85  1153	RICE@SUMEX-AIM.ARPA 	36xx -> Explorer Status.
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  11:53:37 PST
Date: Wed 20 Nov 85 10:40:19-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: 36xx -> Explorer Status.
To: aap@SUMEX-AIM.ARPA
Message-ID: <12160801963.66.RICE@SUMEX-AIM.ARPA>



Progress report on the porting of code to Explorers.

This is the current state of the systems, in which the architectures project
might be interested, that are being ported to the Explorers.

Before you try to load any of these systems you should load up the
compatibility package files.  This is done with the following :-
    (Load "3602:>Tools>36xx-Explorer")
    (Load "3602:>Tools>36xx-Explorer-2")
    (Load "3602:>Tools>36xx-Explorer-3")
These fix a fair number of the known bugs and/or incompatibilities, which we
have addressed.  They may change from time to time so you would do well to look
in the Tools directory for any changes.  Once all of these changes have been
logged properly they will be collapsed into one file so watch out.

As usual any bug reports and/or fixes should be channeled through Mr Acuff so
as to prevent the reinvention of the wheel and unnecessary flamage.


Helios -	No work has been done on this yet but it is not critical.
Simple -	This appears to work except for the timing patch.
Care -		This can be loaded as XCare.  All of the subsystems seem to
		work but for the timing patch in Simple.  Backup (Care)
		systems are yet to be made.
		To load it do the following :-
		    (Make-System 'XCare :Noconfirm :Silent :Nowarn)
System-Manager -This seems to work but for a problem with typeout windows.
		TI are looking at the problem.  Sources will eventually be
		moved to Ardvax but not until TI/Mr Acuff can fix the bugs
		in the pathname system for Unix machines.
		To load it do the following ;-
		    (Make-System 'System-Manager :Noconfirm :Silent :Nowarn)
L100 -		The L100 compiler is fully operational on the Explorers.
		You can load it with :-
		    (Make-System 'L100 :Noconfirm :Silent :Nowarn)
Tina-Language -	This is fully operational on the Explorers.
		To load this please load the tina system first (see below)
		and then do :-
		    (Make-System 'Tina-Language :Noconfirm :Silent :Nowarn)
Cage-Language -	This works as much on the Explorers as it does on the 3600s.
		You can load it with :-
		    (Make-System 'Cage-Language :Noconfirm :Silent :Nowarn)
Tina -		This works fully on the Explorers but for a small problem in
		the trace/debug menus.  This should be fixed soon.
		You can load it with :-
		    (Make-System 'Tina :Noconfirm :Silent :Nowarn)
Czar/Caos -	Work has started on porting this.  It currently gets most of
		the way through its initialisation but is being held up by
		what could be a microcode bug.  This is being investigated.
Parallel-Tina -	This is predicated in Czar so work has not started.

Early next week we expect some new system software and sources.  This may
well change things but with luck it will allow us to respond to bugs more
rapidly.

If you have any questions about these systems please see me.  If you have any
questions about your machine please see Mr Acuff.

Rice
-------

∂20-Nov-85  1158	PATASHNIK@SU-SUSHI.ARPA 	Next AFLB 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  11:58:42 PST
Date: Wed 20 Nov 85 06:16:00-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Next AFLB
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12160753845.12.PATASHNIK@SU-SUSHI.ARPA>

This week's AFLB has turned into a BATS talk on Friday, and next week
is Thanksgiving, so the next AFLB will be in two weeks:

		*************************************

5-Dec-85  :  John Cannon (University of Sydney)

	Finding the order of a permutation group of degree 10,000

			(Abstract forthcoming)

***** Time and place: December 5, 12:30 pm in MJ352 (Bldg. 460) ******
-------

∂20-Nov-85  1205	NEUMANN@SRI-CSL.ARPA 	RISKS-1.24   
Received: from SRI-CSL.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  12:04:55 PST
Date: Wed 20 Nov 85 07:57:30-PST
From: RISKS FORUM    (Peter G. Neumann, Coordinator) <RISKS@SRI-CSL.ARPA>
Subject: RISKS-1.24
Sender: NEUMANN@SRI-CSL.ARPA
To: RISKS-LIST@SRI-CSL.ARPA

RISKS-LIST: RISKS-FORUM Digest  Wednesday, 20 Nov 1985  Volume 1 : Issue 24

        FORUM ON RISKS TO THE PUBLIC IN COMPUTER SYSTEMS 
                 Peter G. Neumann, moderator

Contents:
  Doing Something About Risks in Computer Systems (Brad Davis)
  Space Program Software (Jerome Rosenberg)
  Susceptibility to interference (John Brewer)
  Expecting the unexpected  (Herb Lin)
  Philip W. Anderson's "Case Against Star Wars" (Pete Kaiser)

Summary of Groundrules:
  The RISKS Forum is a moderated digest.  To be distributed, submissions should
  be relevant to the topic, technically sound, objective, in good taste, and 
  coherent.  Others will be rejected.  Diversity of viewpoints is welcome.  
  Please try to avoid repetition of earlier discussions.

(Contributions to RISKS@SRI-CSL.ARPA, Requests to RISKS-Request@SRI-CSL.ARPA)
(FTP Vol 1 : Issue n from SRI-CSL:<RISKS>RISKS-1.n)      

----------------------------------------------------------------------

Date: Tue, 19 Nov 85 11:05:34 MST
From: b-davis@utah-cs.arpa (Brad Davis)
To: RISKS@sri-csl.arpa
Subject: Doing Something About Risks in Computer Systems

Often the discussion has touched on failure of software and hardware, but
rarely on levels and methods of protection that should be built into these
systems.  Is is good to trade cycles for protection?  What are the best ways
to recover from failures?  Does anyone have real experiance with these
questions?
				Brad Davis

    [Clearly these are leading questions!  We have indeed mentioned many
     good techniques of software engineering that help.  But there are no
     easy answers -- especially in the absence of specific requirements.
     But let's see if any of our readers wants to take a crack at this one.
     PGN]

------------------------------

Date: Tue, 19 Nov 85 14:46:49 CST
From: jerome@rsch.wisc.edu (Rosenberg Jerome)
To: RISKS@sri-csl
Subject: Space Program Software

           We have heard a great deal about the great successes of the space
  program but we rarely hear about the difficulties that have to be overcome
  with great effort and dedication. I suggest you direct your readers to the 
  current issue of DATAMATION for an article by Edward Joyce entitled "The
  Art of Space Software". Its subtitle tells a far different story than
  some hand-waving protagonists of the SDI tell about the Space software.
  The subtitle  -- The complicated software labyrinth behind the shuttle is
  still far from error-free -- tells the story. The article should serve to
  alarm those who are quick to discount the sincere critics of the SDI
  software problems.                     jerome @rsch.wisc.edu

------------------------------

Date: Tuesday, 19 Nov 1985 10:21:51-PST
From: brewer%ace.DEC@decwrl.DEC.COM  (too busy for bureaucracy -John 5522026)
To: risks@sri-csl.ARPA
Subject: Re: Susceptibility to interference

	RE: Bennett Smith's comments of emi-rfi susceptibility in automobile
control applications... cb's are low power, limited frequency devices. As an
Amateur radio operator, one has to be aware of much higher output power, as
well as a much wider bandwidths. Amateur Radio frequency allocations include
segments from 1.8Mhz to Ghz ranges.

	As I remember, some of the control modules are also pretty good
emitters of Emi/Rfi hash as well. Typical (legal) output power of a CB is 5
watts or less. A typical ham radio mobile transmitter output power is
100-200 watts.

	Something to think about!
	-John

------------------------------

Date: Tue, 19 Nov 85 15:10:41 EST
From: Herb Lin <LIN@MIT-MC.ARPA>
Subject:  Expecting the unexpected
To: RISKS@SRI-CSL.ARPA

Regarding your comments about spontaneous failure: The Russians have a
saying regarding rifles used on stage in plays: once every decade an
unloaded gun will fire; once every century a rake will fire.

      [Perhaps that is what prompted Stravinsky to stage
       "The Rake's Progress".  PGN]

------------------------------

Date: Wednesday,  2 Oct 1985 21:32:34-PDT
From: kaiser%furilo.DEC@decwrl.ARPA  (Pete Kaiser, 225-5441, HLO2-1/N10)
To: RISKS@SRI-CSLA.ARPA
Subject: Philip W. Anderson's "Case Against Star Wars"

    [The following message was put aside for evaluation before my absence.
     With the reminder that we of course would like to see more informed
     pro-SDI contributions in RISKS as well, Anderson's article seems
     worth including -- not because it breaks new ground, but because it
     represents a position for discussion.  PGN]

The article below, by Professor Philip W. Anderson of Princeton University,
appeared in the Princeton Alumni Weekly of September 25, 1985, and is reprinted
here with the author's permission.  Professor Anderson won the Nobel Prize for
Physics in 1977, and was awarded the National Medal of Science in 1982.

Although what Professor Anderson has to say is couched partly in specific terms
of Princeton University and the discipline of academic physics, it seems to me
relevant to basic research in general, and to computer science research and the
discipline of computer science in particular.  To me, for instance, it seems to
be very personally a social consequence of the military funding of computer
science research that, while I've worked with computers, there have been many
kinds of work which I couldn't conscientiously do because, although they may be
very interesting, they are done essentially only for military purposes and with
military funding.

Finally, Professor Anderson points out that a great deal of sensible thought can
be brought to social issues even by someone who "isn't ... fascinated by the
technical details."  Agreed.  We must remember that we're not priests.

---Pete

Kaiser%BELKER.DEC@decwrl.arpa
{allegra|decvax|ihnp4|ucbvax}!decwrl!dec-rhea!dec-belker!kaiser
DEC, 77 Reed Road (HLO2-1/N10), Hudson MA 01749  617-568-5441

                    ----------------------------------

			The Case Against Star Wars
                      Philip W. Anderson, Princeton

I am not an expert on strategic weapons.  I'm a theoretical physicist who has
been involved in almost all of physics except atomic bombs.  I have not done
classified work since 1945, and that was on radar.  My total contribution to the
laser -- a major technical component of the Strategic Defense Initiative, which
is better known as Star Wars -- was roughly that when one of the scientists at
Bell Laboratories who originated the things asked me to predict whether a
certain seminal version of it would work if they built it, I said "Well, maybe."

Fortunately, most of the scientific issues that come up in discussing Star Wars
are very simple ones which require neither specialized nor especially technical
-- and therefore classifiable -- knowledge.  One needs to know that it costs
everyone about the same amount to put a ton of stuff into a given orbit and that
this is a major portion of the cost of any space system; that signals can't
travel faster than the speed of light; that it takes roughly as much chemical
fuel to burn through a shield with a laser as the shield itself weighs; that
Americans are not measurably smarter than Russians; and a few other simple, home
truths.  Given these, almost everyone comes to much the same conclusions.

If you go through the enormously detailed kinds of calculations on specific
configurations which Richard Garwin and his fellow opponents of SDI felt
necessary to convince the stubborn, you leave yourself open to the kind of
errors of factors of 2 or 4 which Martin Muendel '86 found in his widely
publicized junior paper last spring [Princeton Alumni Weekly, May 8] and which
then -- to the lay person -- seem to weaken the whole structure.  This is a
particularly tough game because Star Wars advocates do not themselves propose
specific configurations and present specific calculations that can be shot down;
their arguments are given in terms of emotional hopes and glossy presentations.
This is why I think it is good for the argument against SDI to be made by a
mentally lazy, non-expert person like myself who isn't particularly fascinated
by the technical details.

The reasons for not building Star Wars are essentially identical to those which
led both us and the Russians to abandon, for practical purposes, the antibal-
listic missile in 1972 and to sign a treaty restricting ABMs.  It is important
to understand that reasoning -- and perhaps it is less emotionally charged than
Star Wars since it is now history and not even controversial history anymore.
Why would anyone feel that a defense against missiles was useless and, in fact,
dangerous and destabilizing?

There are three stages, each more certain than the last: (1) It probably
wouldn't work, even under ideal conditions.  (2) It almost certainly wouldn't
work under war conditions.  This puts us in the dangerous and unstable situation
of the gunfighter who doesn't know if his gun is loaded.  (3) Most certain and
conclusive of all, *each defensive system costs, inescapably, at least 10 times
more than the offensive system it is supposed to shoot down*.  Thus it pays the
other side to increase its offensive arsenal until the defender is bankrupt, and
the net result is an *increase* in armaments and a far more dangerous situation,
without any increase in safety.

The offense has, inescapably, enormous advantages: its missiles are sent at
will, in any desired sequence and quantity, with any number of decoys and other
deceptive countermeasures, preprogrammed at leisure to hit their targets; the
defense has to find them, sort them out, get into space at a time not of its own
choosing, and then kill the warheads it finds with nearly perfect accuracy.  In
the case of ABM, there were other problems, such as that the explosions were
over the defending side and that the first few explosions probably blacked out
the whole shooting match, but that was sufficient argument against.

As far as almost everyone in and out of the Defense Department was concerned,
until March 1983 this situation was an accepted fact.  No technical breakthrough
had or has changed those realities.  The change has been purely political and
emotional, and hence now financial.  President Reagan's March 1983 speech, as
far as anyone can ascertain, was not preceded by any serious technical review,
but quite the opposite: the most recent and urgent internal study of antimissile
defenses had come out negative on all possible schemes.

Apparently, the President based his speech and his subsequent program on a
collection of rather farfetched suggestions -- farfetched but by no means secret
and previously unknown -- which, to the outside scientific observer, seem to
deserve the oblivion that the last pre-Star Wars study consigned them to.  These
schemes amount to a way for the defense to spend more per missile and still let
through a large fraction of the offensive missiles.  The defensive hardware that
has to be got up into space still has to have roughly the same mass as the
offense; in many schemes it has to get there faster; and it still has to be much
more sophisticated and therefore vulnerable and delicate.  Key components, in
most schemes, have to be left in space indefinitely, inviting the enemy to track
them with space mines, perhaps the most dangerous tripwire mechanism for stating
a war that one can possibly imagine.


Some Star Wars advocates will protest that I do not mention the one idea which
doesn't founder just on the problem of total mass in space.  This is the scheme
of exploding hydrogen bombs in space and directing the explosive energy of the
bombs with lasers to kill very many missiles per bomb -- several hundred to
several thousand, if one is to kill an equivalent cost in missiles! If I could
think of any way such a monstrosity could work as opposed to the many ways it
could not work or be frustrated, I would take it more seriously.  Apparently
there has been some good and interesting science done on these lasers, but
unfortunately it is classified; no one, however, seems to claim that it helps
much with the technical problem.  I cannot, incidentally, see any way to do
meaningful development on such a weapon without exploding H-bombs in space, a
terrible pollution as well as a violation of what treaties we have.

I think the above would represent reasonably well the views on the technical
realities of most trustworthy physicists to whom I have spoken, in or out of
academia and in or out of the Star Wars program.  In academic physics depart-
ments, which receive relatively little support from the DOD, a pledge form has
been circulating stating that the signer opposes SDI as unworkable and will not
seek SDI funds; this has had a high percentage of signers everywhere it has been
circulated and its preliminary circulation in Princeton over the summer encoun-
tered only a few holdouts.  Those who do not sign feel, primarily, that research
in any guise shouldn't be opposed, while agreeing personally that the systems
proposed are unworkable and destabilizing.

Perhaps it would be worthwhile, therefore, for me to explain why I feel the
large increment of research funds earmarked by President Reagan for SDI is a
very bad thing for the research community, as well as for the country as a
whole.  You will note that I said *increment*; every year before Star Wars, we
spent $1 billion in ABM research and development.  My main reason is that, on
the whole, Star Wars will represent a further acceleration of three extremely
disturbing trends in the direction of research funding in this country.

First, we are seeing a decrease in basic research relative to mission-oriented,
applied research.  The basic research agencies -- National Science Foundation,
Basic Energy Sciences in the DOE, and National Institutes of Health -- have been
maintained at level funding while their missions have been gently skewed toward
applications and engineering by piling more applied responsibilities on them.
At the same time, while the Administration has cut back on development in some
civilian sectors, it has more than compensated by increasing the amount of
applied work for the military.

Second, there is a trend away from scientific administration of federal research
money -- mostly done by the system of "peer review" -- to micromanagement either
by bureaucrats, or, increasingly, by Congress, with all the logrolling possibil-
ities that entails.  The three institutions mentioned above, especially NSF and
NIH, operate by subjecting each grant to a jury or other scientists.  Like most
democratic procedures, this system is worse than everything except the alterna-
tives; its effect has been reviewed repeatedly and there is no serious doubt
that it works.  Military "research," on the other hand, has always operated on
the arbitrary whim of the contracting officers.  In the early days after World
War II this administration was a benevolent despotism, but the adjective has
long since lost its meaning.  Most of the in-house DOD laboratories have been
rather a scandal in the research community.  The dominant motivation in this
system seems to be the standard bureaucratic one of "empire building."

Third, from the point of view of the country as a whole, perhaps the most
dangerous trend is the shift from civilian to military dominance of our federal
research and development spending.  Under the Reagan Administration, this has
grown to 72 percent military, up from about 50 percent a decade ago.  Everyone
has been told -- and DOD sees to that -- of the great economic benefits of
"spin-off" from military development, but if they exist (and I have never found
an economist who believes in them), they are not evident in our recent economic
performance vis-a-vis Japan and Germany.  In fact, in a country like ours with a
serious shortage of trained engineers and scientists, a shortage which would be
crippling if we did not attract great numbers of them from overseas to staff our
universities and research laboratories, the waste of our precious technical
expertise on military hardware is a serious economic debit.

From Princeton's point of view, all of these trends are disturbing.  As a top-
flight research university, a heavy percentage of our funding is in individual
support of independently functioning basic scientists, mainly peer-reviewed and
to a large extent from the agencies mentioned above.  We have not had to resort
to logrolling political tactics, nor have we had to accept micromanagement, DOD
control of publications, or limitations on citizenship of students to keep our
research funded.  SDI control of funding, and in general the shift of research
funding to the military, is a serious danger to the independence of Princeton as
a research university.

Of course, this is a narrow and slightly parochial view, but it is nonetheless
serious.  Certainly it is more important that the naive emotional appeal of the
Star Wars concept is being used so blatantly to defuse the country's strong
desire for nuclear disarmament, and to turn this emotional pressure into yet
another excuse for enriching the arms manufacturers and building up a dangerous
and worthless arsenal of nonsensical armaments.  To paraphrase Murph Goldber-
ger's testimony on the ABM: Star Wars is "spherically" senseless -- that is,
silly no matter how you look at it.

            [End of Philip Anderson's statement, and of Pete Kaiser's Message.]

------------------------------

End of RISKS-FORUM Digest
************************

-------

∂20-Nov-85  1205	LIBRARY@SU-SCORE.ARPA 	Socrates:  Update On Searching Math/CS Technical Reports By Call Numbers (Accession Numbers)
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  12:04:28 PST
Date: Wed 20 Nov 85 11:52:05-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Socrates:  Update On Searching Math/CS Technical Reports By Call Numbers (Accession Numbers)
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12160815027.41.LIBRARY@SU-SCORE.ARPA>

You may now search out technical reports by our call number (or accession 
number) with or without the leading zeros.  Technical report number 009902
may be found by searching Fin C 009902 or Fin C 9902.

Be aware that besides the call number index there is also a Number index 
which includes the technical report numbers and not call numbers.  Because
this file includes reports from various science libraries, technical report
numbers and call numbers will vary in how they are entered.  Computer 
Science Technical report numbers will vary with some of them having just
a unique number and others have a unique number preceeded by a year notation
such as 84-201 or 85-609.  You need to be very cautious when searching
either the number index or the call number index.  If you have any 
problems, let me know.

Harry Llull
-------

∂20-Nov-85  1238	YAMARONE@SU-CSLI.ARPA 	HUNGRY? X-TRA SANDWICHES AVAILABLE   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  12:36:50 PST
Date: Wed 20 Nov 85 12:27:00-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: HUNGRY? X-TRA SANDWICHES AVAILABLE
To: folks@SU-CSLI.ARPA


TWO TURKEY SANDWICHES ON LIGHT RYE WITH THE WORKS ARE AVAILABLE AT
THE FRONT DESK FOR A MERE $2.50 EACH.

COME AND GET IT! FIRST COME, FIRST FED!

THANK YOU, THE VENTURA SANDWICH CORP.
-------

∂20-Nov-85  1407	@SU-SUSHI.ARPA:PAPA@SU-SCORE.ARPA 	Talk by Victor Klee 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  14:07:28 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 20 Nov 85 14:02:00-PST
Date: Wed 20 Nov 85 14:01:55-PST
From: C. Papadimitriou <PAPA@SU-SCORE.ARPA>
Subject: Talk by Victor Klee
To: aflb.local@SU-SCORE.ARPA
Message-ID: <12160838663.47.PAPA@SU-SCORE.ARPA>

Victor Klee is giving a Colloquium today on:

``The Complexity of Linear Programming:  The d-step Conjecture After 30 Years''

The colloquium is in the Materials Science Auditorium (next to the
Earth Sciences building, room 550?), Thursday Nov. 22 at 4:30.

---CHP
-------

∂20-Nov-85  1502	@SU-SUSHI.ARPA:PAPA@SU-SCORE.ARPA 	V. Klee's talk 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  14:58:06 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 20 Nov 85 14:20:30-PST
Date: Wed 20 Nov 85 14:20:23-PST
From: C. Papadimitriou <PAPA@SU-SCORE.ARPA>
Subject: V. Klee's talk
To: aflb.local@SU-SCORE.ARPA
Message-ID: <12160842024.15.PAPA@SU-SCORE.ARPA>

Victor Klee's seminar is TODAY WEDNESDAY, and not tomorrow Thursday.
Sorry for the (multiply) contradictory message...
---Christos.
-------

∂20-Nov-85  1506	YAMARONE@SU-CSLI.ARPA 	HUNGRY?? TURKEY & AVO AVAILABLE 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  15:04:54 PST
Date: Wed 20 Nov 85 14:53:44-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: HUNGRY?? TURKEY & AVO AVAILABLE
To: FOLKS@SU-CSLI.ARPA


I KNOW THESE SANDWICH MESSAGES ARE STUPID AND A P.I.T.A...BUT
I'M NO WHEEL, SO PARDON ME.

THERE IS HOWEVER A DELICIOUS SANDWICH AVAILABLE TO THE FASTEST
PERSON TO RESPOND.

A SPECIAL TURKEY AND AVOCADO ON SLICED SOURDOUGH WITH EVERYTHING
EXCEPT MUSTARD.....$3.00 ....


COME TO THE FRONT DESK IF YOU'RE INTERESTED....

SORRY FOR THE INTERRUPTION, THE VENTURA SANDWICH CORP.
-------

∂20-Nov-85  1604	@SU-SUSHI.ARPA:SILVERBERG@SU-SCORE.ARPA 	Talk by Victor Klee
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  16:03:36 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 20 Nov 85 14:21:41-PST
Date: Wed 20 Nov 85 14:21:32-PST
From: Ellen Silverberg <SILVERBERG@SU-SCORE.ARPA>
Subject: Talk by Victor Klee
To: aflb.local@SU-SCORE.ARPA
Message-ID: <12160842234.10.SILVERBERG@SU-SCORE.ARPA>

Victor Klee's talk is today, Wednesday Nov. 20 at 4:30 (in case the previous 
message had you confused).  The room is 550A Materials Science Engineering.

Ellen
-------

∂20-Nov-85  1814	EMMA@SU-CSLI.ARPA 	Newsletter November 21, No. 3  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 20 Nov 85  18:14:43 PST
Date: Wed 20 Nov 85 17:05:29-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter November 21, No. 3
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
November 21, 1985               Stanford                        Vol. 3, No. 3
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, November 21, 1985

   12 noon		TINLunch
     Ventura Hall       Parsing as Deduction?
     Conference Room    by Fernando Pereira and David Warren
			Discussion led by Mark Johnson (Johnson@su-csli.arpa)
			(Abstract on page 2)

   2:15 p.m.		CSLI Seminar
     Redwood Hall	Interactive Modularity
     Room G-19		Ron Kaplan, Xerox PARC (Kaplan.pa@xerox.arpa)

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Redwood Hall	An Introduction to Information-based Complexity
     Room G-19		J. F. Traub, Computer Science Department, Columbia
                              ←←←←←←←←←←←←
                              ANNOUNCEMENT

   Please note that there will be no activities and no newsletter on
   Thursday, November 28, because of the Thanksgiving Holiday.  Thursday
   activities will resume on December 5.

!
Page 2                     CSLI Newsletter                   November 21, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                          THIS WEEK'S TINLUNCH
                          Parsing as Deduction?
                  by Fernando Pereira and David Warren

      Pereira and Warren's paper is exceptional in both its scope and its
   content.  It begins by proposing a translation of conventional phrase
   structure rules into (Horn clause) logic that can be given to a
   theorem prover, which then uses the logical translation to ``prove''
   the well-formedness of sentences with respect to the original grammar
   (hence the title, ``Parsing as Deduction'').
      Secondly, Pereira and Warren show how standard context-free parsing
   algorithms can be generalized as inference procedures that can
   ultimately be used to mimic parsers for certain non-context-free
   languages: thus showing us how to extend our parsing techniques for CF
   languages (which we know fairly well) to non-context-free languages in
   a straight- forward way.  Thus we can talk about ``the Earley parsing
   algorithm for LFG,'' for instance.
      And finally, they make some theoretical comparisons between the
   parsers so obtained for various different frameworks, and derive
   various properties regarding the parsing complexity.  Quite a lot for
   an eight-page paper!
      While not wanting to restrict discussion, I suggest that we
   concentrate on only one of these issues, namely the central claim that
   parsing can be viewed as deduction.  In what sense is it correct to do
   so?  Does it make sense computationally or psychologically?  Or
   linguistically or (dare I say it at CSLI) philosophically?
      Secondly, what about the logical translations that Pereira and
   Warren suggest?  Their translation into logical for a rule like

   	VP --> V NP PP

   is something like the following (expressed informally)

   	a V followed by an NP followed by a PP
   	implies the existence of a VP.

   But consider a sentence like

   	I saw the man with the telescope

   on the reading where the man, not me, had the telescope.  The antecedent
   of the logical translation of the rule is met, so the VP with the reading
   where I used the telescope to see the man should exist, simultaneously
   with the VP with the reading where the man has the telescope as well.
   That is, we are forced to infer the simultaneous existence of VPs
   corresponding to BOTH readings of the sentence.
      Is there a problem here?  And if so, why doesn't Pereira and
   Warren's  ``deductive'' run into problems with such ambiguous
   sentences?						--Mark Johnson 
!
Page 3                     CSLI Newsletter                  November 21, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
             ``Morphological Structure of Kunparlang Verbs''
                  Carolyn Coleman, (Coleman@csli.arpa)
             Thursday, November 21, 10 a.m., Conference Room

      Kunparlang verbs are extremely complex morphologically.  They
   cross- reference Subject and Object functions, incorporate nominal
   roots, use `applicative' derivational morphology, carry modal,
   directional and aspectual affixes, and inflect for Tense and Mood.
   There are two levels of hierarchical morphological structure:
    (i) The stem, which carries all morphology having compositional semantics.
   (ii) The lexical base, which caries all semantically idiosyncratic
        morphology.
      Kunparlang verbs undergo two types of reflexive operation which
   have a partially complementary distribution and which have different
   semantic effects on the verbs to which they apply.  With the first
   reflexive operation the reflexive subject is always an Actor; with the
   second the reflexive subject may be an Undergoer.  The second
   reflexive operation has a range of meanings which match those of
   mediopassive constructions in other languages as well as the reflexive
   reading.  Both reflexivizing operations are derivations that apply at
   the level of the lexical base; given that they have the same
   morphological status, there is a problem of how to semantically
   characterize them in a manner that will clearly show the semantic
   similarities and differences between them.		--Carolyn Coleman
                               ----------
                          PIXELS AND PREDICATES
                              Cleo Huggins
          Wednesday, November 27, 1 p.m., Ventura Seminar Room

      Graphic design is a profession that addresses problems with visual
   communication.  The issues that are covered by this area of work are
   significant in the design of symbols.
      There are problem solving methods used in design.  One is the use
   of semiotics, the study of signs.  I will describe how this field of
   study can be applied to the design process.
      Graphic designers also study the interaction of graphic elements.
   One might say that design is about harnessing these ``visual basics''
   and using them to reach a communication goal.  I will look at some of
   these elements and indicate how they are used in visual communication.
      Designers seldom work on the design of isolated symbols.  Often
   visual communication takes on the responsibility of systems.  Signage
   in a building and instructions/directions for a process, are examples
   of graphic systems.  In the design of a communication system one
   develops a graphic language.  Consistency is an important element in
   the design of this language.  I will look at a collection of computer
   related graphic elements and emphasize the role of design at a global
   level.
      I recommend this talk to anyone who:
   	* Has a tendency to redesign icons to make them witty or clever.
	* Believes that all graphic designers are advertising artists.
   	* Has their cousin design the graphic interface on their applications.
   	* Thinks that type is generally boring and would rather use
	exciting typefaces more often.
   	* Is interested in how graphic design may be integrated into
	work in other fields.
!
Page 4                     CSLI Newsletter                  November 21, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                              LOGIC SEMINAR
             Proving Properties of Destructive LISP Programs
                  Ian Mason, Philosophy Dept., Stanford
    Friday, November 22, 12:00-1:00, Math. Faculty Lounge, Room 383-N
                               ----------
                          PIXELS AND PREDICATES
                          Fragments of Behavior
                       Luca Cardelli, Digital SRC
           Wednesday, December 4, 1 p.m., Ventura Seminar Room

      The talk discusses some general issues concerning visual
   programming.  A particular user interface paradigm is then presented,
   where software tools can be visually assembled into larger and more
   complex tools. The basic abstraction is called ``Fragment of
   Behavior'' which is a thread of control with an interface (for
   connecting to other fragments) and an interaction protocol (for direct
   user interaction).  Composition of fragments involves both a
   composition of interfaces and of interaction protocols, and determines
   how the different fragments behave and interact concurrently.
      The goals are (1) allow different programmers to develop
   ``features'' of an application independently, including the user
   interfaces, (2) provide a library of very basic tools which users can
   custom-assemble, and (3) allow users to modify existing compound tools
   by adding, removing, or changing features.
                               ----------
                                SRI TALK
                          Unification Revisited
         Jean-Louis Lassez, IBM Thomas J. Watson Research Center
                  Monday, November 25, SRI, Room EJ242

      There are three main approaches to finitely represent sets of
   solutions of equations in the Herbrand Universe.  In Robinson's
   classical approach the set of solutions is represented by an mgu which
   is computed from the set of equations.  We introduce a dual approach,
   based on Plotkin's and Reynold's concept of anti-unification in which
   the finite representation (mgs) is now ``lifted'' from the set of
   solutions.  A third approach proposed by Colmerauer is based on the
   concept of eliminable variables.
   The relationships between these three approaches are established.
      This study provides an appropriate setting to address the problem
   of solving systems of equations and inequations which arises in recent
   extensions to Prolog.  A key result is that the meta-equation

                        E = E1 v E2 v ... v En

   admits solutions only in trivial cases.  Two important corollaries
   follow naturally.  The first is Colmerauer's property of independences
   of inequations.  This means that deciding whether a system of
   equations and inequations has solutions can be done in parallel.  The
   other corollary is a negative result; the set of solutions of a system
   of equations and inequations can be finitely represented by mgu's only
   in trivial cases.  Consequently, one cannot obtain a simplified system
   which is in ``solved'' form.  This is unlike the case when only
   equations are considered.  Similar properties hold in inductive
   inference when one attempts to generalize from sets of examples and
   counter-examples.
!
Page 5                     CSLI Newsletter                  November 21, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                         FOUNDATIONS OF GRAMMAR
               Foundations of Grammar group announces HUG

      The FOG group met last week to hear Lauri Karttunen report on HUG,
   a new development environment for unification-based grammars on Xerox
   1100 workstations.  HUG consists of four basic parts: (i) a
   unification package, (ii) input/output routines for directed graphs,
   (iii) an interpreter for rules and lexical entries, and (iv) an
   Earley-style chart parser.  All four are written in simple Interlisp-D
   for transportability to other dialects of Lisp.  The format for
   grammar rules and lexical entries in HUG is based on SRI's PATR
   system.  In addition to its generic core, HUG contains facilities for
   grammar development and debugging.  These routines take full advantage
   of the graphic capabilities of D-machines.
      The grammar formalism in HUG is based on PATR.  It is designed to
   make it easy to encode anything from simple phrase structure grammars
   to categorial grammars.  From the parser's point of view, a grammar
   rule is a single directed graph whose subparts correspond to syntactic
   constituents.  Lexical generalizations are expressed by means of
   templates and lexical rules as in PATR.  A Prolog-style treatment of
   long-distance dependencies is built in the system.
      HUG is now available for use at CSLI.  The documentation is
   currently in two sections.  HUG.DOC (11 pages) in {HEINLEIN:}<HUG>
   explains HUG's format for rules and lexical entries.  HUGTOOLS.DOC (24
   pages) is a user's manual.  A section HUG's parser and unification
   routine is in preparation.  For hard copies of these documents, see
   Carol Kiparsky (Carol@csli.arpa).
                               ----------
                        SUMMARY OF LOGIC SEMINAR
          Truth, the Liar, and Circular Propositions, Cont.
           Jon Barwise and John Etchemendy (Barwise@csli.arpa)
                      Friday, November 15, 1985

      The previous time John Etchemendy gave an informal introduction to
   Peter Aczel's set theory ZF/AFA, and showed how to use it to model the
   Austinian conception of proposition.  He then discussed how the Liar,
   the truth-teller, and other paradoxes and puzzles come out in this
   model.  This time, I reviewed ZF/AFA very briefly and then used it to
   model the Russellian conception of proposition, and discuss how the
   same puzzles come out in this model.			--Jon Barwise

   (The announcement of the above Logic Seminar was accidently omitted
   from last week's newsletter.)
-------

∂21-Nov-85  0933	EMMA@SU-CSLI.ARPA 	Newsletter addition  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  09:33:34 PST
Date: Thu 21 Nov 85 09:26:28-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter addition
To: friends@SU-CSLI.ARPA
Tel:  497-3479


The SRI talk by Jean-Louis Lassez on November 25 will be at 2:00.

-------

∂21-Nov-85  1100	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	International CS Institute 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  10:56:29 PST
Received: from su-aimvax.arpa by SU-SCORE.ARPA with TCP; Thu 21 Nov 85 10:44:56-PST
Received: by su-aimvax.arpa with Sendmail; Thu, 21 Nov 85 10:44:49 pst
Date: Thu, 21 Nov 85 10:44:49 pst
From: Jeff Ullman <ullman@diablo>
Subject: International CS Institute
To: faculty@score

I had a visit from Ronald Kay yesterday, and apparently the
GMD-funded "International Inst. of CS" is moving along seriously.
We are among 5 sites being considered, and there will be discussions
on Jan. 8 with Kay and GMD director Szyperski.  The outline that
I was able to get from Kay is that the institute would be
independent, initially with funding of 3M/year.
A fraction of that would support postdoctoral visitors from Germany,
who would work on projects also funded from that 3M.

I suspect that in the long run, the directors of the institute
would determine whose projects were funded, and these might be
anywhere.  The quid-pro-quo for helping to set it up and hire
the first management would be that the inst. would be located
in Palo Alto and its management and guiding committee would
be dominated by SU people.  Further, the initial projects funded
would be SU projects.

Thus, by Jan 8 we need to get together a few, no more than 3,
proposed projects.  There should be room for some postdoctoral
visitors, and from what I can gather, Kay and Szyperski would
most like to see new projects that cut across (sub)discipline lines.

I think Nils is going to try to run a discussion of possibilities
next Tuesday, at lunch.  I'm going to try to "spearhead" this
proposal if there is sufficient interest that it might fly, so
I'd be happy to know of interest beforehand.
Ernst has some information on GMD which he is going to give
to Rosemary Napier, if anyone is curious.
				---Jeff

∂21-Nov-85  1134	RICE@SUMEX-AIM.ARPA 	3600 -> Explorer Update.
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  11:31:57 PST
Date: Thu 21 Nov 85 11:31:02-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: 3600 -> Explorer Update.
To: aap@SUMEX-AIM.ARPA
Message-ID: <12161073339.30.RICE@SUMEX-AIM.ARPA>


Czar seems to be working now.

Q-Scheme works as well on the Explorers as it does on the 3600s.

Rice.
-------

∂21-Nov-85  1258	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #37  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  12:57:29 PST
Date: 21 Nov 85 1245-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #37
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Thursday, 21 Nov 1985      Volume 1 : Issue 37

Today's Topics:

                      SIMD vs MIMD (2 messages)
                     Functional languages on Cray
                           Job announcement

----------------------------------------------------------------------

Date: Wednesday, 20 November 1985  06:21-PST
From: Bradley C. Kuszmaul <bradley at THINK.COM>
Subject: MIMD vs SIMD - A response I receive directly

The following response was delivered directly to me, and M. Li tells
me that it is ok to send it along with my response on to Parsym.  His
meta-response is forthcoming.
-Brad

 From: Tony Li <tli@usc-eclb>

 Yow!  I can't resist that challenge.  I can resist sending it to
 Parsym for a bit though.

 For a start, I'm not sure that I agree with your definition of SIMD.
 In your M68000 SIMD machine, there is no real synchronization of
 instructions.  Even with the common clock, and one macro-instruction,
 your machine is not SIMD in the common sense.  Each processor would
 NOT be doing the same thing at exactly the same time.  I cite the
 definition of SIMD in [Stone, Intro to Computer Architecture, pg.
 321].

 In particular, as soon as you begin evaluation of EVAL on all
 processors, they will be in sync.  However, as soon as you hit a COND,
 things will start to happen differently.  A traditional SIMD machine
 will execute all parts of the COND, disabling processors until they
 find a valid branch.  Your machine (unless you're doing something that
 you're not telling me about) is executing different instructions on
 different processors.  This is exactly a MIMD machine (Multiple
 instruction streams).  In some sense, it is irrelevant that all of the
 streams started out as a single program.  Since each processor will
 make its own path (possibly unique) through the code, there are
 definitely multiple instruction streams.


  On the other hand you could conceivably be disabling processors in the
  traditional SIMD (i.e. ILLIAC IV) case.  In this case, I argue that
  any application which will have substantial computation done by one
  branch of a COND will be inefficient.


Cheers,
Tony ;-)


But I DEFINED the instruction set to be the singleton containing the
EVAL instruction.  I don't understand what you mean by "your machine ...
is executing different instructions on different processors".  At any
given time all the processors are running the EVAL instruction.

I don't really think of a network of 68000's as being SIMD, but it is
possible to argue that they are.  The point is that it is hard to draw
the line between SIMD and MIMD.

Perhaps if I start in the domain that definitely seems SIMDish and
gradually move towards MIMD the problem of the beard will become more
apparent (i.e. one whisker is not a beard, two is not a beard, but 4096
whiskers is a beard.   Is 2048 a beard?  If so is 1024 a beard?  If so
is 512 a beard?  Where is the dividing line?  The same question can be
asked about MIMD.  We know that a net of 68000's is not SIMD and that a
connection machine is SIMD.  Suppose that we "enhance" the connection
machine instruction set....)

You claim that traditional SIMD machines are allowed to either perform
the current instruction or do nothing (i.e. be disabled).  The problem
is that even when all the processors are enabled and all are doing the
same thing (e.g. a bitwise AND operation), the resulting state of each
processor is a function not only of the instruction being broadcast, but
also of the data that was there to start with.  Now suppose that we make
the instruction set a little more complex.  For example we might see
instructions like
  "All processors with a 0 in bit X perform an add, others perform a
   multiply"
  It seems clear that a machine which runs instructions of
this type could be built.  The required bandwidth of the instruction
stream might be increased (e.g. we might supply two of the old
instructions at one time, all processors with a 0 in bit X would perform
the first instruction, all processors with a 1 in bit X would perform
the second instruction).  This mechanism still seems to be SIMD (perhaps
2IMD - two instruction multiple data) because the architecture is very
similar to the ILLIAC-IV in that the instruction set is being broadcast
from some central location, and the local location is not doing control
(i..e. no goto's), it is just doing data manipulation)

Now suppose that I really increase the bandwidth of the instruction
stream to the point where I have the following instructions
   "Consider the eight bit field starting at bit X.
      All processors with 0 in those eight bits do a NOP,
      All processors with 1 in those eight bits do an ADD,
         '  '  '  '       2  ' ' '   '   '  '   ' '   MULTIPLY,
       ....                        ...
        .
        .
        .
      All processors with 256 in those eight bits do a FLOAT MULTIPLY"
Now I have really increased the bandwidth of the instruction set,
because I need to provide 256 instructions to every processor all the
time.  There may be implementation difficulties with this scheme, but if
I had a lot of pins on my VLSI chips, I might be able to shove 256
instructions into the chip every clock cycle, each processor could chose
which instruction it will execute bsed on the bits starting at X.  Now I
DEFINE a single instruction to be 256 of the old instructions, and I
still have what seems to be a SIMD machine.

I can add a little more hair, and have every processor get the value X
indirectly out of location 0, and change the instruction set so that we
might see the following fragment:
    "Let X=<0> in every processor
      All processors with 0 in <X> do a NOP and increment <0>
      All processors with 1 in <X> do a ADD and increment <0>
       ...
      All processors with 155 in <X> set <0> to be the value computed
            during the previous instruction
       ..."
And it still seems like a SIMD machine.  However `opcode' 155 looks
suspiciously like a GOTO in that <0> is set to some arbitrary value, and
on the next instruction, we will be looking at <0> to see what
instruction to execute.  It is starting to look more like a MIMD
machine, but given my DEFINITION, this is still a SIMD machine (with a
huge instruction bandwidth)

Now suppose I notice that I always broadcast the same huge instruction.
I.e. that `opcode' 0 always means a NOP, `opcode' 1 always means ADD,
and so on.  I have not lost any of the computational power of the
original machine since the `goto' like operation gives me a lot of
power.  I, as a hardware engineer, might decide to solve all my
pin-boundedness problems by hard wiring that instruction into the
processors.  I might as well since I always broadcast the same thing.
Once I have hardwired that instruction into the chip I have dramatically
reduced the instruction bandwidth, I only need to send the EVAL opcode
which takes no bits to represent.  Now I have a network that looks like
a MIMD machine but it is a SIMD machine by the argument of the beard.

Now that I understand how to simulate any MIMD machine by building a
special SIMD machine, I can know how to simulate any MIMD machine with any
SIMD machine.  I select all processors that want to run opcode 0 and
then broadcast the opcode 0 instruction.  Then I select all processors
that want to run opcode 1 and broadcast the opcode 1 instruction, and so
on.  I can perform this simulation with a performance degradation of
only a factor of 256.  I wanted to keep my processors busy doing useful
work, so if I say "a processor is busy enough if it is doing useful
work 1/256 of the time" then I am all set.

------------------------------

Date: Thursday, 21 November 1985  12:31-PST
From: AGRE%MIT-OZ at MIT-MC.ARPA
Subject: simd and mimd

Seth Steinberg mentions work Alan Bawden and I did writing a Scheme
interpreter for the Connection Machine.  Although Seth's comments are to the
point, I want to make sure nobody thinks we actually intended Scheme as a
serious CM programming language.  The Scheme interpreter was just what we
wrote to try out Alan's first CM language, CL1.  Having been in MIT Scheme
culture for so many years we figured we knew every possible variant of the
algorithm.  (We were wrong, by the way.  A CM Scheme interpreter seems to
need an eccentric modularity if arguments are to be evaluated in parallel
without unnecessarily tying up the processors implementing the source code.
Weird.)

We learned some things about programming fine-grained parallel architectures
in general, but I'll just mention what we learned about programming SIMD
machines in particular.  The idea behind running a Scheme interpreter (or
any interpreter) on a SIMD machine is to simulate a MIMD machine.  If this
is what you want then Scheme is a bad language for it, for three reasons.
One is that you need to multiplex the instruction stream among too many
hunks of interpreter code, one per cell type.  (This turns out not to be so
bad if you're careful.)  Another is that the primitives are so small that
any decent program takes 50 processors to implement.  (Other languages might
be better on this count.)

The third problem with using a Lisp-like language to program a SIMD machine
is the hardest.  The Lisp interpreter, and most Lisp programs as well, does
a tremendous amount of dynamic storage allocation.  There are gazillions of
cons nodes and stack frames, with a median lifetime near zero.  It would be
nice to avoid all this constant reallocation of processors, for a number of
reasons.  One is that a rapidly changing process structure is hard to debug.
Another is that the unpredictability of resource requirements makes load
balancing difficult.  By the time you do a hairy analysis to find which
processes are competing for resources the processes have finished.  Another
is that storage management takes time.  The published algorithms for the CM
are pretty poor; presumably you can do better but I'll have to be convinced.
A fourth is that there is a substantial overhead each time you have to set
up a new pattern of router connections.  If the connection pattern was
relatively static you could store common router connectivities and change
them incrementally.

These considerations argue for a style of programming in which processors
are assigned roughly permanent functions and their patterns of connectivity
are roughly static.  If you take this to an extreme you get connectionism
(more the Rochester sort than the CMU sort).  Whether we need to go this far
remains to be seen.  Of course the answer to "SIMD vs MIMD" is "SIMD is
better for some things and MIMD is good for others", so we should see what
we can do best with relatively static process organizations.  This isn't the
time for detailed speculations on the subject, but here's a hint.  Almost
all of the dynamically allocated processors our Scheme interpreter used in
running a program were variable binding nodes.  (In a related project,
William Foster verified this assertion statistically for an ACT interpreter
he wrote in CL1.)  To a first approximation, you need to do dynamic storage
allocation when you need to bind variables to values.  How hard is it to
program without variables?

/phil

------------------------------

Date: Wednesday, 20 November 1985  09:05-PST
From: Paul Kelly <mcvax!kcl-cs!pkelly at seismo.CSS.GOV>
Subject: PARSYM Digest   V1 #31

Re: Performance of Lisp on Crays etc.

There is some work going on to produce optimising compilers for functional
languages, to run on Crays.

S.K. Skedzielewski and M.L. Welcome of Lawrence Livermore National
Laboratory, USA, report in their paper "Data Flow Graph Optimisation
in IF1" (in LNCS 201, "Functional Programming and Computer
Architecture", Nancy, 1985), their efforts to compile the
single-assignment language SISAL to Cray code via a standard graphical
intermediate form, IF1.

A single assignment language seems to be (correct me if I'm wrong), a
functional language with syntactic constructs for forms common in
numerical programming.

The paper doesn't give any performance figures relative to other
languages, or machines.  Perhaps the authors can be persuaded to tell
us how well it worked?

Yours, Paul Kelly

Dept of Computing, King's College, London, UK.

------------------------------

Date: Wednesday, 20 November 1985  09:12-PST
From: Alfred Hartmann <alfred%mcc-pp at mcc.arpa>
Subject: Position announcement for PARSYM

Microelectronics and Computer Technology Corporation - Position Announcement
----------------------------------------------------------------------------

Parallel Processing Theoretician --

The MCC Parallel Processing Program is seeking a senior research staff
member with a working knowledge of computational complexity theory
(both traditional algorithmic and VLSI theories) who desires to
explore unifying principles and fundamental limits for abstract
parallel processing models.  The individual we seek has a strong
theoretical background (especially in the rapidly developing area of
VLSI complexity theory) with a desire to extend existing theories
towards a general theory of parallel processing architecture supplying
asymptotic limits on system power, density (in three dimensions),
interconnection network bandwidths, latencies, and layouts, locality,
efficiency, resource utilization, etc. The ultimate goal is the
development of a theory of parallel processing that establishes
computational metrics and fundamental limits upon the metrics, to be
used in evaluating new paradigms of parallel computation (intended to
replace the von Neumann paradigm of sequential computation).

Typical qualifications would be a doctorate in computer science or
applied mathematics, demonstrated competence in the field, and strong
academic and research associations.  Frequent interaction with applied
research and applications in computer architecture, parallel
algorithms, performance analysis and prediction, and computer-aided
design should be expected as part of this position.  Recent doctorate
earners with worthy dissertation work in this area are also encouraged
to apply.

Reply with vitae to:  George Black, VP and Director
                      Human Resources
                      MCC
                      9430 Research Blvd., Ech. I/200
                      Austin, TX 78759
                      (512) 343-0860

------------------------------

End of PARSYM Digest
********************

∂21-Nov-85  1525	@MIT-MC.ARPA:JCMA@MIT-MC.ARPA 	Classic Quotes (Positivism Assesses Progress in Meaning)   
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  15:24:17 PST
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 21 NOV 85  18:25:36 EST
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 21 Nov 85 18:21-EST
Date: Thu, 21 Nov 85 18:23:52 EST
From: "John C. Mallery" <JCMA@MIT-MC.ARPA>
Subject:  Classic Quotes (Positivism Assesses Progress in Meaning)
To: MINSKY@MIT-MC.ARPA
cc: PHIL-SCI@MIT-MC.ARPA
Message-ID: <[MIT-MC.ARPA].727439.851121.JCMA>

``In sum half a century of talk about meaning has been useless to
scientists and has, if anything, increased confusion among
philosophers.'' 

-- Mario Bunge, Chapter on Meaning in his Treatise on Basic Philosophy, 
Volume 2, Semantics II: Interpretation and Truth, Boston: D. Rediel,
1974, p.43.

∂21-Nov-85  1541	@MIT-MC.ARPA:JCMA@MIT-MC.ARPA 	Classic Quotes (Positivism Assesses Progress in Meaning)   
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  15:39:09 PST
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 21 NOV 85  18:25:36 EST
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 21 Nov 85 18:21-EST
Date: Thu, 21 Nov 85 18:23:52 EST
From: "John C. Mallery" <JCMA@MIT-MC.ARPA>
Subject:  Classic Quotes (Positivism Assesses Progress in Meaning)
To: MINSKY@MIT-MC.ARPA
cc: PHIL-SCI@MIT-MC.ARPA
Message-ID: <[MIT-MC.ARPA].727439.851121.JCMA>

``In sum half a century of talk about meaning has been useless to
scientists and has, if anything, increased confusion among
philosophers.'' 

-- Mario Bunge, Chapter on Meaning in his Treatise on Basic Philosophy, 
Volume 2, Semantics II: Interpretation and Truth, Boston: D. Rediel,
1974, p.43.

∂21-Nov-85  1639	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH -- Haim Gaifman 
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  16:32:58 PST
Date: Thu 21 Nov 85 16:22:22-PST
From: LANSKY@SRI-AI.ARPA
Subject: Next Monday's PLANLUNCH -- Haim Gaifman
To: planlunch.dis: ;
cc: halpern@IBM-SJ.ARPA, moses@IBM-SJ.ARPA

		 LOGIC OF POINTERS AND EVALUATIONS: 
	       THE SOLUTION TO THE SEMANTIC PARADOXES
				
                            Haim Gaifman  (GAIFMAN@SRI-AI)
          Mathematics Department, Hebrew University, Jerusalem
 	    	     Visiting SRI International, AI Center

	 	    11:00 AM, MONDAY, November 25
       SRI International, Building E, Room EJ228 (new conference room)

ABSTRACT:

Imagine the following exchange:

Max:          What I am saying at this very moment is nonsense
Moritz:       Yes, what you have just said is nonsense.

Evidently Max spoke nonsense and Moritz spoke to the point.
This is rather puzzling, because Max and Moritz appear to have asserted
the same thing, namely: that Max spoke nonsense. 
A philosophically more refined  version of the puzzle consists in exhibiting 
the following two-line text:

line 1:    The sentence written on line 1 is not true.
line 2:    The sentence written on line 1 is not true.

Our natural intuition is that the self-referring sentence on line 1
is not true (whatever sense could be made of it). Therefore the sentence
on line 2, which asserts this very fact, should be true. But what is written on
line 2 is exactly the same as what is written on line 1. 
The very same sentence is judged differently on its two different occurences. 

I shall argue that the unavoidable conclusion is that truth values
should be assigned here to sentence-tokens (i.e., the particular occurences of
sentences). In systems (like Kripke's) which assign
only type-dependent truth values one cannot express some facts 
easily expressible in common daily discourse. Terrible things	
happen in such systems like Holes and even Black Holes.

I shall present a formalism which distinguishes between tokens and
a method of evaluating their truth values (including gaps)
which takes into account the whole network of pointings. 
One soon realizes that one should treat more general entities than tokens.
These are the ``POINTERS'' referred to in the title of my talk.
A pointer is any object which is used to point to a sentence (i.e.,
sentence-type). A sentence-token is a special case of a pointer, it points
to the sentence of which it is a token. The basic formalism  is extremely
simple. It can serve as a basis  for  various extensions (with quantifiers 
over pointers) and in all cases the resulting language will contain a truth
predicate taking as arguments pointers to its own  sentences. There are also
versions with additional modality and epistemic predicates taking pointers
as arguments.

The term ``EVALUATION'' refers to the method by which
truth values are associated with pointers. A basic guiding heuristic	
is to regard pointers as names of procedures in some computer program.
The procedure is started by calling the pointer which in turn may call
other pointers, etc. Eventually it returns a value,  TRUE, FALSE or GAP. 

The procedure is such that, calling `the sentence on line 1' returns GAP, but
calling `the sentence on line 2' returns  TRUE.  It yields
similarly acceptable results in self referential situations in general, where
any number of pointers can be involved in nasty complicatd loops.
 
I shall discuss various related topics, other semantic paradoxes, and
show that the smallest number undefinable in less than twenty words
is alive and well and can be defined after all.
 
P. S. Every sentence in this abstract is of course true.
Question: How about this last very same sentence?

-------

∂21-Nov-85  1704	ullman@su-aimvax.arpa 	paper recieved   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  17:04:04 PST
Received: by su-aimvax.arpa with Sendmail; Thu, 21 Nov 85 17:01:48 pst
Date: Thu, 21 Nov 85 17:01:48 pst
From: Jeff Ullman <ullman@diablo>
Subject: paper recieved
To: nail@diablo

"The Alexander method, a technique for the processing of recursive
axioms in deductive databases," by Rohmer and Lescoeur, of Bull
Research (that's Machines Bull!).

Apparently the Alex. referred to is "The Great," as in Gordian knot.
It seems to be another version of "magic sets".

∂21-Nov-85  1721	ullman@su-aimvax.arpa 	vowel fans take note  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  17:16:02 PST
Received: by su-aimvax.arpa with Sendmail; Thu, 21 Nov 85 17:11:17 pst
Date: Thu, 21 Nov 85 17:11:17 pst
From: Jeff Ullman <ullman@diablo>
Subject: vowel fans take note
To: nail@diablo

of the weird configuration in the second author's name, on that
last paper.

Also, I noted in the references to that paper I am listed
as "J. D. Shapiro"  close enough...

∂21-Nov-85  1754	WINOGRAD@SU-CSLI.ARPA 	No ENVIRONMENTS meeting until Dec 9  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  17:54:12 PST
Date: Thu 21 Nov 85 17:47:16-PST
From: Terry Winograd <WINOGRAD@SU-CSLI.ARPA>
Subject: No ENVIRONMENTS meeting until Dec 9
To: friends@SU-CSLI.ARPA, su-bboards@SU-CSLI.ARPA

Sorry to have missed the newsletter with this.  There will
be no meeting for the next two weeks.  We will resume with
Danny Bobrow (Xerox) on the handling of object storage in LISP
(with COMMONLOOPS) on Dec 9, then Ron Kaplan (Xerox and CSLI) on
the grammar writer's workbench on Dec 16.
--t
-------

∂21-Nov-85  2104	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	ADVANCED COURSE ON PETRI NETS  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 21 Nov 85  20:56:10 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Thu 21 Nov 85 20:54:09-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Thu 21 Nov 85 20:53:31-PST
Received: by rsch.wisc.edu; Thu, 21 Nov 85 22:36:19 CST
Date: Thu, 21 Nov 85 22:26:54 CST
From: udi@rsch.wisc.edu (Udi Manber)
Message-Id: <8511220426.AA16743@rsch.wisc.edu>
Received: by rsch.wisc.edu; Thu, 21 Nov 85 22:26:54 CST
Subject: ADVANCED COURSE ON PETRI NETS
To: udi@rsch.wisc.edu
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 21 Nov 85 22:36:06 CST (Thu)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

Sent by request of Dr. W. Reisig, GMD:

Preliminary Announcement


                     ADVANCED COURSE ON PETRI NETS

	Theory     Applications     Relationship to Other Models

	     Bad Honnef, West Germany, September 8-19, 1986


Petri Nets represent a long and sustained effort to develop concepts,
theories and tools to aid in distributed systems design and analysis.
They are used in many areas of computer science including software
engineering, data base and information systems, computer architecture
and operating systems, communication protocols and computer networks,
process control, and socio-technical systems such as office communication
and man-machine interaction. A rich body of theory has been developed
for Petri Nets. It reflects all major problem areas of distributed
systems and covers many successfully applied principles and analysis
techniques for systems organization.

The course addresses those who are

  interested in systems design and would like to use Petri Nets,

  familiar with subareas of the theory or the applications of
  nets and wish to become acquainted with the whole area,

  interested in learning about recent results presented within a
  uniform framework,

  wanting to learn about successfully applying Petri Nets in
  various practical situations.

The course is a successor of the ``Advanced Course on General Net Theory
of Processes and Systems'' held in Hamburg in October 1979.

There will be introductory lectures (accompanied by exercise sections),
survey talks on the most relevant issues as well as presentations on
recent results in theory and applications.

The course will also discuss relationships to other models of concurrent
systems such as COSY, CCS, CSP and ACP. In the field of general systems
design (requirements engineering) we relate nets to techniques such as
SADT. 

Several computer-aided tools for editing and analyzing nets will also be
presented.

Also an opportunity will be provided for the participants to discuss
their research problems and results.

Lectures will be given by leading experts in the area. The course will
take place in Bad Honnef, near Bonn, Germany, from September 8th to
September 19th, 1986.

Directors of the course are:
	Prof. Dr. W. Brauer, Hamburg University, West Germany,
	Dr. W. Reisig, GMD Bonn, West Germany,
	Prof. Dr. G. Rozenberg, Leiden University, The Netherlands.

Scientific advisors of the course are:
	Prof. Dr. G. Roucairol, BULL, Paris, France,
	Dr. P.S. Thiagarajan, Aarhus University, Denmark.

The course will be organized by the Gesellschaft fur Mathematik und
Datenverarbeitung (GMD) under the auspicies of AFCET, AICA, BCS, EATCS,
and GI.

For details and application forms please contact:
Dr. W. Reisig, GMD - Fl, Postfach 1240, D-5205 Sankt Augustin 1, West
Germany; Tel. (02241) 14-2797.


--------------
TN Message #10
--------------

∂22-Nov-85  1248	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #38  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 22 Nov 85  12:47:52 PST
Date: 22 Nov 85 1218-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #38
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Friday, 22 Nov 1985       Volume 1 : Issue 38

Today's Topic:

                            MIMD vs. SIMD

----------------------------------------------------------------------

Date: Thursday, 21 November 1985  12:58-PST
From: Sal Stolfo <sal at CS.COLUMBIA.EDU>
Subject: PARSYM Digest   V1 #36

Reply to: the SIMD Camp
From: The MIMD&SIMD Camp


Its probably true that a programmer (or user) can be protected from
most or all details of any machine (thank God).  OH, the poor
developer, however!

It is not sufficient to say that the code for the SIMD machine for the
unification problem I posed is the same as for a sequential machine
(modulo intermittent code sequences to disengage processors from the
"else streams" and the "then streams" in conditional expressions).

There were two issues being raised:
 1- The unification problem might require the execution of every
possible branch in the program many times because at each branch point
every possible type of object might be present in the SIMD machine.

 2- The SIMD program (be it by the hand of a system hacker or a user)
must concern itself at some level sometime with data spanning across
several (possibly many) PE's.  In a sequential machine you add one to
an address for the next word of data, or load an index register with a
pointer from memory.  That must be done.  In an SIMD machine with data
spanning over several PE's, a similar operation must be done which
includes bit passing and disabling instructions in addition to address
computation.

The second issue lies at the heart of the "granularity problem".  It is
probably true that if the "task granularity" is larger than the
machine's PE granularity, overhead must be paid by the mismatch.  It is
probably also true that if the task granularity is much smaller than
the machine's PE granularity, then parallelism inherent in the task will
be lost (although hardware utilizaton can be quite high).  There is a
continum here that waits to be clearly understood from a computational
complexity perspective.  The bottom line is, however, that whatever PE
granularity you choose and fix in your machine, you can probably find
a grain-mismatched task that performs oh so poorly.  Note in the
extreme case a uniprocessor has a severe grain mismatch with
inherently parallel tasks resulting in maximum utilization and
absolute minimum parallelism.

The remark I made about "speed naming the game" is indeed too
general.  "Cost/performance" should be our champion.  Thus, relative
cost of SIMD and MIMD devices in terms of DELIVERED MIPS per dollar is
probably the true name of the game.  What I mean by DELIVERED MIP is
that the MIP should be measured on the basis of number of useful
objects or results computed per second for the application of
interest.  Thus, one should not rate in terms of billion of bit
operations per second, but factor out all of the overhead instructions
burned and lossage due to poor grain fitting to derive the actual
delivery time of useful results.  Caution on both sides is in order.
At the risk of not taking my own advice, here is another nuance.

One argument I heard repeatedly is the savings a SIMD PE enjoys by
amortizing one functional unit over several processing sites.  It might
be advantageous, however, to trade in all of the area consumed by the
wires to connect the multiple processing sites for a larger, more
powerful, single functional unit that operates on wider data or
performs more useful instructions in one duty cycle.  It may,
therefore, be possible that 4 concurrent processing sites would each
have to execute say 4 instructions to equal one instruction by the
single larger processing site.  In 4 machine cycles, both compute the
same 4 results.

Let me say this about silicon cost.  Silicon area alone is insufficient
to judge cost of a device.  A square centimeter's cost may be
insignificant due to amortization, or extremely expensive due to
development effort or cooling requirement.  Judging the cost per MIP of
two devices is thus not easily measured by counting square
centimeters.  The cost of fabrics by the yard varies according to the
cost of the raw material and labor.  Computing by the silicon yard is
no different.

In summary, the large number of parameters involved and the lack of
experience and knowledge available about these issues will undoubtedly
work hard at maintaining this SIMD vs MIMD controversy for some time.

(Parenthetically, I'm a DADO follower.)


Sal Stolfo

------------------------------

Date: Thursday, 21 November 1985  14:45-PST
From: HENRY%MIT-OZ at MIT-MC.ARPA
Subject: SIMD vs. MIMD

 GLS: I think that SIMD appears to many people compare unfavorably to
 MIMD because it is tempting, in the mind's eye, to picture a single
 processor, whether SIMD or MIMD, as being of about the same size.  I
 think we tend to mentally compare a 1000-processor SIMD machine with a
 1000-processor MIMD machine. 

GLS's point is well taken -- we tend to avoid mentally factoring in
the cost of processors in terms of their hardware size.  There's
another side to this, however, which tends to be unfair to MIMD.
Again, since we are using the word "processor" in both cases, we tend
to assume that the processors are of equal computational POWER.  In
the case of connectionist SIMD machines, the processors may be so
small that it may well take a group of these processors to do some
given computational task that a single, larger processor could do.
[Is a Vax a 32-processor SIMD machine?]

My own view is that SIMD vs. MIMD is that it will be a short-term vs.
long-term tradeoff.  In the short term, SIMD and shared memory
parallel processors will have a cost advantage, especially for
specialized applications with static process and communication
patterns.  In the long term, with more capable processors, more
parallelism and less structured applications, MIMD will win out.

------------------------------

Date: Thursday, 21 November 1985  22:14-PST
From: Simon Kasif <simon at mimsy.umd.edu>
Subject: MIMD vs SIMD

              A Non-Folk Theorem on SIMD = MIMD Question

A common folklore in the parallel computation circles asserts that
highly regular computations such as matrix multiplication are most
suitable for SIMD while symbol manipulation is best done on MIMD
machines.  The examples given in the MIMD vs SIMD discussion certainly
conformed to this folk tale.

I will argue that this is not necessarily the case.

To prove my point I will use the degree of freedom in the SIMD
definition.  Specifically, it is possible to define an SIMD
computation using a shared memory concept (called PRAM in the theory
community and ultracomputer in New York).  Now, don't get alarmed.  To
achieve practicality, let us assume that we have P processors and M
memories interconnected using a (switching) NETWORK.  This SIMD model
is somehow more general than the one where each PE has its own memory.

From now on things are easy.  Consider the unification problem
suggested by Sal Stolfo where one needs to unify P terms with a given
term.  At each point of execution each processor performs a REGULAR
operation, namely unifying one atomic term with another (or applying a
substitution to an atomic term).

Data Flow languages also can naturally be mapped to this formalism (as
noted by Jack Schwartz and A. Gotlieb).  Consider any state of the
data flow graph: Node X passes token MESSAGE(X,Y) to node Y.  As
before, this is also a highly regular operation since it can be
implemented in our model by copying MESSAGE(X,Y) from location X to
location Y.

In fact, ANY rewriting system can be naturally mapped to this model by
having every processor always perform a reduction of one term to
another.  This includes functional languages (FP, pure LISP), logic
programming languages (pure Prolog, Equational Logic, Concurrent
Prolog, PRISM, Tablog).  Needless to say that currently these
languages are the main candidates for supporting so called AI
applications.

Efficiency considerations in this model of computation create the
problem of distributing information on the memories to minimize memory
conflicts.  This issue is independent of the SIMD = MIMD question.

			Q.E.D (or where did I go wrong)



A CLOSING REMARK:

The observations above support the messages by Brad and V. Pratt.
That is, the question of MIMD vs SIMD is really a question of the
level we are observing the system from.  If we look down enough, every
systen looks like it is sending a signal from one gate to another:
SIMD.

Personally, I am a great favorite of MIMD machines for the following
reasons:

1. Flexibility: one can create a MIMD machines by simply connecting
   several SIMD machines (uniprocessors).
   Example: Consider I/O processors and a CPU: a clear instance of
   MIMD.

2. Multi-user environments are clearly easier implemented on SIMD
   machines.

3. Two brains are better than one.

------------------------------

Date: Friday, 22 November 1985  02:07-PST
From: Alan Bawden <ALAN at MIT-MC.ARPA>
Subject: SIMD/MIMD and all that jazz

In the context of the current SIMD/MIMD discussion Seth Steinberg
mentioned the paper Phil Agre and I wrote about the Scheme interpreter
we implemented in the experimental connection machine language CL1.
In that paper Phil and I were not trying to address the issues of SIMD
vs. MIMD machines or SIMD vs. MIMD programming languages.
Unfortunately we inadvertently brought attention to those issues by
choosing a Scheme interpreter as our sample program.  I doubt if
anyone has ever read our paper without becoming distracted by the
SIMD/MIMD issue.  If I had to do that bit of research over again
today, I would choose to implement something more mundane as a
demonstration project.  Say a text justifier or an assembler or
something equally dull.  (There are other flaws in the paper that make
it hard to understand.  Perhaps someday we will revise it.)

Although we did not set out to learn about SIMD vs. MIMD, everybody
wants to talk to us about it.  And Phil did think about some of the
issues while he was implementing the Scheme interpreter.  Consequently
we now -do- have something to say about it, even though we didn't
write about it in the paper Seth referred to.  In fact, we have two
-different- things to say.  Phil already made his point in a separate
message to this forum.  I have a couple of things to say about the
issues of simulating a MIMD architecture with an interpreter on a SIMD
machine.


A Little Background: CL1

The tool we used to implement the Scheme interpreter is an
experimental low-level connection machine programming language called
CL1.  Datatypes in CL1 consist of atomic types such as numbers and
addresses and small record structures that can be built out of atomic
objects.  CL1 has primitives for performing operations such as
arithmetic and primitives for sending and receiving messages.

The CL1 compiler processes a collection of descriptions (written in
the Lisp-like CL1 notation) of the behavior of various classes of
processing elements.  The output of the compiler can be thought of as
a finite state machine describing every state a processing element can
be in at runtime, when each state-transition is legal, and how each
state-transition is to be effected.  ("Every state" does not mean all
the 2↑400 states a processing element with 400 bits of storage can be
in.  States have associated state variables that can contain various
values at runtime.  For an estimate of the number of states I am
talking about, go and count the labels in an assembly language
program.)

On a SIMD machine (such as a connection machine) such a FSM can be
executed by cycling through each state broadcasting instructions for
the benefit of processing elements that happen to be in that state.
You can think of this as multiplexing the single instruction bus of
the SIMD machine.  If the FSM has N states, and each state has about
the same amount of associated code (a good estimate), and you
broadcast instructions in the most simpleminded way, then each
processing element only runs about 1/N of the time.

On a MIMD machine such a FSM can be executed by giving every processor
a local copy of the code which it can use to fetch the instructions
appropriate for its current state.  You pay for this convenience in
hardware of course, since each processing element needs to have some
local instruction memory, addressing logic, etc.  Also, the bigger the
program you plan to run, the more storage you need at each processing
element.


But SIMD Need Not Be That Slow: A Digression About Scheduling

At first glance it looks as if a SIMD machine will always execute an N
state FSM at 1/N'th the speed of the same FSM on a MIMD machine,
although the SIMD machine will be smaller and easier to engineer.  It
does not need to be this bad, however.  Given a single bit of feedback
from the processing elements (such as the Global OR flag on a
connection machine), you can avoid broadcasting all of the
instructions for every state.  If you broadcast the instructions: "If
you are in state #259, put a 1 on your Global OR output", and the
output of the Global OR network is 0, then you can skip broadcasting
the instructions for state #259.  Hopefully there will be lots of
unpopulated states at any given moment.

Furthermore, analysis of the FSM should enable you to skip even
considering broadcasting the instructions for certain states.  If
nobody was in state #259 a moment ago, and you just broadcasted the
code for state #17 for which the only possibly transitions are from
#17 to #23 and #17 to #1729, then you -know- that no new cells have
entered state #259.  You don't even need to ask.  A good model of what
states are populated at what time, kept at runtime, coupled with an
analysis of the FSM, performed at compile time, might make a
significant reduction in the factor of N.

Even cleverer algorithms can be devised for improving the performance
of a SIMD machine executing such a FSM program.  We might imagine ways
of deciding what state to schedule next that would tend to cluster
processing elements together in the state graph.  This might tend to
increase the number of processing elements listening to the
instruction bus at any given moment, which would effectively reduce
that factor of N.

Nothing in this section represents more than speculation on my part.
I would be very interested in learning about anyone who has
experimented with actual scheduling algorithms for running such FSM
programs on SIMD machines.  My intuition is that for a "typical" FSM
you should be able to reduce the factor of N to something more like
log N.  But that's just an intuition, I'd really like to see some
simulations.


Why Writing an Interpreter Looks Attractive

Given that the performance loss for a SIMD machine over a MIMD machine
is (perhaps) on the order of the size of the FSM you are trying to
run, it would seem that we can beat the game by writing an
interpreter!  That is, we could write a CL1 program that instructs a
collection of processing elements how to behave like the data
structures in an interpreter for some other language, say some
parallel dialect of Scheme.

That single CL1 program would be compiled into a single FSM with
perhaps 25 states.  We suffer a factor of 25 performance loss if we
run it on a SIMD machine rather than a MIMD machine, but we save a
great deal of hardware complexity.  Furthermore we can run -any-
parallel Scheme program with exactly the same factor of overhead,
rather than writing our code in CL1 where we suffer a different, and
perhaps huge, performance loss for each different piece of code.

Given that we never plan to write another piece of CL1 code again, we
can afford to really bum the hell out of the code for the interpreter.
We can even do the kind of scheduler analysis suggested in the last
section until we have squeezed every last drop of performance out of
that particular FSM.  Perhaps we can get down to a factor of 20.

Five years later we are manufacturing our SIMD machine out of some
technology that is 50 times faster anyway and we forget about the
factor of 20 we threw away once.  Life is golden, flowers bloom, and
little Cherubs fly about our machine room.


The Fly in the Ointment

But it doesn't work out that way in reality.  Our Scheme interpreter
had terrible behavior for a parallel programming language
implementation.  Although we never ran it on any actual parallel
machine, it was only necessary to look at the structures in our
simulations to see that we had written a loser.  (But please, before
you write us off as a couple of bozos, remember that we didn't care
how badly the interpreter ran.  We were trying to learn something
about using CL1.  The Scheme interpreter was just a test program.)

We did exactly what we had set out to do: we had written an
-interpreter-.  We had written a CL1 program that manipulated as data,
objects that represented Scheme programs.  At runtime the data that
represented the text of the FACT procedure would be accessed whenever
anyone anywhere wanted to apply that procedure to anything.  Given
that there is a limited bandwidth to the memory holding the text of
the FACT procedure, this results in processing elements queueing up
for their turn.

This is a major problem since it means that no matter how many
processing elements a machine has, only a fixed number of them can be
using any procedure at any given time.  By not thinking about this
problem we managed to throw away the whole advantage of building a
parallel computer in the first place.

There were other problems with our interpreter, other places where
processing elements inefficiently queued up waiting for access to
something [see Phil's message in PARSYM V1 #37], but the fact that we
had processing elements waiting for the -text- of procedures I find
especially significant.  We are right back where we started, having
trouble distributing instructions to the people who wanted to execute
them!

We could have avoided this bottleneck in one of two ways: we could
have distributed copies of the text of the FACT procedure to everyone,
or we could have broadcasted the text of the FACT procedure
periodically somehow.  In the first case we would have reinvented the
MIMD machine, and would have had to pay the price in memory
proportional to the size of the programs we wanted to run.  In the
second case we would have had a SIMD machine, and would have suffered
a slowdown (perhaps) proportional to the size of our program.

Any other solutions anyone can devise will, I am sure, have analogs in
hardware architectures we could have built in the first place, and
analogous cost/benefit tradeoffs.

All of this would still be true, by the way, if we were using a MIMD
machine instead of a SIMD machine.  If we had decided to build a MIMD
machine, and save money by only giving each processing element enough
memory to store the parallel Scheme interpreter FSM, we would still be
facing exactly the same dilemma here.

The lesson I would draw from this goes something like:

If at some level of abstraction different parts of the machine are
doing different things at the same time, and if those differences
occur dynamically as a result of what data happens to be where (as
opposed to statically where we planned in advance that processing
element 69 would compute pi from 5 PM to 6 PM), and if those
differences are sufficiently diverse, then there will be a problem
distributing something we might as well call "instructions".  And you
won't be able to get rid of the problem by any slight of hand tricks.

------------------------------

End of PARSYM Digest
********************

∂22-Nov-85  1455	@MIT-MC.ARPA:MERRY.pa@XEROX.ARPA 	Re: Classic Quotes (Positivism Assesses Progress in Meaning) 
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 22 Nov 85  14:51:05 PST
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 22 NOV 85  17:45:11 EST
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 22 Nov 85 17:42-EST
Received: from Xerox.ARPA by MIT-MC.ARPA 22 Nov 85 16:54:34 EST
Received: from Semillon.ms by ArpaGateway.ms ; 22 NOV 85 11:34:25 PST
Date: Fri, 22 Nov 85 11:34:11 PST
From: MERRY.pa@Xerox.ARPA
Subject: Re: Classic Quotes (Positivism Assesses Progress in Meaning)
In-Reply-To: <[MIT-MC.ARPA].727439.851121.JCMA>
To: "John C. Mallery" <JCMA@MIT-MC.ARPA>
cc: MINSKY@MIT-MC.ARPA, PHIL-SCI@MIT-MC.ARPA
Message-ID: <851122-113425-2280@Xerox>

Dear Chris,

Thanks, but I'm not quite sure I get your meaning.

Diana . . . . I think

By the way my mother's maiden name is Bunge.  Hmmmm?


∂22-Nov-85  1902	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Call for Algorithms Papers
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 22 Nov 85  19:01:50 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 22 Nov 85 18:29:27-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Fri 22 Nov 85 18:29:25-PST
Received: by rsch.wisc.edu; Fri, 22 Nov 85 20:16:46 CST
Message-Id: <8511222123.AA05439@rsch.wisc.edu>
Received: from WASHINGTON.ARPA by rsch.wisc.edu; Fri, 22 Nov 85 15:23:26 CST
Date: Fri 22 Nov 85 13:21:07-PST
From: Larry Ruzzo <Ruzzo@WASHINGTON.ARPA>
Subject: Call for Algorithms Papers
To: theory@RSCH.WISC.EDU
Status: R
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 22 Nov 85 20:16:24 CST (Fri)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

From: Paul Purdom <pwp%indiana.csnet@CSNET-RELAY.ARPA>

The Fall Joint Computer Conference will meet in Dallas Texas on Nov. 2-6,
1986. The deadline for papers is March 15, 1986. I am organizing the session
on algorithms. There also will be many other sessions on topics of interest to
people in theoretical computer science. An effort is being made to have a
conference of higher technical quality than that of recent general
conferences. If you have some important results and are willing to present
them in a form that is accessable to a wide audience, you are urged to submit
them to this conference. Five copies of your paper should be submitted. You
may send papers on algorithms (other than algorithms for artifical
intelligence) directly to me if you wish. Papers on any topic may be sent to
Harold Stone, Program Chair, FJCC86, IBM T J Waston Research Center, P.O. Box
218, Yorktown Heights, N.Y. 10598.

Paul Purdom, 101 Lindley Hall, Computer Science, Indiana University,
Bloomington, Ind., 47405.

--------------
TN Message #11
--------------

∂22-Nov-85  1941	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 22 Nov 85  19:30:27 PST
Received: from ames-vmsb.ARPA ([192.12.123.6].#Internet) by SU-SCORE.ARPA with TCP; Fri 22 Nov 85 19:28:22-PST
Date: 22 Nov 85 19:10:00 PST
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA


               ASSOCIATION FOR COMPUTING MACHINERY
                San Francisco Golden Gate Chapter
               "SIGBIG" Special Interest Committee
                 For Large High Speed Computers

 Meetings on  the first Wednesday of each month at 7:30 PM.   Speakers 
 who  can give insights to various aspects of  SUPERCOMPUTING are 
 featured each month.

 Next meeting:     Wednesday, December 4,1985,  7:30 PM

     Speaker:      Denny Cole/Intel   Beaverton,ORE (Hypercube)

     Topic:        The iPSC: An Environment for Concurrent Processing

     Location:     AXIOM Systems
		   1589 Centre Pointe Drive
		   Milpitas, CA 

     Directions: 17 South to Montague Expressway East. Left from
		 Montague onto Centre Point (before Capitol).
	or	 17 South to Capitol Expressway East. Right from
		 Capitol onto Centre Point (before Montague).
	or	 680 South to Montague Expressway West. Right from
		 Montague onto Centre Point (after Capitol).

 ---------------------------------------------------------------
 Tape-recordings  of  most of the previous  may  be obtained
 in exchange for a tape cassette or $5.00 by contacting: 
                Mary Fowler (415)261-4058 (rec)
                Supercomputing  #192, BOX 2787
                Alameda, CA. 94501-0787

 For information contact Mary Fowler, Chairperson (415) 839-6547
                     or  Mike Austin, Publ. Chair (415) 423-8446

------

∂23-Nov-85  0351	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 26 (E. Clark, Stanford)    
Received: from UCBVAX.BERKELEY.EDU by SU-AI.ARPA with TCP; 23 Nov 85  03:49:30 PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA00636; Thu, 21 Nov 85 12:17:32 PST
Received: by cogsci (5.31/5.16)
	id AA09584; Thu, 21 Nov 85 12:14:19 PST
Date: Thu, 21 Nov 85 12:14:19 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8511212014.AA09584@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 26 (E. Clark, Stanford)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                     Cognitive Science Seminar - IDS 237A

                      Tuesday, November 26, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

                  ``Contrast as a Constraint in Acquisition''
                                 Eve V. Clark
                Department of Linguistics, Stanford University
		            [eclark@su-psych.arpa]

        Speakers of a language tacitly subscribe to what  I  will  call
        the  Principle  of  Contrast,  namely that a difference in form
        marks a difference in  meaning.   This  principle,  I  propose,
        offers  a  powerful  tool  to  children acquiring language.  It
        serves to constrain the inferences  they  make  about  possible
        meanings  for  new  forms in the lexicon, in morphology, and in
        syntax, by distinguishing  them  from  the  meanings  of  forms
        already  familiar.  If the Principle of Contrast is observed by
        children, three major predictions follow:  (i)  differences  in
        form  should  be  taken  to signal differences in meaning, (ii)
        established forms should take priority  over  innovative  ones,
        and  (iii) gaps in the lexicon should be filled on the one hand
        by unfamiliar words and on the other by lexical innovations.

             In this talk, I examine these predictions  and  show  that
        each  is  strongly  supported  by  acquisition  data.  Children
        appear to observe the Principle of Contrast from very early.  I
        will  also argue that this principle offers a means for getting
        rid of unconventional, over-regularized forms in  the  lexicon,
        in  morphology,  and  in syntax.  The assumption that different
        forms have different meanings is as indispensable  in  acquisi-
        tion as it is in everyday use.
        ---------------------------------------------------------------
        UPCOMING TALKS
        December   3:  Bernard Baars, Langley Porter, UCSF
        ---------------------------------------------------------------
 

∂23-Nov-85  0352	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Nov. 26 (E. Clark, Stanford)
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 23 Nov 85  03:52:18 PST
Received: from ucbvax.berkeley.edu by SU-CSLI.ARPA with TCP; Sat 23 Nov 85 03:42:36-PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA00636; Thu, 21 Nov 85 12:17:32 PST
Received: by cogsci (5.31/5.16)
	id AA09584; Thu, 21 Nov 85 12:14:19 PST
Date: Thu, 21 Nov 85 12:14:19 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8511212014.AA09584@cogsci>
To: cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Nov. 26 (E. Clark, Stanford)

                      BERKELEY COGNITIVE SCIENCE PROGRAM
                     Cognitive Science Seminar - IDS 237A

                      Tuesday, November 26, 11:00 - 12:30
                        240 Bechtel Engineering Center
                 Discussion: 12:30 - 1:30 in 200 Building T-4

                  ``Contrast as a Constraint in Acquisition''
                                 Eve V. Clark
                Department of Linguistics, Stanford University
		            [eclark@su-psych.arpa]

        Speakers of a language tacitly subscribe to what  I  will  call
        the  Principle  of  Contrast,  namely that a difference in form
        marks a difference in  meaning.   This  principle,  I  propose,
        offers  a  powerful  tool  to  children acquiring language.  It
        serves to constrain the inferences  they  make  about  possible
        meanings  for  new  forms in the lexicon, in morphology, and in
        syntax, by distinguishing  them  from  the  meanings  of  forms
        already  familiar.  If the Principle of Contrast is observed by
        children, three major predictions follow:  (i)  differences  in
        form  should  be  taken  to signal differences in meaning, (ii)
        established forms should take priority  over  innovative  ones,
        and  (iii) gaps in the lexicon should be filled on the one hand
        by unfamiliar words and on the other by lexical innovations.

             In this talk, I examine these predictions  and  show  that
        each  is  strongly  supported  by  acquisition  data.  Children
        appear to observe the Principle of Contrast from very early.  I
        will  also argue that this principle offers a means for getting
        rid of unconventional, over-regularized forms in  the  lexicon,
        in  morphology,  and  in syntax.  The assumption that different
        forms have different meanings is as indispensable  in  acquisi-
        tion as it is in everyday use.
        ---------------------------------------------------------------
        UPCOMING TALKS
        December   3:  Bernard Baars, Langley Porter, UCSF
        ---------------------------------------------------------------
 

∂23-Nov-85  1159	ACUFF@SUMEX-AIM.ARPA 	3674 and 3647 status   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 23 Nov 85  11:59:46 PST
Date: Sat 23 Nov 85 11:58:53-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: 3674 and 3647 status
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12161602697.55.ACUFF@SUMEX-AIM.ARPA>

   The monitor on 3647 has been replaced and so 3647 should be fully operational
now.

   An IO board in 3674 has been replaced and seems to have fixed the network
failure problem.  Please use 3674, especially it's network interface (ie. send
and retreive files on it from other machines), and let me know right away if
you have problems.

	-- Rich
-------

∂23-Nov-85  1233	ACUFF@SUMEX-AIM.ARPA 	Air Conditioning in the Welch Road Machine Room 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 23 Nov 85  12:31:53 PST
Date: Sat 23 Nov 85 12:31:00-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Air Conditioning in the Welch Road Machine Room
To: ksl-LispM@SUMEX-AIM.ARPA
Message-ID: <12161608544.55.ACUFF@SUMEX-AIM.ARPA>

Folks,
   We are very short of A/C capacity in the WR machine room, so in the
short term, while solutions are being studied, I'm going to try to keep
some of the lisp machines powered down.  Until further notice, I'd like
to turn off unused machines at night and over weekends.  In particular,
I'm planning on cutting off all but one of the Explorers in the lobby
of bldg. C, Explorers in offices that are not used during off hours
(Delagi, Brown, Nii, and Delaney), and 3547 (which has no file system).
To avoid accidents, I'd like to take care of the on/off switching myself,
though if you need to get files off of a powered down Explorer, go
ahead and turn it ON (Don't turn anything OFF unless there is an emergency!)
by pressing the button on the FRONT of system unit in.  All of the Explorers
have tags on the front identifying them by host name.

   Also, if you notice that the machine room seems unusually hot,
please notify me, Nick Veizades, or some other staff person.  Don't be
afraid to call me at home.

   We realise that this is a gross inconvenience, and are doing all we
can to remedy the problem, so please bear with us.

	-- Rich
-------

∂23-Nov-85  1255	JOHNSON@SU-CSLI.ARPA 	Intro to Computational Linguistics    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 23 Nov 85  12:55:02 PST
Date: Sat 23 Nov 85 12:47:38-PST
From: Mark Johnson <JOHNSON@SU-CSLI.ARPA>
Subject: Intro to Computational Linguistics
To: linguists@SU-CSLI.ARPA, folks@SU-CSLI.ARPA


Several people have commented that they feel a need
for an introductory course in computational linguistics
and computation here at Stanford.

Such a course might include

	- introduction to LISP
	- implementing FSAs, and some of their applications
	- basic CF parsing techniques (bottom up, earley)

with a practical rather than theoretical emphasis.

I am in principle interested in teaching such a course, although
I may have time constraints, due to thesis, etc.  But Spring this
year is a distinct possibility...

So if anybody is interested in having such a course here at Stanford,
let me or Joan Bresnan know.

Thanks,
Mark
-------

∂24-Nov-85  1238	LANSKY@SRI-AI.ARPA 	tomorrow's planlunch -- 11AM  
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 24 Nov 85  12:38:42 PST
Date: Sun 24 Nov 85 12:35:36-PST
From: LANSKY@SRI-AI.ARPA
Subject: tomorrow's planlunch -- 11AM
To: planlunch-reminder.dis: ;

		 LOGIC OF POINTERS AND EVALUATIONS: 
	       THE SOLUTION TO THE SEMANTIC PARADOXES
				
                            Haim Gaifman  (GAIFMAN@SRI-AI)
          Mathematics Department, Hebrew University, Jerusalem
 	    	     Visiting SRI International, AI Center

	 	    11:00 AM, MONDAY, November 25
       SRI International, Building E, Room EJ228 (new conference room)

-------

∂25-Nov-85  0924	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Nov 85  09:24:39 PST
Date: Mon 25 Nov 85 09:06:21-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
cc: library@SU-SCORE.ARPA, maslin@SUMEX-AIM.ARPA
Message-ID: <12162095577.16.RICHARDSON@SU-SCORE.ARPA>


Tomorrow's lunch discussion will be about plans for a proposal to establish
an international institute of computer science, with initial funding from
GMD. (MJH 146 at 12:15)
-------

∂25-Nov-85  1405	BERG@SU-SCORE.ARPA 	evaluation forms    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 25 Nov 85  14:05:39 PST
Date: Mon 25 Nov 85 14:01:12-PST
From: Kathy Berg <BERG@SU-SCORE.ARPA>
Subject: evaluation forms
To: instructors@SU-SCORE.ARPA, tas@SU-SCORE.ARPA
cc: berg@SU-SCORE.ARPA, modica@SU-SCORE.ARPA
Stanford-Phone: (415) 497-4776
Message-ID: <12162149252.8.BERG@SU-SCORE.ARPA>

We have just received the Fall quarter evaluation forms from
Tau Beta Pi.  Please stop by room 030 and get the forms and
instruction sheets from Gina Modica.  Please instruct the 
students to fill out the top of the form with the pertinent
information about the class, in addition to answering the
questions on the sheet.  Please have a volunteer bring
the forms back to either MJH 256 or MJH 030.

Thank you so much for your kind cooperation!

Kathryn Berg
-------

∂25-Nov-85  1600	EMMA@SU-CSLI.ARPA 	The Next TINLunch    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Nov 85  15:59:58 PST
Date: Mon 25 Nov 85 15:51:51-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: The Next TINLunch
To: friends@SU-CSLI.ARPA
Tel:  497-3479

Here is a description for the December 5 Tinlunch:

            ``A Humanistic Rationale for Technical Writing''
                         by Carolyn R. Miller
                   Discussion led by Geoff Nunberg

   This paper is typical of a number of recent articles by
sociologists, rhetoricians, and humanistically-trained writing
specialists, which insist that scientific writing is no less
rhetorical in its means and effects than is writing of an explicitly
belletristic sort. Whether or not we find their arguments compelling,
these articles raise interesting questions for producers and consumers
of technical prose, especially in intellectually self-conscious
disciplines like philosophy, AI, and linguistics. For example: What is
the common understanding of the research enterprise that underlies the
linguistic conventions characteristic of scientific prose, such as the
avoidance of ``I'' and the unusual uses of ``we,'' the frequent use of
impersonal constructions, the numbering of paragraphs, and so on? Can
we apply the apparatus of traditional rhetoric to the evaluation of
the expository usefulness of particular formal languages and
notational conventions? Is there grounds for distinguishing between a
``rhetoric of information,'' concerned with the selection and
arrangement of factual observations, and a ``rhetoric of
description,'' concerned with the linguistic means used to report such
observations?



The TINLunch will be held in the Ventura Conference Room at noon.
-------

∂25-Nov-85  1920	ZEC@SU-CSLI.ARPA 	linguistcs department colloquium
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 25 Nov 85  19:20:14 PST
Date: Mon 25 Nov 85 19:11:02-PST
From: Draga Zec <ZEC@SU-CSLI.ARPA>
Subject: linguistcs department colloquium
To: folks@SU-CSLI.ARPA
cc: zec@SU-CSLI.ARPA





           LINGUISTICS DEPARTMENT COLLOQUIUM



         Nick Clements will be giving a talk on 

          THE  PHONOLOGY  AND  MORPHOLOGY  OF

                REDUPLICATION  IN  BANTU

    on Tuesday, 26 November at 3:15, in History 217. 

   There will be a reception following the colloquium in 
Joseph Greenberg Room, at Linguistics Department, Bldg. 100

-------

∂26-Nov-85  0745	@SU-SUSHI.ARPA:karp@ernie.berkeley.edu  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 26 Nov 85  07:45:22 PST
Received: from ucbvax.berkeley.edu by SU-SUSHI.ARPA with TCP; Tue 26 Nov 85 07:40:32-PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA00650; Mon, 25 Nov 85 14:29:30 PST
Received: by ernie (5.31/5.16)
	id AA10846; Mon, 25 Nov 85 14:28:57 PST
Date: Mon, 25 Nov 85 14:28:57 PST
From: karp@ernie.berkeley.edu (Richard Karp)
Message-Id: <8511252228.AA10846@ernie>
To: aflb.local@su-sushi.arpa, allmsgs@ernie.berkeley.edu,
        csfac@ernie.berkeley.edu, csmsgs@ernie.berkeley.edu,
        theory-b@ernie.berkeley.edu, theory-f@ernie.berkeley.edu

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley, Calif.  (415)642-0143

Tuesday, November 26

SEMINAR IN COMPLEXITY
2:00  MSRI Lecture Hall
Michael Saks  "Dynamic Location Problems on Graphs"

COMPLEXITY COLLOQUIUM
4:00   MSRI Lecture Hall
Umesh Vazirani  "A Fast Parallel Matching Algorithm"

∂26-Nov-85  0953	YAMARONE@SU-CSLI.ARPA 	NO LUNCH WEDNESDAY    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 26 Nov 85  09:44:31 PST
Date: Tue 26 Nov 85 09:20:16-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: NO LUNCH WEDNESDAY
To: folks@SU-CSLI.ARPA


DUE TO PERSONNEL SHORTAGES AND THE IMPENDING ONSLAUGHT OF
THANKSGIVING CALORIES, THE VENTURA SANDWICH CORPORATION HAS
DECIDED TO SUSPEND OPERATIONS ON WEDNESDAY, NOVEMBER 27.

HAPPY THANKSGIVING!!

SEE YOU NEXT WEEK!!

V.S.C.
-------

∂26-Nov-85  1012	RICHARDSON@SU-SCORE.ARPA 	CSD Lunch
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 26 Nov 85  10:12:21 PST
Date: Tue 26 Nov 85 08:23:04-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
cc: library@SU-SCORE.ARPA, maslin@SUMEX-AIM.ARPA
Message-ID: <12162349843.19.RICHARDSON@SU-SCORE.ARPA>


Lunch today in MJH 146 at 12:15!
-------

∂26-Nov-85  1050	ullman@su-aimvax.arpa 	meeting tomorrow 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 26 Nov 85  10:50:55 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 26 Nov 85 10:44:33 pst
Date: Tue, 26 Nov 85 10:44:33 pst
From: Jeff Ullman <ullman@diablo>
Subject: meeting tomorrow
To: nail@diablo

Let's meet 11AM as usual.  Shukey has some ideas on algebraic
approaches to thinking about "all derivations" of a given logic
program, and if there is time I'd like to talk about extensions
to Allen's "polynomial fringe theorem."

∂26-Nov-85  1435	MODICA@SU-SCORE.ARPA 	Grades  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 26 Nov 85  14:33:18 PST
Date: Tue 26 Nov 85 14:20:06-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: Grades
To: instructors@SU-SCORE.ARPA
Message-ID: <12162414837.25.MODICA@SU-SCORE.ARPA>


This is to remind you of the procedures regarding End Quarter Reports.
All grades are due within 96 hours of the final exam.  All EQRs must
come to me (not Victoria any more) before they are mailed to the
Recorder's Office. If you want a copy of the grade sheet, you should
make a copy before you bring me the original. The Department Copy
remains in the Department Office.

Grades for students not on the EQR should be submitted to me on
Supplementary Report/Grade Change Cards (also known as yellow grade
change cards).

The policy on grades is as follows: End-Quarter grades are final and
not subject to change by reason of a revision of judgment on the
instructor's part; nor are passing grades to be revised on the basis
of a second trial (e.g. a new exam or additional work). Changes may be
made at any time to correct an actual error in computation or
transcribing, or where some portion of the student's work has been
unintentionally overlooked.

Feel free to contact me with questions. I'm modica @score, and e-mail
is the best way to reach me.
-------

∂26-Nov-85  2228	BRESNAN@SU-CSLI.ARPA 	interactions of morphology, syntax, and discourse    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 26 Nov 85  22:28:36 PST
Date: Tue 26 Nov 85 22:19:29-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: interactions of morphology, syntax, and discourse
To: folks@SU-CSLI.ARPA

←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
Next Thursday, December 5, following Thanksgiving, Draga Zec will
present her work on Obligatory Control in Clausal Complements in the
CSLI conference room at 10:00 a.m.  She derives the phenomenon of
obligatory anaphoric control from the theory of obviation together
with the semantics of control verbs.  Serbo-Croatian, from which much
of her evidence is drawn, beautifully illuminates the two separate
factors in obligatory control, which are obscured in English.
Everyone interested in the syntax and semantics of control
constructions and their implications for linguistic theory is invited.
Written copies of the paper will be available on Monday, December 2 at
the CSLI Receptionists' desk.

              INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                ``Obligatory Control in Clausal Complements''

	It is generally held that obligatory control correlates with
	the non-finiteness of the complement. Both syntactic and
	semantic theories of control have crucially depended on this
	particular assumption.  My intention is to present a case of
	obligatory control into clausal complements, develop an
	analysis within the LFG framework, and then explore the
	implications of this case for an adequate treatment of
	control.

	Serbo-Croatian has two types of clausal complements, Type 1
	which is generally uncontrolled, and Type 2 which allows
	obligatory control with predicates like 'try', 'intend',
	'persuade', 'force', etc. It will be shown that Type 2
	complements cannot be dealt with in terms of the LFG theory
	of functional control, or any other syntactic theory of
	control. Rather, it will be argued that these complements
	are a clear case of what in LFG is referred to as anaphoric
	control. Certain differences in anaphoric binding properties
	between the two complement types are attributed to the
	phenomenon of obviation which is found with Type 2 but not
	with Type 1 complements.

	Since anaphoric control cannot capture the systematic
	controller/controllee relation characteristic of obligatory
	control, one will have to make reference to the semantic or
	more precisely, thematic properties of control-inducing
	predicates. This may have implications for syntactic
	theories of obligatory control, whose aim is to make
	predictions about controller/controllee relations solely in
	syntactic terms. This case will also be relevant for the
	semantic analyses that account for control solely in terms
	of entailment.  --Draga Zec
	-------
-------

∂27-Nov-85  0015	EMMA@SU-CSLI.ARPA 	Ventura Security
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  00:15:28 PST
Date: Tue 26 Nov 85 15:09:33-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Ventura Security
To: folks@SU-CSLI.ARPA, consultants@SU-CSLI.ARPA
Tel:  497-3479


   Please be extra careful about locking up over the Thanksgiving break.

Remember to lock both your office door and the exit door of the building
when you leave.

Emma Pease
-------

∂27-Nov-85  0842	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #39  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  08:42:45 PST
Date: 27 Nov 85 0830-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #39
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 27 Nov 1985     Volume 1 : Issue 39

Today's Topics:

                                  SISAL
                    Connection Machine Programming
        Call for Papers -- Principles of Distributed Computing

----------------------------------------------------------------------

Date: Thursday, 21 November 1985  19:51-PST
From: eugene at AMES-NAS.ARPA (Eugene Miya)
Subject: SISAL

SISAL does not run on the Cray.  Yet.  God, we wish it did.  SISAL is
a joint project between DEC, LLNL, U of Manchester, CSU Fort Collins
and several sites willing to be mice: NASA Ames among them.  We can
run SISAL on several VAXen.  Last year, we had a dataflow workshop
with Jack Dennis.  SISAL is a cleaned up version of Jack's VAL
language.  We had physicist users willing to run SISAL or VAL, but
their 3-D problems quickly exceeded the capability to realistically
run problems on a VAX.  It would take days of VAX time to get some
type of Cray run comparison, say "programmability."  The Cray 2 with C
and Pascal (the latter vectorizing) introduces an interesting
prospect: SISAL on the 2.  [I thought about bringing up Tim Budd's APL
on the 2 BTW but w/o hardware vectorization (C does not vectorize yet,
and it would be an exercise for the APL user :-))].  The 2 running
UNIX brings a candy store of different packages never in this arena
(we have XLISP: 1st non-Cray package brought up).

If you want details of SISAL the contact is skedz@lll-crg (Steve
Skedzielewski is the local driving force behind dataflow at Livermore
now.

Comment on the SIMD vs MIMD discussion: not much code, but the
definitions of MIMD and SIMD seem interesting.  My boss was an ILLIAC
IV programmer.  While he agrees MIMD is "more powerful."  He views the
discussion as something of a joke.  One of the reasons SIMD was more
powerful on the ILLIAC was the capability for higher than Cray I/O
speeds (4096 head disks).  The numerical world appears to view MIMD at
two levels: simple IF-statement processing which seems to easily break
vector and parallelization, and true general purpose MIMD (few real
MIMD algorithms of high order [Monte Carlo, some pruning]).  Keep up
the discussion, its amusing to watch and let us know when you can
sustain a computation rate greater than large memory Crays [note: no
use of units].

:-)

--eugene miya
  NASA Ames Research Center

------------------------------

Date: Tuesday, 26 November 1985  05:49-PST
From: Seth Steinberg <sas at BBN-VAX.ARPA>
Subject: Golly!

When I read the Connections Machine book I came away fascinated but
very frustrated.  The description was almost completely at the
hardware level but aside from a bit of handwaving about xectors and
storage allocation nothing was said about the SOFTWARE!  When I read
Agre and Bawden's MIT AI Memo (Was it #786?) I was able to understand
how a connections machine could be programmed.  You can learn a lot
more from a few hundred lines of code than a few dozen papers full of
algorithms.

I would like to thank Phil Agre, who sent me a copy of this paper, and
Alan Bawden, who should publish his bit bucket since it contains some
of the most interesting code I have read for writing this gem.  I hope
this discussion encourages more writing on SOFTWARE ENGINEERING for
parallel computers.


						Seth Steinberg

------------------------------


Date: 18 Nov 85 14:23:51 PST
From: HALPERN@IBM-SJ.ARPA
To:   arpanet-bboards@mit-mc
Subject: Call for Papers -- Symposium on Principles of Distributed Computing

[Forwarded by Steven A. Swernofsky <SASW at MIT-MC.ARPA>]

                       CALL FOR PAPERS:
5th ACM SIGACT-SIGOPS Symposium on Principles of Distributed Computing
                            (PODC)

                  Calgary, Alberta, Canada
                     August 11-13, 1986


Original research contributions are sought that address fundamental
issues in the theory and practice of distributed and concurrent
systems.  Topics of interest include, but are not limited to, the
following aspects of concurrent and distributed systems:

* Principles of distributed computation derived from
  practical experience with working systems
* Algorithms and complexity
* Specification, semantics, and verification
* Programming languages and programming language constructs
* Fault tolerance

Important Dates:
Jan. 31, 1986: Abstracts due
Apr. 11, 1986: Authors informed of acceptance or rejection
May 16, 1986:  A final copy of each accepted paper due, typed on special
               forms for inclusion in the conference proceedings

Please send ELEVEN copies of a detailed abstract
(not the complete paper), with the address,
telephone number, and NET ADDRESS (if available) of a
contact author on the cover page, to the program chair:

     Dr. J. Halpern
     Department K53/801
     IBM Almaden Research Center
     650 Harry Road
     San Jose, CA 95120-6099

The abstract should be no more than 10 DOUBLE-SPACED TYPEWRITTEN PAGES.
It must include a clear description of the problem being discussed,
comparisons with extant work, and a section on major original
contributions.  There should be enough detail provided for the
program committee to make a decision.

SUBMISSIONS ARRIVING LATE OR DEPARTING SIGNIFICANTLY FROM THESE
GUIDELINES RISK REJECTION WITHOUT CONSIDERATION OF THEIR MERITS.

The Program Committee consists of:

  David Cheriton, Stanford
  Cynthia Dwork, IBM Almaden
  Nissim Francez, Technion
  Hector Garcia-Molina, Princeton
  Joseph Halpern, IBM Almaden
  Butler Lampson, DEC
  Richard Ladner, U. of Washington
  Paul Leach, Apollo Computer, Inc.
  Michael Merritt, AT&T Bell Laboratories
  Doron Rotem, Waterloo University/Lawrence Berkeley Labs
  Sam Toueg, Cornell

Conference Chair: Ernest Chang, Alberta Research Council
Publicity Chair: Tiko Kameda, Simon Fraser University

------------------------------

End of PARSYM Digest
********************

∂27-Nov-85  1056	ACUFF@SUMEX-AIM.ARPA 	Patches for Some Explorer Bugs   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  10:54:23 PST
Date: Wed 27 Nov 85 10:52:31-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Patches for Some Explorer Bugs
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12162639191.25.ACUFF@SUMEX-AIM.ARPA>

   I've made up patches for 7 of the bugs in our current software (2.12).
I advise everyone to do (make-system 'ksl-patches :noconfirm) in her
login-init.lisp.  As always, please let me know if there are problems.

   Also, a fairly complete list of known bugs can be found in
Sumex:<LispM>Exp-Pre2.bugs.

	-- Rich
-------

∂27-Nov-85  1411	ullman@su-aimvax.arpa 	algebra of derivations
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  14:11:29 PST
Received: by su-aimvax.arpa with Sendmail; Wed, 27 Nov 85 14:06:26 pst
Date: Wed, 27 Nov 85 14:06:26 pst
From: Jeff Ullman <ullman@diablo>
Subject: algebra of derivations
To: nail@diablo

Following up on what Shukey said this morning, perhaps the
most interesting algebraic object is the thing that describes
the pattern(s) of the variables in the strings of literals
that get generated.  For example, is there a tableau whose
pattern must match the variables in successive literals?


e.g., if the string looks like p(X1,X2) p(X2,X3) p(X3,X4)...
we could describe it by the tableau

		b1	b2
		b2	b3

Happy thanksgiving to all.

∂27-Nov-85  1446	SCHOLZ@SU-SUSHI.ARPA 	meeting of the minds   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  14:46:04 PST
Received: from SU-SUSHI.ARPA by su-aimvax.arpa with Sendmail; Wed, 27 Nov 85 14:41:19 pst
Date: Wed 27 Nov 85 14:41:17-PST
From: Karin Scholz <SCHOLZ@SU-SUSHI.ARPA>
Subject: meeting of the minds
To: mugs@SUMEX-AIM.ARPA
Cc: nail@SU-AIMVAX.ARPA
Message-Id: <12162680837.28.SCHOLZ@SU-SUSHI.ARPA>


allen vangelder has graciously agreed to speak and to answer questions
on the nail! project at the upcoming mugs (dec9, 4pm, mjh252).
y'all please be nice to him.
nail! meetings are conducted according to serial transmission protocol,
interrupts disabled, 1200 baud, 60db.  allen is probably unaccustomed
to screaming, clawing, kicking, and cries of the form "that's trivial!",
and "that is subsumed by X!" and "that's not what i want to know!" from
bulgy-veined audience members oblivious to concurrency control issues.

by the way, jeff mentioned that anyone who is interested in attending
the nail! meetings on wednesdays at 11am is welcome.
karin
-------

∂27-Nov-85  1648	LANSKY@SRI-AI.ARPA 	future plans for PLANLUNCH    
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  16:47:56 PST
Date: Wed 27 Nov 85 16:45:43-PST
From: LANSKY@SRI-AI.ARPA
Subject: future plans for PLANLUNCH
To: planlunch.dis: ;

Due to various circumstances, there will be NO planlunch this coming Monday.
The schedule for the rest of the quarter and for early next quarter follows:

Monday, Dec. 2:  BREAK
Monday, Dec. 9:  Lenny Wesley
Monday, Dec. 16: Marcel Schoppers
HOLIDAY BREAK
Monday January 20: Dave Smith
Monday January 27: Peter Ladkin

Don't forget:  if you would like to give a talk next quarter, 
	       please let me know!

-Amy Lansky  (LANSKY@SRI-AI)
-------

∂27-Nov-85  1708	BSCOTT@SU-SCORE.ARPA 	Tuesday Faculty Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Nov 85  17:06:32 PST
Date: Wed 27 Nov 85 17:05:29-PST
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: Tuesday Faculty Lunch
To: Faculty@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12162707088.12.BSCOTT@SU-SCORE.ARPA>


Since there have been so many concerns expressed about the new Patent Agree-
ment and Stanford's patent and copyright policy, I have arranged to have two
representatives from the Sponsored Projects Office present at Tuesday's
faculty lunch on December 3 to respond to your questions.  I am not sure just
who will be here to represent SPO, but I will let you know on Monday.

If you do have questions, I urge to you attend.

General discussion had been planned for next week, but Nils agrees that the
SPO visit is probably a good idea.

Betty
-------

∂28-Nov-85  1917	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #40  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 28 Nov 85  19:17:47 PST
Date: 27 Nov 85 2035-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #40
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Thursday, 28 Nov 1985      Volume 1 : Issue 40

Responses to PARSYM Hardware Survey:

        Aquarius (Berkeley)
        Tagged-Token Dataflow Architecture and
            Multiprocessor Emulation Facility (MIT)
        Butterfly (BBN)

[This digest begins the publication of the responses to the PARSYM
Survey on Hardware for Parallel Symbolic Computation (V1 #33).  Five
responses have been received so far.  Because of message size
limitations, the five have been divided into two digests: three
responses here, two in a subsequent digest.  Thank you very much,
respondents, for taking the time to let PARSYM know about your work.
I'm sure PARSYM readers will be impressed and enlightened by the
information you have provided.

I hope these early returns will stimulate other PARSYM readers to
reciprocate.  Further contributions are always welcome.  -- BD]

----------------------------------------------------------------------

Date: Saturday, 16 November 1985  09:13-PST
From: fagin at dali.berkeley.edu (Barry Steven Fagin)
Subject: Response to survey about hardware

Greetings fellow readers!

Here's my response to the digest's survey on hardware for parallel
symbolic computation:

> What computer hardware are you using or developing?

We're working on the Aquarius multiprocessor, a heterogeneous computing
system with multiple processing elements for symbolic and numeric
computation.  The heart of the system is a VLSI Prolog processor chip,
based on a TTL version designed and built here.

> What are its components (processors, memory) and how many; how do the
> processors and memory communicate?

We envision 16 processors connected with 16 main memory modules through
a crossbar.  Each processor has a cache between it and the
crossbar.  There is also a separate bus connecting the processors
to a synchronization memory, with snooping caches between the bus
and the processors.

> SIMD/MIMD?  Shared memory or distributed memory?

MIMD, shared memory.

> What model of concurrency are you investigating with this system?

A simplified version of Conery's AND/OR process model.




		  Part II (Optional, for extra credit)


> 0. Project summary:

> What model of concurrency (e.g., dataflow, communicating sequential
> processes, actors, futures, etc.) is your work based on?  What
> language, if any, are you using in your work?

We're using conventional Prolog.  Clause processes are responsible
for AND-parallelism, procedure processes for OR-parallelism.
Processes communicate with one another by sending messages to parents
and children.

> 1. Architectural classification:

> SIMD or MIMD?

MIMD.

> 2. Processing elements:

> Describe each processor ...

The processor is a VLSI version of a TTL Prolog processor designed and
built here at Berkeley.  Simulations indicate 400 KLips on
deterministic concatenate benchmark.  Each processor addresses a
maximum of 16 megabytes.  A functional and a register transfer level
simulator, both written in C, exist for the uniprocessor, and a
functional simulator exists for the multiprocessor.

Our system is a heterogeneous one, consisting of several parallel
Prolog processors (PPP's), processors for numeric computation,
and a processor for I/O and operating system functions.

> How many processors are there in the minimum and maximum configurations?
> How many are you currently using?

We envision 16 processors in the final system.  The simulator as it stands
now assumes an infinite number of processors are available.

> 3. Memory system:

> Describe how memory is distributed and accessed in the system ...

There are 16 main memory modules connected to the processors through
a crossbar, and a synchronization memory connected to the processors
over a bus.  There is a single address space.  Traffic to main
memory from each processor is intercepted by a write-through cache, to
the synchronization memory by a snooping cache.

> What are the minimum and maximum amounts of memory?  How much are you
> currently working with?

Unknown at this point.  The simulated version assumes infinite memory.

> 4. Communication system:

> How are processors connected to each other and to memory?  Describe in
> terms of topology (e.g., tree, systolic array, pipe, grid), switching
> type (e.g., LAN, bus, Butterfly switch), and throughput (e.g., bus
> transmission rates, time to access memory on remote nodes, time to
> broadcast a message to all nodes).

Can't answer these yet.

> 5. Synchronization:

> What hardware mechanisms does the system supply to support
> synchronization?

We haven't decided yet, but we'll probably use test and set.

> 6. Pragmatics

> Is your system geared towards either specific programming languages or
> applications?

We're looking for problems that can be solved rapidly by processing in
both the symbolic and the numeric domain.  Currently, we're using Prolog
as the language for symbolic processing.

> Is the hardware of your system generally available or, if custom-made,
> likely to be available in future?

We're working on agreements with companies to sell the chip we produce.
Will keep you posted.

--Barry Fagin

(fagin@berkeley)

------------------------------

Date: Wed, 20 Nov 85 10:50 EST
From: Soley@MIT-MC.ARPA
Subject: PARSYM Digest   V1 #33

		   The Third PARSYM Survey: Computer Hardware

There are essentially two major projects taking place in the Computation
Structures Group (formerly Functional Languages & Architectures Group)
at LCS (MIT) under Professor Arvind.  The end goal of one is a dataflow
machine known as the Tagged-Token Dataflow Architecture (TTDA), a
decidedly MIMD fine-grain-computation (though complex processors [at the
moment]) model.  The other project, known as the Multiprocessor
Emulation Facility (MEF), is a simulation/emulation testbed for
multiprocessor architectures.  The TTDA is being modelled under the MEF
environment, as well as under simulation on an IBM 4381.

				  Part I (Easy)

    What computer hardware are you using or developing?

The TTDA uses imaginary hardware (which we will construct when we
understand it).  The MEF uses any number of Lisp machines (we are
targeted at 8 Symbolics 3600's and 64 TI Explorers; we currently have 3
3600's, 5 3670's, and about 20 Explorers).  All are connected via
Ethernet (Chaos protocols), and the Explorers are also connected via a
globally clocked circuit-switched 3Mbyte/second/circuit network of our
own concoction, though it is based on the BB&N Butterfly's network.  We
are also working on a network interface card for the Explorer (called
NuCA, for Nubus Channel Adapter) and a packet-switching hypercube
topology network for the Explorers.

    SIMD/MIMD?  Shared memory or distributed memory?

Both the TTDA and the MEF are MIMD.  The TTDA uses a distributed memory
model known as "I-structures."  The MEF was originally intended
primarily to model packet-communicating (distributed memory)
architectures, but will likely do well with shared memory designs with
our new communication networks.

    What model of concurrency are you investigating with this system?

TTDA: Obviously, dataflow (fine-grained parallelism with no programmer
specification of that parallelism).  MEF: Anything goes.

    What would you like to see in your next hardware system?

Lower-latency interconnection network.


		      Part II (Optional, for extra credit)


    0. Project summary:

    What model of concurrency (e.g., dataflow, communicating sequential
    processes, actors, futures, etc.) is your work based on?  What
    language, if any, are you using in your work?

TTDA: Dataflow.  MEF: Anything goes.  The TTDA project is using the
language ID to express programs, though the emulation of the processor
is done in Lisp.  There are also simulations of the machine in Lisp and
Pascal.  The MEF is written entirely in Lisp, and emulations for the MEF
are also written in Lisp.

    1. Architectural classification:

    SIMD or MIMD?

Both MIMD.


    2. Processing elements:

    Describe each processor ...

TTDA: approximately 68000 power.  MEF: each host is a Lisp machine.

    How many processors are there in the minimum and maximum configurations?
    How many are you currently using?

TTDA: Min one processor, max probably several thousand.  Currently
simulating and emulating between one and sixteen.  MEF: Min one
processor, max 256.  Currently using between one and 28.

    3. Memory system:

    Describe how memory is distributed and accessed in the system. ...

TTDA: The only non-local memory access is through a mechanism called
I-structures.  The dataflow model is highly resistant to memory latency,
for reasons that I won't go into here (feel free to ask for references).
MEF: though originally intended for message-passing (no global memory),
we will be experimenting with remote memory reference over our faster
connection media.  Many different styles are being considered.

    What are the minimum and maximum amounts of memory?  How much are you
    currently working with?

MEF: Each of our Lisp machines has between four and six megabytes.

    4. Communication system:

    How are processors connected to each other and to memory?  Describe in
    terms of topology (e.g., tree, systolic array, pipe, grid), switching
    type (e.g., LAN, bus, Butterfly switch), and throughput (e.g., bus
    transmission rates, time to access memory on remote nodes, time to
    broadcast a message to all nodes).

MEF: Ethernet, circuit switch based on Butterfly switch, future packet
switch of our own design (in conjunction with on-site IBM engineers).

    5. Synchronization:

    What hardware mechanisms does the system supply to support
    synchronization?

TTDA: Tags on the tokens, and a unit known as "waiting-matching store."

    6. Pragmatics

    Is your system geared towards either specific programming languages or
    applications?

The TTDA is geared toward functional languages, such as ID (Irvine
Dataflow).  The MEF is geared toward Lisp.

    Is the hardware of your system generally available or, if custom-made,
    likely to be available in future?

TTDA: Not generally available.  MEF: Generally available.

    If you're now using a multiprocessor system, how does the architecture
    of your development system correspond to your research target
    architecture?

The TTDA and the MEF (on which the TTDA is emulated) are both comprised
of relatively complex processors, with high-bandwidth interconnections.

------------------------------

Date: Sat, 23 Nov 85 02:06 EST
From: "Curtis A. Scott" <cscott@BBN-VAX.ARPA>
Subject: Survey #3 reply

Disclaimer:

Since I am not only a user of the hardware described, but an employee
of the company which builds it (BBN), my views may be assumed to be
biased.  However, my views are just that-- my views-- and are not to
be taken as BBN statements or policy.



	       The Third PARSYM Survey: Computer Hardware

                (Reply to PARSYM-Survey@SUMEX-AIM.ARPA)

			      Part I (Easy)

What computer hardware are you using or developing?

A:  We are using BBN's Butterfly.


What are its components (processors, memory) and how many; how do the
processors and memory communicate?

A:  1 to 256 68000 processors, ultra-high-speed packet-switching interconnect
    network.


SIMD/MIMD?  Shared memory or distributed memory?

A:  MIMD.  Memory mapping is at the discretion of the systems programmer.
    Most applications run with code and a value cache in local memory
    and a global shared linear address space.  While it is possible
    to fetch and execute code across the switch, it's not a winning
    strategy.


What model of concurrency are you investigating with this system?

A:    FUTURE-based.


What would you like to see in your next hardware system?

A:  Faster processors, and/or more processors.  Larger address space
    (the processor is limited to 24 bits of virtual byte address).




		  Part II (Optional, for extra credit)


0. Project summary:

What model of concurrency (e.g., dataflow, communicating sequential
processes, actors, futures, etc.) is your work based on?  What
language, if any, are you using in your work?

A:  We are building a FUTURE-based version of Scheme (which we call
    "MultiScheme", since it is derived from both Scheme and Bert
    Halstead's Multilisp work).  This is intended to serve as a
    base for a concurrent Common Lisp and experimentation with
    other concurrent programming systems.  Scheme is implemented
    in a combination of 'C' and assembler; the Scheme compiler emits
    68000 code for simple operations.


1. Architectural classification:

SIMD or MIMD?

    MIMD.  We doubt that any SIMD machine can efficiently use its hardware
    performing general-purpose symbolic computation, although as
    subengines SIMD machines will be quite useful.


2. Processing elements:

Describe each processor ...

A:  The BBN Butterfly is built of processor nodes plus an ultra-high-speed
    interconnect network.  Each node has a 68000 (or 68020) processor, a
    node controller, and some amount (typically 1-4 Meg) of local memory.
    The processor node controller is a microprogrammed bitslice processor.

    The network nodes are single custom VLSI chips.  The interconnect
    network for 16 processor nodes fits on a single card.


How many processors are there in the minimum and maximum configurations?
How many are you currently using?

A:  Butterflies can be run in any configuration from 1 to 256 processors.
    We use a 16-processor machine for most of our development, and use
    the 128-processor machine at BBN occasionally.


3. Memory system:

Describe how memory is distributed and accessed in the system. ...

A:  The node controller mediates all memory references made by the processor,
    performing the mapping of local and distant memory into the address
    space of processes running on that processor and handling requests
    from other processors.  If a reference is non-local, the node
    controller stops the processor, sends a memory request through the
    network to the processor on which the physical memory for that virtual
    address actually resides, and restarts the processor when the value
    returns.  This is transparent to the processor and typically slows
    a 16-bit nonlocal read from the normal local 1 usec to 4-5 usec.

    There is no hardware cache.  Snoopy caches are not possible on
    non-bus-based systems, but bus-based systems will never have
    sufficient bandwidth for multiprocessing on the scale we
    envision.


What are the minimum and maximum amounts of memory?  How much are you
currently working with?

A:  The minimum is a single node with 256K bytes.  This size is no longer
    being built.

    The maximum is currently 256 processors X 4 Meg = 1 Gigabyte of
    memory.  Such a machine has not yet been built; the maximum
    configuration to date has 128 processors X 1 Meg = 128 Megabytes.


4. Communication system:

How are processors connected to each other and to memory? ...

A:  Butterfly switch.  Time to read memory varies depending on
    network load, but usually is ~5 usec.  Writes take
    about half that time since a return packet is unnecessary.
    If "broadcasting a message to all nodes" is viewed as the
    time it takes to notify a process on each node that some event
    has occurred, this time is minimized (on the current machines) by
    simply writing to a flag residing in the local memory of each processor
    node.  This takes approximately 8 * NumProcessors usecs.  There are
    also higher-level event notification mechanisms available, but in the
    simplest case a write is cheaper.  Total best-case throughput of the
    network on a 128-processor machine is around 32 Gb, and scales with
    the number of processors.  Switch congestion is rarely a problem,
    but memory hotspotting (when some piece of physical memory
    accessed through a single node controller is used heavily by
    other processors) can cause performance degradation for naively-
    parallelized programs.


5. Synchronization:

What hardware mechanisms does the system supply to support
synchronization?

A:  Simple atomic operations (add, subtract, ior, xor) are implemented
    in microcode, and are somewhat slower than simple reads and writes,
    typically <30 usec.  The simplest form of semaphore uses atomic←ior
    to get a flag and simple write to clear it, and thus takes <35 usecs
    of overhead.

    Microcode-supported high-level constructs include Events (an object
    on which a process may wait, and which may be atomically posted
    with a single datum), and DualQueues, which have atomic EnQueue and
    DeQueue operations and allow processes to leave Events to be posted
    if there is no data available in the queue.


6. Pragmatics

Is your system geared towards either specific programming languages or
applications?

A:  The hardware was originally built for use as a voice packet switching
    node, but is now completely general-purpose.  Systems programming
    and applications are now programmed mostly in 'C', but our project
    aims to make the machine more programmable for AI and 'symbolic'
    (yes, I know, all computing is symbolic) applications.


Is the hardware of your system generally available or, if custom-made,
likely to be available in future?

A:  The hardware is available to anyone who wants to plunk down money.
    Most machines are currently installed in university and defense
    contractor labs.


If you're now using a multiprocessor system, how does the architecture
of your development system correspond to your research target
architecture?

A:  Exactly, since our 'application' is a good symbolic programming
    environment for this particular machine.


What are the advantages of your hardware setup, particularly for the
problems you are investigating?  What are the disadvantages?  Is there
a system that would better meet your needs?

A:  Advantages: Direct access to the hardware developers and many of
    the Butterfly's current users; reliable hardware; rapidly improving
    system amenities (e.g., Internet access to Butterflies, file
    servers).

    Current disadvantages: no disk drives directly attached to the
    machine; compilation currently is done on a VAX or Sun and files
    downloaded via Ethernet to the Butterfly.

    There is no stable, available multiprocessor that I am aware of which
    has the capacity of the Butterfly in useful interconnect bandwidth
    and processing power.  Since our project is intended to result in a
    system for the fastest possible general symbolic computing, this machine
    is the best engine for it.

------------------------------

End of PARSYM Digest
********************

∂29-Nov-85  0848	NILSSON@SU-SCORE.ARPA 	Computing Needs  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Nov 85  08:48:45 PST
Date: Fri 29 Nov 85 08:47:54-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Computing Needs
To: faculty@SU-SCORE.ARPA
Message-ID: <12163140794.8.NILSSON@SU-SCORE.ARPA>

On November 1, Les Earnest sent a message to the faculty list requesting
5 year projections of computer facility needs, to be used in several
planning efforts.  To date, the following responses have been received

 Activity	Respondents	 Remarks

Comp. Linguist. T. Winograd
& Specification

Analysis of	D. Knuth
Algorithms

 -		R. Floyd	Minimal support needed

A.I. & Formal	L. Earnest
Reasoning	& J. McCarthy

Knowledge	T. Rindfleisch  Instructional computing sketchy
Systems Lab.	& P. Rosenbloom

Of course, we are not going to assume that a "no response" means
that you will have no computing needs during the next 5 years--we'll
just have to try to guess what they might be.  I think we'll be 
able to guess the "average needs" as well as anyone, but if you
know about special needs, now is the time to get them into Les.

The reason for being concerned about this now, is that budget
forecasts are being submitted, temporary (1985-1990) space needs
are being planned, and certain decisions regarding our computing
environment and how it ought to be organized are being made.  We take
our job to be one of creating the flexibility we all want during the
next few years--not one of thinking up constraints on what we do.  BUT,
we do need to hear from you to do that job well.

If your projected needs have not been reported yet, please send them to
Les @SAIL not later than December 2.  If we don't know what you need,
we can't make adequate plans for space, computers, and staffing.

Thanks,  -Nils
-------

∂29-Nov-85  1028	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books--CS   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Nov 85  10:28:44 PST
Date: Fri 29 Nov 85 10:27:36-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books--CS
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA, cs%Playfair@SU-SCORE.ARPA
Message-ID: <12163158945.9.LIBRARY@SU-SCORE.ARPA>

Parallele Algorithmen. Informatik-Fachberichte No. 64. by F. Hossfeld.
QA184.H83 1983.

Proceedings Of The 1981 KL-One Workshop.  Fairchild Technical Report
No. 618. FLAIR Technical Report No. 4. May 1982. by James G. Schmolze
and Ronald J. Brachman. (8512116).

1985 ACM Thirteenth Annual Computer Science Conference March 12-14, 1985.
(8512393)

Proceedings Graphics Interface '83. Comptes Rendus Interface Graphique '83.
May 1983. Edmonton, Alberta  T385.P78 1983A f.

Brinch Hansen On Pascal Compilers. by Per Branch Hansen. QA76.73.P2B75 1985.

Invitation To Database Processing. Petrocelli Books. by Leonard Gorney.
QA76.9D3G67 1985.

H. Llull
-------

∂29-Nov-85  1527	NILSSON@SU-SCORE.ARPA 	Patent agreement 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Nov 85  15:27:22 PST
Date: Fri 29 Nov 85 15:26:43-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Patent agreement
To: faculty@SU-SCORE.ARPA
Message-ID: <12163213398.8.NILSSON@SU-SCORE.ARPA>

I still have some questions about the new patent agreement that the
university is asking us to sign.  Maybe these will be cleared up
at Tuesday's faculty lunch (maybe they won't too).  Anyway, I'm
personally not signing anything yet!  -Nils
-------

∂29-Nov-85  1537	LIBRARY@SU-SCORE.ARPA 	Paper Index To Technical Reports No Longer Available
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Nov 85  15:37:09 PST
Date: Fri 29 Nov 85 15:36:21-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Paper Index To Technical Reports No Longer Available
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12163215151.9.LIBRARY@SU-SCORE.ARPA>

We have removed the Index TO Uncataloged Materials from the reference
shelf in the Math/CS Library.  This paper index was only accurate
as of January 1985.  The only complete and up-to-date index to our
technical reports is the Technical Reports file in Socrates.

H. Llull
-------

∂01-Dec-85  1235	JOHNSON@SU-CSLI.ARPA 	My books..   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 1 Dec 85  12:34:56 PST
Date: Sun 1 Dec 85 12:23:54-PST
From: Mark Johnson <JOHNSON@SU-CSLI.ARPA>
Subject: My books..
To: folks@SU-CSLI.ARPA, linguists@SU-CSLI.ARPA



I noticed apon return that my bookcase is looking rather
bare these days..


Would anyone that has a book or journal from me please
either return it or send me mail telling me that they have
it?

Thanks,

Mark
-------

∂01-Dec-85  1639	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #41  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Dec 85  16:38:56 PST
Date: 1 Dec 85 1630-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #41
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 2 Dec 1985        Volume 1 : Issue 41

Responses to PARSYM Hardware Survey:

        COBWEB (King's College London)
        GRIP (University College London)

----------------------------------------------------------------------

Date: Wednesday, 20 November 1985  09:32-PST
From: Paul Kelly <mcvax!kcl-cs!pkelly at seismo.CSS.GOV>
Subject: PARSYM Digest   V1 #33

	       The Third PARSYM Survey: Computer Hardware

FROM: Paul Kelly,
      Dept of Computing,
      King's College London, (was Westfield College)
      UK. (mail address at the end)

[Please note that there are several individuals and several
organisations involved in the project I am describing.  This does not
purport to be description of the project as a whole - only my
perception of it.]

			      Part I (Easy)

What computer hardware are you using or developing?
   We are using High Level Hardware's Orion supermicro, running Unix 4.1,
   and have been using this machine's excellent user-microcode facility
   although I'm not involved with that.

   More interesting are our computer architecture designs.  I'm
   only going to talk about one - the COBWEB.  An early version
   is described in the Proceeding of the Nancy (1985) Conference
   on Functional Programming and Computer Architecture.
     COBWEB is an as-yet hypothetical machine directed at
   wafer-scale integration.  It is specifically directed at
   functional languages.  The early versions, at least, run a
   machine code based on Turner's combinators augmented with
   directors (see Kennaway & Sleep's paper "Director Strings as
   Combinators").

What are its components (processors, memory) and how many; how do the
processors and memory communicate?
   We intend there to be thousands of processing elements in a
   COBWEB, although each processing element (PE) is very small.
   PEs' local memory is the only storage in the machine, and
   each PE's memory is quite small - perhaps only big enough for
   2 - 4 reduction tokens.

   PEs are arranged in a grid - possibly 6-connected, or
   8-connected.  We don't yet have the detailed figures to make a
   choice at the moment.  They communicate by passing tokens
   between immediate neighbours.  The early prototypes use Ivor
   Catt's spiral both for fault-tolerance, and for token routing.
   (the spiral starts in the middle, and grows as the end PE
   passes control to its left-most working neighbour.  A spiral is
   created which buries faulty PEs between layers.  Communication
   occurs both around the spiral, and radially.)

SIMD/MIMD?  Shared memory or distributed memory?
   MIMD - this is a highly-parallel very small granularity
   machine.  Memory is distributed.

What model of concurrency are you investigating with this system?
   The first design views concurrency as a parallel graph
   reduction process, guaranteed to be semantically equivalent
   to the standard 'normal' reduction order.
      The machine interprets functional programs to give a
   fully-lazy semantics.  In order to extract parallelism,
   strictness analysis is used during compilation to find
   expressions reduceable earlier than the standard reduction
   order, whilst still guaranteeing that expressions not needed
   by the computation are not computed, so termination
   correctness is preserved.
      With directors, the graph reduction process has many
   similarities with data-flow, as might be expected since
   functional languages and data-flow languages are in essence
   indistinguishable.

What would you like to see in your next hardware system?
   No COBWEB design has yet been implemented, but we are already
   looking beyond the machine described in the Nancy paper.
   We are looking at different routing and fault tolerance
   techniques within the grid structure, and also at more exotic
   logical interconnection topologies.  We are trying to stick to
   fairly conservative fabrication technology - wafer-stepping
   rather than direct writing of whole wafers, and self-test and
   self-reconfiguration.  This should keep the price down, which
   I think is important.  I think COBWEBs should be components of
   affordable home computers, or at least cheap desk-top
   workstations.
      We're also looking at different evaluation mechanisms,
   though we're sticking with functional languages.  These
   include super-combinators, or serial combinators, and
   data-flow.

This seems to answer your other questions.

I must emphasize that this is a personal view.

I am Paul Kelly (Postgraduate),
     King's College Dept of Computing,
     Westfield College,
     Kidderpore Avenue,
     Hampstead,
     London NW3.     ('phone (01) 435 7141 ext 636)
     UK.

------------------------------

Date: Thu, 21 Nov 85 10:59:47 GMT
From: Simon PJ <simonpj@ucl-cs.arpa>
Subject: GRIP

        Third PARSYM survey - Computer Hardware

GRIP (Graph Reduction In Parallel) is the name of a parallel machine
designed to execute functional languages using graph reduction, under
development at University College London.

I've answered PART 1, and appended a brief writeup which gives some
more details.

Simon L Peyton Jones
Dept of Computer Science
University College London
Gower St
London WC1E 6BT

simonpj@ucl.cs          (try via ISID if you have a problem)

--------------------------

        Brief answers to survey

GRIP is a parallel machine designed to execute functional languages
using graph reduction.  (Actually the hardware is amenable to other
languages - one group wants to put Prolog on it).  It is being
developed with the support of the Alvey Directorate, in collaboration
with 3 companies.

We are developing our own hardware, consisting of microprocessor based
Processing Elements (PEs) and bit-slice Intelligent Memory Units
(IMUs).

They communicate over a bus, the IEEE P896 Futurebus.  This is the
most fundamental (and maybe surprising) design decision.  It allows us
to achieve substantial concurrency without using clever switches or
solving locality issues.  So we can address one issue at a time -
parallel graph reduction in particular.  Of course it places a limit
on the size of the machine, which we simply accept!  Up to that limit
we should achieve better price/performance than more extensible
machines.

We hope that each board will contain 4 PEs and 1 IMU, and the GRIP
machine will contain a maximum of 32 boards.

GRIP is definitely MIMD!  Each PE has some private memory, but the
IMUs implement a global shared structured memory (a heap, in fact).

The model of concurrency is GRAPH REDUCTION (does everyone know what
this is?).

Next system:
        - migrate functionality from the PEs into the IMUs.
        - rip out 4 PEs and replace with one bitslice PE
          to give same or better performance with less
          concurrency required.

-------------------



		    A brief overview of GRIP

	       A parallel graph reduction machine



	Simon L Peyton Jones, Chris Clack and Jon Salkild

		    University College London
		       21st November 1985


GRIP is a parallel Fifth Generation machine designed to  execute
functional  languages  using  graph reduction.  In this paper we
give a brief outline of graph reduction and GRIP's architecture.

1.  Functional languages and graph reduction

Functional languages are Fifth Generation  programming  languages
which  address  two  of  the major current challenges to computer
science, namely the challenges of correctness and parallelism.

The challenge of correctness is the difficulty we  experience  in
writing  large correct programs, a problem which Hoare eloquently
outlines in his Turing award lecture [Hoare81].  Darlington gives
an  introduction  to  functional  programming showing how some of
these problems may be alleviated [Darl84].

The organisation of a number of  independent  processors  to  co-
operate  in  executing  a  single  program  is  the  challenge of
parallelism.  Functional languages contain inherent  parallelism,
and  so  are  a  suitable  medium  in  which  to express parallel
programs.  To understand where the parallelism comes from we will
look at a functional program:

	let f x = (x+1)*(x-1)
	in  f 4

The "let" defines a function "f" of a single argument "x",  which
computes  "(x+1)*(x-1)".   The  program executes by evaluating "f
4", that is the function "f" applied to 4.  We can think  of  the
!
			  GRIP overview



program like this:

		@
	       / \
	      f   4

where the "@" stands for function application.  Applying "f" to 4
gives

		*
	       / \
	      /   \
	     +     -
	    / \   / \
	   4  1  4   1

We  may  now   execute   the   addition   and   the   subtraction
simultaneously, giving

		*
	       / \
	      5   3

Finally we can exeucte the multiplication, to give the result

		15

From this simple example we can see that:

   (i) Executing a functional program consists of  evaluating  an
       expression.

  (ii) A functional program has a  natural  representation  as  a
       tree (or more generally, a graph).

 (iii) Evaluation proceeds by  means  of  a  sequence  of  simple
       steps, called reductions.  Each reduction performs a local
       transformation  of  the  graph  (hence  the   term   graph
       reduction ).

  (iv) Reductions may safely  take  place  simultaneously,  since
       they cannot intefere with each other.

   (v) Evaluation is complete when there are no further reducible
       expressions.

GRIP  (Graph  Reduction  In  Parallel)  is  designed  to  execute
functional  programs  by  performing  parallel  graph  reduction.
Despite the opportunities for parallelism offered  by  functional
languages,  only  the  ALICE project (at Imperial College) has so
far  attempted  a  parallel  implementation  [Darl81].   GRIP  is
intended to provide state-of-the-art performance at moderate cost
by extracting maximum performance from a fast  bus.   This  means
that within its performance range, GRIP should provide more power
!
			  GRIP overview



for unit cost than more extensible designs (such as ALICE).   Our
performance  target  for  a  fully  populated GRIP is one million
reductions per second.

2.  The GRIP architecture

Most proposals for parallel graph reduction  machines  look  like
Figure  1.  The Processing Elements (PEs) traverse the graph held
in the Intelligent Memory  Units  (IMUs),  discovering  reducible
expressions  and  reducing them.  The principal variation between
different  designs  lies   firstly   in   the   the   choice   of
communications  network,  and secondly in the intelligence of the
IMUs.



    --------------  --------------       --------------
    |            |  |            |       |            |
    | Processing |  | Processing |  ...  | Processing |
    | Element    |  | Element    |       | Element    |
    |            |  |            |       |            |
    --------------  --------------       --------------
	 |                |                     |
    ------------------------------------------------------
    |           Communications medium                    |
    ------------------------------------------------------
	  |                 |                      |
    ---------------  ---------------       ---------------
    |             |  |             |       |             |
    | Intelligent |  | Intelligent |       | Intelligent |
    |   Memory    |  |    Memory   |  ...  |   Memory    |
    |   Unit      |  |    Unit     |       |   Unit      |
    |             |  |             |       |             |
    ---------------  ---------------       ---------------

Figure 1.   Physical structure of a parallel graph reduction machine


2.1  The bus

In the case of GRIP we have chosen to use a fast  bus,  the  IEEE
P896  Futurebus, for the communications network.  A bus offers an
extremely cost effective switch, but at the cost  that  only  one
transaction  can take place between a PE and an IMU at once, thus
limiting concurrency.  This places a  fundamental  limit  on  the
parallelism  achievable  using  GRIP,  but  it gives an extremely
cost-effective solution up to this limit.  In the case of GRIP we
anticipate  being  able  to  integrate up to 120 PEs or so, on 30
boards, before running out of bus bandwidth and physical space.

The use of  a  bus  allows  us  to  address  one  research  issue
(parallel  reduction) at a time, rather than try to solve several
!
			  GRIP overview



difficult problems at once.  By using a bus we  therefore  expect
to  exploit a cost/performance/concurrency "window", and to build
a working prototype in a relatively short time (2-3 years).

2.2  The Intelligent Memory Units

GRIP's IMUs will consist of a megabyte of RAM arranged in  40-bit
words, with a fairly simple bit-slice microprogrammable processor
on the front.  Instead of supporting just READ and WRITE commands
as  normal  memories do, the IMUs will support an instruction set
of  high-level  operations,  chosen  to  support  parallel  graph
reduction.   These  operations  are  the  unit  of indivisibility
supported by GRIP (all  concurrent  machines  must  provide  some
indivisible  operations  to  assure  correct  synchronisation  of
parallel  activities).   In  addition  the  use   of   high-level
operations reduces the bus bandwidth required to communicate with
the IMUs.

2.3  The Processing Elements

The  PEs  are  autonomous  units   responsible   for   performing
reductions  on  the  graph  held  in  the  IMUs.  They will be of
straightforward  design,  based  around  a   microprocessor   (eg
Transputer, MC68020, NS32332), and will include their own private
memory, which is inaccessible to the rest of the system.

The processor within a  PE  executes  a  program  held  in  local
memory.

2.4  Physical arrangements

Although they are logically separate, we will  integrate  several
PEs  and  one  IMU  on  each  board  plugged  into the bus.  This
maximises the use of our scarcest resource (bus slots), and  also
maximises  the  number  of concurrent activities in the system by
putting several on each board.

Most transactions between a PE and an IMU will be carried out  on
a  split  cycle  basis,  to  make  the  best  use  of  scarce bus
bandwidth.  The PE will write a transaction request into  a  fast
transaction  buffer  held  in the bus interface section.  The bus
interface will then acquire the bus (which may take  some  time),
send  all  pending  requests  to  the corresponding buffer on the
destination board and relinquish the bus.  The  request  will  be
processed  by  the  recipient  IMU,  which  will  write  a  reply
transaction into the transaction buffer.  The reply will then get
transferred back to the requesting PE by the same mechanism.

Achieving an implementation of  this  protocol  without  imposing
substantial  latency on transactions is one of the major hardware
challenges of the project.
!
			  GRIP overview



3.  Project status

The GRIP  project  is  funded  by  the  Alvey  directorate  as  a
collaborative  project  between  University  College London, ICL,
High Level Hardware Ltd and Research Software  Ltd.   Three  full
time  Research  Assistants  form  the main team based at UCL, and
work began in the late Autumn of 1985.
!
			  GRIP overview

			   References

[Darl81]    Darlington J and Reeve M, "ALICE - a  multiprocessor
	    reduction  machine  for  the  parallel evaluation of
	    applicative  languages",  Proc  ACM  conference   on
	    Functional   Programming   Languages   and  Computer
	    Architecture, New Hampshire, pp65-75, Oct 1981.

[Darl84]    Darlington   J,   "Functional    programming",    in
	    Distributed  Computing,  ed  Duce, Peter Peregrinus,
	    1984.

[Hoare81]   Hoare CAR, "The Emperor's old clothes", CACM  24(2),
	    pp75-83, Feb 1981.

------------------------------

End of PARSYM Digest
********************

∂01-Dec-85  2336	ACUFF@SUMEX-AIM.ARPA 	New Explorer Software  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Dec 85  23:36:37 PST
Date: Sun 1 Dec 85 23:34:40-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: New Explorer Software
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12163826515.73.ACUFF@SUMEX-AIM.ARPA>

  I've been using the newest Explorer software (2.48) for several
days, and I think it's at least as stable as the 2.12 that most people
are using.  Also, I've generated a few more bug fixes for it, and we
have most sources for it.  Thus, I'd like everyone to start using this
new sub-version.  I'll be putting it on all the pool machines soon,
and, unless I hear differently I'll put it on LOD2 on private machines
as well, if the LOD2 has 2.12 now, so if you feel you don't want to
use the new version, please let me know quickly.

  With 2.48, X8 (pool machine outside of my office) will become the
SYS host, where system sources and various declarations are stored, so
you might want to avoid using it if another machine is free, since
others will be using it from the net for file access.

   As a reminder, you should do (make-system 'ksl-patches :nowarn
:noconfirm :silent) in your login-init.lisp file to fix as many known
bugs as we have fixes for.

  There are many new bug reports in Sumex:<LispM>exp-pre2.bugs.  Just
the new ones can be found in x1:acuff;new-bugs.text.  You might want
to glance through these so you know what problems have been
encountered.

   Again, if you find a bug or problem with an Explorer, please let me
know.

	-- Rich
-------

∂01-Dec-85  2347	ACUFF@SUMEX-AIM.ARPA 	Compatibility between Symbolics 36xx and TI Explorer 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 1 Dec 85  23:47:05 PST
Date: Sun 1 Dec 85 23:45:23-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Compatibility between Symbolics 36xx and TI Explorer
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12163828463.73.ACUFF@SUMEX-AIM.ARPA>

   If you're porting code from a Symbolics to an Explorer, you should
take a look at Sumex:<LispM>36xx-Explorer.lisp.  This file lists many
differences with suggestions about what can be done to circumvent
them, and some code that provides functionality on the Explorer.  I
suggest putting (make-system '36xx-Explorer :noconfirm :silent) into
your login-init.lisp if you plan on developing on both machines.

   Also, if you discover a major incompatibility not covered by
36xx-Explorer, *PLEASE* let me know.

	-- Rich
-------

∂02-Dec-85  0859	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  08:59:27 PST
Date: Mon 2 Dec 85 08:51:50-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
cc: library@SU-SCORE.ARPA, maslin@SUMEX-AIM.ARPA
Message-ID: <12163927944.12.RICHARDSON@SU-SCORE.ARPA>

The topic for the CSD Lunch on Dec. 3 at 12:15 in MJH will be Stanford
Patent Agreements with two representatives from the Sponsored Projects
Office attending to answer questions.
-------

∂02-Dec-85  0947	@SU-CSLI.ARPA:admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Dec. 3, 1985 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  09:46:56 PST
Received: from ucbvax.berkeley.edu by SU-CSLI.ARPA with TCP; Mon 2 Dec 85 09:37:20-PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA16750; Mon, 2 Dec 85 09:44:17 PST
Received: by cogsci (5.31/5.16)
	id AA29615; Mon, 2 Dec 85 09:43:33 PST
Date: Mon, 2 Dec 85 09:43:33 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8512021743.AA29615@cogsci>
To: allmsgs@cogsci.berkeley.edu, cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Dec. 3, 1985
Cc: admin@cogsci.berkeley.edu

                         BERKELEY COGNITIVE SCIENCE PROGRAM
                                     Fall 1985
                        Cognitive Science Seminar - IDS 237A
                         Tuesday, December 3, 11:00 - 12:30
                           240 Bechtel Engineering Center
                    Discussion: 12:30 - 1:30 in 200 Building T-4

                       ``An Approach to Conscious Experience''
                                  Bernard J. Baars
                 Langley Porter Neuropsychiatric Institute, U.C.S.F.

                Conscious experience has been widely viewed as a confusing
           and  ill-defined  issue, and most psychologists have avoided it
           until quite recently.  However, there are straightforward  ways
           to  specify reliable empirical constraints on the problem, sim-
           ply by contrasting comparable pairs of events, one of which  is
           conscious  and  the  other  not.  For example, we are typically
           unconscious of highly  predictable  stimuli,  though  there  is
           strong evidence that such stimuli continue to be represented in
           the nervous system.  We are unconscious of automatized actions,
           of  the unattended stream in a selective attention paradigm, of
           conceptual presuppositions, of the unconscious meaning of  per-
           ceptual  and linguistic ambiguities, of lexical access, syntac-
           tic rule-application, etc.  In all these cases the  unconscious
           information  continues  to  be  represented and processed.  Any
           complete theory of conscious experience is bounded by, and must
           ultimately account for, the entire set of such contrasts.

                The empirical constraints converge on a model of the  ner-
           vous  system  as  a  distributed  collection  of specialists---
           automatic, unconscious, and very efficient.   Consciousness  is
           associated  in this system with a "global workspace"---a memory
           whose contents are broadcast to all the specialists.   Special-
           ists  can  complete  or  cooperate  for  access  to  the global
           workspace, and those that succeed can recruit and control other
           specialists  in  pursuit  of  their goals.  Over the past seven
           years this Global Workspace approach has  been  extended  to  a
           number  of  puzzling  issues,  including action control and the
           neurophysiological basis of consciousness.
           ----------------------------------------------------------------
           ELSEWHERE

           Peter  Labudde,  SESAME  Group  visiting   scholar   from   EMS
           Samedan/Switzerland, will speak on "Experiments for students in
           everyday physics" at the SESAME Colloquium on Monday,  December
           2 at 4:00 p.m. in 2515 Tolman Hall, Campus.

           Jim Greeno, EMST and Cognitive Science Program, will  speak  on
           "How Problems Differ" at the Cognitive Psychology Colloquium on
           Friday, December 6 at 4:00 p.m. in the Beach Room, 3105  Tolman
           Hall, Campus.
           ----------------------------------------------------------------

∂02-Dec-85  0947	admin%cogsci@BERKELEY.EDU 	UCB Cognitive Science Seminar--Dec. 3, 1985
Received: from UCBVAX.BERKELEY.EDU by SU-AI.ARPA with TCP; 2 Dec 85  09:47:43 PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA16750; Mon, 2 Dec 85 09:44:17 PST
Received: by cogsci (5.31/5.16)
	id AA29615; Mon, 2 Dec 85 09:43:33 PST
Date: Mon, 2 Dec 85 09:43:33 PST
From: admin%cogsci@BERKELEY.EDU (Cognitive Science Program)
Message-Id: <8512021743.AA29615@cogsci>
To: allmsgs@cogsci.berkeley.edu, cogsci-friends@cogsci.berkeley.edu
Subject: UCB Cognitive Science Seminar--Dec. 3, 1985
Cc: admin@cogsci.berkeley.edu

                         BERKELEY COGNITIVE SCIENCE PROGRAM
                                     Fall 1985
                        Cognitive Science Seminar - IDS 237A
                         Tuesday, December 3, 11:00 - 12:30
                           240 Bechtel Engineering Center
                    Discussion: 12:30 - 1:30 in 200 Building T-4

                       ``An Approach to Conscious Experience''
                                  Bernard J. Baars
                 Langley Porter Neuropsychiatric Institute, U.C.S.F.

                Conscious experience has been widely viewed as a confusing
           and  ill-defined  issue, and most psychologists have avoided it
           until quite recently.  However, there are straightforward  ways
           to  specify reliable empirical constraints on the problem, sim-
           ply by contrasting comparable pairs of events, one of which  is
           conscious  and  the  other  not.  For example, we are typically
           unconscious of highly  predictable  stimuli,  though  there  is
           strong evidence that such stimuli continue to be represented in
           the nervous system.  We are unconscious of automatized actions,
           of  the unattended stream in a selective attention paradigm, of
           conceptual presuppositions, of the unconscious meaning of  per-
           ceptual  and linguistic ambiguities, of lexical access, syntac-
           tic rule-application, etc.  In all these cases the  unconscious
           information  continues  to  be  represented and processed.  Any
           complete theory of conscious experience is bounded by, and must
           ultimately account for, the entire set of such contrasts.

                The empirical constraints converge on a model of the  ner-
           vous  system  as  a  distributed  collection  of specialists---
           automatic, unconscious, and very efficient.   Consciousness  is
           associated  in this system with a "global workspace"---a memory
           whose contents are broadcast to all the specialists.   Special-
           ists  can  complete  or  cooperate  for  access  to  the global
           workspace, and those that succeed can recruit and control other
           specialists  in  pursuit  of  their goals.  Over the past seven
           years this Global Workspace approach has  been  extended  to  a
           number  of  puzzling  issues,  including action control and the
           neurophysiological basis of consciousness.
           ----------------------------------------------------------------
           ELSEWHERE

           Peter  Labudde,  SESAME  Group  visiting   scholar   from   EMS
           Samedan/Switzerland, will speak on "Experiments for students in
           everyday physics" at the SESAME Colloquium on Monday,  December
           2 at 4:00 p.m. in 2515 Tolman Hall, Campus.

           Jim Greeno, EMST and Cognitive Science Program, will  speak  on
           "How Problems Differ" at the Cognitive Psychology Colloquium on
           Friday, December 6 at 4:00 p.m. in the Beach Room, 3105  Tolman
           Hall, Campus.
           ----------------------------------------------------------------

∂02-Dec-85  1002	BSCOTT@SU-SCORE.ARPA 	Tuesday Lunch
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  10:02:02 PST
Date: Mon 2 Dec 85 09:51:25-PST
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: Tuesday Lunch
To: Faculty@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12163938790.29.BSCOTT@SU-SCORE.ARPA>


The persons representing Sponsored Projects to discuss the patent and copy-
right issues are Clive Liston, Associate Director for Intellectual Property
Administration, his assistant, Cornelia Shonkwiler, and Phyllis Hughes, the
SPO contract officer for Computer Science.

Betty
-------

∂02-Dec-85  1015	ZEC@SU-CSLI.ARPA 	interactions of morphology, syntax, and discurse    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  10:15:20 PST
Date: Mon 2 Dec 85 10:04:07-PST
From: Draga Zec <ZEC@SU-CSLI.ARPA>
Subject: interactions of morphology, syntax, and discurse
To: folks@SU-CSLI.ARPA
cc: zec@SU-CSLI.ARPA


Copies of the paper by Draga Zec to be discussed in this Thursday's meeting of
the Interactions of Morphology, Syntax, and Discourse group, will be available
at the CSLI Receptionists' desk on Wednesday morning.
-------

∂02-Dec-85  1041	BSCOTT@SU-SCORE.ARPA 	RFP from IBM 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  10:41:35 PST
Date: Mon 2 Dec 85 10:41:17-PST
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: RFP from IBM
To: AC@SU-SCORE.ARPA
cc: BScott@SU-SCORE.ARPA
Message-ID: <12163947866.29.BSCOTT@SU-SCORE.ARPA>


Nils has received a Request for Proposal OEM DP Survey for Mobile 
Environment.  The statement of work includes:

1.  General.  This survey will identify alternative data processing systems
    and architectures for performing in a mobile environment while meeting
    the performance objectives established for this study.  Baseline performance
    and capacity for evolutionary growth are of primary interest.

2.  Objectives.  The objectives of this project are:

    a.  Identify candidate computers which achieve projected performance re-
        quirements.

    b.  Rank the candidate computers which demonstrate the potential of meeting
        performance requirements, and which demonstrate a growth capability for
        future needs.

Proposals are due at IBM (Westgate Village, CA) on 12/20/85.  The RFP says
the performance period is 2/3/85 - 5/30/86--but they must mean 2/3/86 -
5/30/87.

No budget dollar amount is mentioned.

If you are expecting this RFP, or if you are interested in submitting a pro-
posal, I have full information in my office.

Betty
-------

∂02-Dec-85  1106	RICHARDSON@SU-SCORE.ARPA 	Faculty Accomplishments 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  11:06:44 PST
Date: Mon 2 Dec 85 11:06:16-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Faculty Accomplishments
To: ac@SU-SCORE.ARPA
Message-ID: <12163952415.12.RICHARDSON@SU-SCORE.ARPA>

The "Faculty Accomplishments" which were distributed to you on 11/19/85
are due today. I will be delivering those that I receive by today to SOE
tomorrow.

Anne
-------

∂02-Dec-85  1121	EMMA@SU-CSLI.ARPA 	Bibtex emacs library 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  11:16:14 PST
Date: Mon 2 Dec 85 11:03:18-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Bibtex emacs library
To: folks@SU-CSLI.ARPA
Tel:  497-3479


  I am looking for the source file for the bibtex emacs library (i.e., the
file bibtex.emacs) and I strongly suspect that it is somewhere at SRI, but
I can't find it.  Do any of you have it?

  
Emma

ps. I got the compiled library from Hans Uszokoreit who, I believe, adapted
the library from the biblio library (I have the source file for that).  
However, Hans is unavailable at this time (and for another week at least).
-------

∂02-Dec-85  1121	REGES@SU-SCORE.ARPA 	anyone free on 11/5?    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  11:14:10 PST
Date: Mon 2 Dec 85 11:13:05-PST
From: Stuart Reges <REGES@SU-SCORE.ARPA>
Subject: anyone free on 11/5?
To: faculty@SU-SCORE.ARPA
Office: Margaret Jacks 210, 497-9798
Message-ID: <12163953656.49.REGES@SU-SCORE.ARPA>

I have asked several people individually, but I still seem to be without a
speaker for this Thursday's department lecture series.  This is CS300, the
series intended primarily for new PhD students who are trying to find out what
kind of research is going on here.  The talk is scheduled from 2:45 to 4.  If
you can speak, please let me know.  Otherwise I will have to cancel.
-------

∂02-Dec-85  1123	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Announcing a Logic Meeting at UCLA 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  11:22:42 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 2 Dec 85 11:12:14-PST
Date: 02 Dec 85  1109 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Announcing a Logic Meeting at UCLA
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    


			  Logic Meeting at UCLA
			 January 24, 25, 26 1986

There will be a very informal gathering of logicians at UCLA, starting
with lunch at 12:00 noon on Friday, January 24 and ending at noon on Sunday
January 26.  As usual, there will be two talks on Friday afternoon, three
talks and a session on problems and short announcements on Saturdy, and two
more talks on Sunday morning.  The entertainment will include a get together
on Friday afternoon and a party on Saturday night.

    The speakers this year are M. Beeson, H. Gaifman, Ph. Kolaitis, A.
Louveau, M. Lerman, M. Magidor and J. Steel.  In addition, the following
logicians have indicted that they are likely to attend: J. Barwise, C. C.
Chang, R. Dougherty, S. Feferman, M. Foreman, A. Horn, E. Kastanas, A. 
Kechris, P. Maddy, D. A. Martin, W. Mitchell, J. R. Moschovakis, Y.
Moschovakis, R. Solovay and H. Woodin.

                           EVERYONE IS INVITED

    This first announcement is being sent to a small number of persons on a
a hastily contrived mailing list, so please mention the meeting to your friends
who might be interested in coming.  If you plan to come, drop me (Y.Moschovakis)
a note  or call the number below.

    We have reserved rooms for the meeting at the following hotels which are
quite near the Campus.

        Claremont, 1044 Tiverton Ave., (213) 208 5957
        Royal Palace Westwood Hotel (the old Tiverton)
                Tiverton Ave., (213) 208 6677
        Holiday Inn - Brentwood, 170 North Church Lane, (213) 476 6411
        Bel-Air Sand Hotel, 11461 Sunset Blvd.  (213) 476 6571

    The first tow of these are the closest and the least expensive.  Please
make your own reservations, mentioning the UCLA Logic meeting.

    In accordance with the informal nature of the meeting, the lectures will
be primarily expository, and there will be plenty of time for talk.

A second and final announcement will be mailed on January 6.  That wil
contain the formal program of talks, instructions for getting to UCLA, etc.,
as well as the names of all those who notify me (YM) before Christmas that 
they are planning to come.

Telephones: For additional info call our secretary Elain Barth during working
hours at (213) 825 1148, or myself (YM) at home, evenings,  (213) 394 5201.

Yiannis N. Moschovakis
Department of Mathematics
University of California
Los Angeles, CA 90024
math.ynm@LOCUS.UCLA.EDU

∂02-Dec-85  1205	EMMA@SU-CSLI.ARPA 	bibtex emacs library 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  12:05:03 PST
Date: Mon 2 Dec 85 11:48:26-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: bibtex emacs library
To: folks@SU-CSLI.ARPA
Tel:  497-3479


  The file I asked for is being retrieved from archives over at SRI-AI.

Many thanks to those of you who sent in suggestions.

Emma 

ps. Sorry for sending these messages to folks.
-------

∂02-Dec-85  1620	RICHARDSON@SU-SCORE.ARPA 	Near West Campus Invitation  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  16:20:06 PST
Date: Mon 2 Dec 85 15:45:02-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Near West Campus Invitation
To: ac@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12164003163.39.RICHARDSON@SU-SCORE.ARPA>

Recently you received a memo from Jim Gibbons inviting you to attend a
Near West Campus review of work to date on December 16 from 4-6. I just
spoke with Jane Johnston on this and she has advised me that this meeting
will take place in Durand 450. Please make a note of it.

-------

∂02-Dec-85  1716	DAVIES@SUMEX-AIM.ARPA 	No meeting this Wednesday  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  17:16:04 PST
Date: Mon, 2 Dec 1985  17:14 PST
Message-ID: <DAVIES.12164019531.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: No meeting this Wednesday
cc:   Davies@SUMEX-AIM.ARPA

There will be no weekly (?) meeting this Wednesday, December 4.

	-- Byron

∂02-Dec-85  1727	MODICA@SU-SCORE.ARPA 	survey forms 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  17:26:55 PST
Date: Mon 2 Dec 85 16:50:11-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: survey forms
To: instructors@SU-SCORE.ARPA
Message-ID: <12164015024.25.MODICA@SU-SCORE.ARPA>

This is a reminder to pick up survey forms from me (Gina) in MJH 030.
It is very important that the class be surveyed, so please come by for
the forms.
-Gina
-------

∂02-Dec-85  1732	MODICA@SU-SCORE.ARPA 	grades  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  17:31:59 PST
Date: Mon 2 Dec 85 16:57:54-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: grades
To: instructors@SU-SCORE.ARPA
Message-ID: <12164016427.25.MODICA@SU-SCORE.ARPA>

End Quarter Reports will be available in my office (MJH 030) 
starting tomorrow afternoon.
Please come by to pick them up or let me know where you want
yours mailed. Please remember to specify which course you are
teaching.
Thank You.
Gina
-------

∂02-Dec-85  1802	TAJNAI@SU-SCORE.ARPA 	Bell Fellowship   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 2 Dec 85  18:02:02 PST
Date: Mon 2 Dec 85 18:00:15-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: Bell Fellowship
To: Faculty@SU-SCORE.ARPA
Message-ID: <12164027778.17.TAJNAI@SU-SCORE.ARPA>

We have not received the "call" for nominations for the Bell
Fellowship, but we will receive it shortly.  Last year the deadline
was January 15 at Bell.  The nominees need at least a month to get
their packets together -- especially considering lost time over the
holidays.

Feel free to send your nominations now.  I will notify you as soon
as I get the "call" and when nominations will be closed.

We want to nominate 3 first or second year PHD students. 

Carolyn
-------

∂03-Dec-85  0923	@SU-CSLI.ARPA:CLT@SU-AI.ARPA 	Seminar in Logic and Foundations   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  09:23:21 PST
Received: from SU-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 3 Dec 85 09:17:41-PST
Date: 03 Dec 85  0909 PST
From: Carolyn Talcott <CLT@SU-AI.ARPA>
Subject: Seminar in Logic and Foundations  
To:   "@DIS.DIS[1,CLT]"@SU-AI.ARPA    


Speaker: Ian Mason, Philosophy Dept., Stanford

Title:   Proving properties of destructive LISP programs (cont'd)

Time:    Friday Dec. 6, 1985, 12:00-1:00 P.M.

Place:   Math. Faculty Lounge, Stanford, Room 383N.

                                 S. Feferman

∂03-Dec-85  0923	ullman@su-aimvax.arpa 	CS300 talk  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  09:23:21 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 3 Dec 85 09:19:19 pst
Date: Tue, 3 Dec 85 09:19:19 pst
From: Jeff Ullman <ullman@diablo>
Subject: CS300 talk
To: mugs@sumex, nail@diablo

It appears that I will be giving a NAIL! overview at the
CS300 lecture, 2:45 PM, Thursday 12/5, in Hist. Corner 202.

∂03-Dec-85  1101	TAJNAI@SU-SCORE.ARPA 	Bell Fellowship   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  11:01:13 PST
Date: Tue 3 Dec 85 10:54:10-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: Bell Fellowship
To: faculty@SU-SCORE.ARPA
Message-ID: <12164212357.10.TAJNAI@SU-SCORE.ARPA>


I called Bell this morning and got some details.  The deadline for
applications is January 15, 1986.  Therefore, the deadline for
nominations is Friday, Dec. 13.

U.S. citizens or permanent residence required.

Carolyn
-------

∂03-Dec-85  1148	LANSKY@SRI-AI.ARPA 	Next Monday's PLANLUNCH  
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  11:48:18 PST
Date: Tue 3 Dec 85 11:44:55-PST
From: LANSKY@SRI-AI.ARPA
Subject: Next Monday's PLANLUNCH
To: planlunch.dis: ;

	REASONING ABOUT CONTROL IN A HIGH-LEVEL COMPUTER VISION SYSTEM

			  Leonard Wesley
		    SRI International, AI Center

	 	    11:00 AM, MONDAY, December 9
       SRI International, Building E, Room EJ228 (new conference room)


If you built an expert system, how would you expect it to decide what to 
do next in complex situations? Typically there are several alternative 
actions it might take to reach some goal. In some cases, the best alternative 
is clear or the choices do not warrant extensive analysis. At times the 
consequences of pursuing some action justify expending the effort to obtain 
the necessary information to analyze the pros and cons of choosing a 
particular alternative. 

Most would agree that the information that is needed to reach any decision is,
to some degree, uncertain, imprecise, and occasionally inaccurate (called 
"evidential" information). Clearly knowledge about the certainty, precision, 
and accuracy of information can be used to improve a system's 
ability to reason about (i.e., control) its actions. In this talk, we shall 
describe how this might be accomplished by an expert system in the domain 
of high-level computer vision. We shall explain why we view Shafer's theory 
of belief functions as being better suited than some other models as a 
theoretical foundation for representing evidential information and reasoning 
about control. Results from a large number of image interpretation 
experiments will be presented to demonstrate how a system's performance can be
improved when Shafer's theory is soundly exploited. Finally, we shall briefly 
describe how our approach to control might be extended to an evidential-based 
framework for planning under uncertainty. 

-------

∂03-Dec-85  1203	NILSSON@SU-SCORE.ARPA 	Faculty Meeting  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  12:03:29 PST
Date: Tue 3 Dec 85 12:03:16-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Faculty Meeting
To: ac@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12164224937.32.NILSSON@SU-SCORE.ARPA>

Here's a tentative agenda for next Tuesday's special faculty meeting.
Please let me know if you have additional agenda items that cannot
wait until our regular meeting in January.  -Nils

Tentative Agenda for December 10, 1985 General Faculty Meeting
(Special Meeting since we have a lot to talk about and ought not
to wait until regular meeting in January.)

MJH 146, 1:15 pm  [NOTE earlier time--since it's a long agenda]

1.  Reports
     Financial, Betty Scott
          (items: salary research offsets, ...)
     Facilities, Lester Earnest
          (items:  Courtesy Computer Accts Policy, APS, Long-range Plans, ...)
     PhD Program Committee Progress Report, Terry Winograd
     Forum, Carolyn Tajnai
     Others as may be appropriate

2.  Report on Long-Range Planning Process, Nils Nilsson

3.  Report on Undergraduate Major Planning
     Curriculum Matters, Jeff Ullman
     Summary of Curriculum approval process with UG Council and C-US, Jeff 
                                        Ullman and Nils Nilsson
     Summary of discussions with Dean Gibbons regarding resources, Nils Nilsson
     Discussion

4.  Report on Progress on Searches, Nils Nilsson
     Robotics
     Systems
     Possible New Searches

5.  Proposed Consulting Professors (Tenenbaum and Hayes), Nils Nilsson

6.  New Patent and Copyrights Policy, Betty Scott

7.  Proposal on "Visiting Professorship,"  John McCarthy and Nils Nilsson

8.  Suggestion for a Survey of CSD Graduates, Bob Floyd

Adjourn 3:45 (at the latest because we get kicked out of the room then).
-------

∂03-Dec-85  1315	MODICA@SU-SCORE.ARPA 	End Quarter Reports    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  13:12:53 PST
Date: Tue 3 Dec 85 13:11:12-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: End Quarter Reports
To: instructors@SU-SCORE.ARPA
Message-ID: <12164237302.24.MODICA@SU-SCORE.ARPA>

I am putting EQRs in your mailboxes this afternoon.
Also -- please please come by and pick up survey forms.
-Gina
-------

∂03-Dec-85  1915	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #42  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  19:15:08 PST
Date:  3 Dec 85 1824-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #42
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 4 Dec 1985      Volume 1 : Issue 42

Today's Topics:

         Response to PARSYM Hardware Survey: CEDAR (Illinois)
                    SEAI Parallel Computing Survey

----------------------------------------------------------------------

Date: Fri, 29 Nov 85 12:31:36 cst
From: harrison@uicsrd.CSRD.UIUC.EDU (Luddy Harrison)

>>        The Third PARSYM Survey: Computer Hardware
>>
>>             Part I (Easy)
>>What computer hardware are you using or developing?
   The CEDAR machine will be constructed from eight-processor multiprocessors
   designed and built by Alliant computers, and modified by us.  The principle
   piece of hardware we are designing is the omega network which will connect
   the processors to a large, interleaved shared memory.  In addition, we
   are designing vector prefectching boards, "processors" which sit between
   each memory bank and the network (I'll describe their function later)
   and possibly, but not certainly, a cache system for the global memory.

>>What are its components (processors, memory) and how many; how do the
>>processors and memory communicate?
   We will double the machine size periodically.  Currently,
   we have 16 processors (2 Alliants); given the machine room size and
   the price of the Alliant machines, ~128 is the practical maximum.

>>SIMD/MIMD?  Shared memory or distributed memory?
   MIMD.  Well, ... Each of the eight processors in an Alliant machine
   has a full complement of vector instructions, so that the machine looks
   like a pure MIMD machine when the processors are in scalar mode, but
   a hybrid when both inter-processor concurrency and vector concurrency
   are being used.  Shared memory.  Each "cluster" (Alliant machine) has
   a shared memory, to which the processors are connected via a crossbar.
   There is a cache between each processor and its corresponding cluster
   memory, which is kept coherent by hardware.  In addition, there is
   a large "global" memory, added by us, to which every processor is connected
   via an omega network (packet switching) composed of 8x8 crossbars.
   The data path of the network is 64 bits wide.

>>What model of concurrency are you investigating with this system?
   ?? Is MIMD a model?  If so, then MIMD.

>>What would you like to see in your next hardware system?
   Not currently planning a next system.  (Still at work on this one!)



>>      Part II (Optional, for extra credit)


>>0. Project summary:

>>What model of concurrency (e.g., dataflow, communicating sequential
>>processes, actors, futures, etc.) is your work based on?  What
>>language, if any, are you using in your work?
   I hope that it does not reflect an essential inelegance that I can't
   come up with a model name.  We have a big, fast, shared memory machine
   and we will be running big, fast programs on it!

   Fortran is the first language, for better or worse, to appear on
   the machine.  We are planning a compiler / run-time environment
   for lisp.  Alliant is talking about making a parallel C available,
   but no specifics have come forth yet.  Probably, many langauges
   will be supported in the end.

  >>1. Architectural classification:

>>SIMD or MIMD?
  MIMD + vector instructions in each processor.

>>2. Processing elements:

>>Describe each processor in the system in terms of type (e.g., 68000
>>equivalent, Lisp machine, custom), "power" (e.g., in MIPS or relative
>>to a Vax or 68000), and memory size.  If the processors are simulated
>>then also describe the uniprocessor and mechanism for simulation.  If
>>the processors in the system are not of uniform type then describe
>>each type and their relationship to each other.

   Each cluster has eight so-called "CE's" (computational elements), which
   runs a 68020 instruction set, in addition to the vector instructions
   which are sped along by some fast floating point hardware.  While
   the 68020 instruction set is used, the processors are not 68020s: they
   are built from gate arrays, and each is roughly 2-3 times as fast as a VAX
   785.   There are also a number of "IP's" (interactive processors) in
   each cluster, which are used primarily to handle sequential jobs, such
   as editing.


>>How many processors are there in the minimum and maximum configurations?
>>How many are you currently using?

   We have 16 processors (two clusters) at present.  The idea is for the
   machine to be easily scalable.  No "new" hardware is required to
   double the machine size, just more of the old.  And money.
   Price (and machine room size) aside, 1K+ processors are feasible.


>>3. Memory system:

>>Describe how memory is distributed and accessed in the system.  Shared
>>memory or distributed memory?  Is there a single address space or does
>>each processor have its own address space?  What sort of caching is
>>employed, if any: snoopy, write-through, or something else?  What
>>sizes are the caches?

   One great big address space.  Data is labeled by the compiler or
   by the user as shared or private, at the processor, cluster, or global
   level.  There are 64Kbytes of cache between the processors of a cluster
   and a cluster memory.  Snoopy cache.  The cache is "a complex beast,"
   as one of our hardware people says.


>>What are the minimum and maximum amounts of memory?  How much are you
>>currently working with?

   Each module of global memory will have 8Mbytes (1Mword - 64 bit words).
   There will be one module per processor.  For a 32 processor machine, this
   means 256Mbytes of global memory.  If a 1Mbit chip becomes available
   soon, the figure will quadruple.   The cluster comes with up to 8Mbytes
   of cluster memory; extensions to the backplane to allow more than this
   are planned.


>>4. Communication system:

>>How are processors connected to each other and to memory?  Describe in
>>terms of topology (e.g., tree, systolic array, pipe, grid), switching
>>type (e.g., LAN, bus, Butterfly switch), and throughput (e.g., bus
>>transmission rates, time to access memory on remote nodes, time to
>>broadcast a message to all nodes).

Each stage in the interconnection network causes a delay of 170ns;
this gives a peak bandwidth, for a 32 processor machine, of 12Gbits/s
for the interconnection network. To realize such a bandwidth, each processor
will have to keep its memory port busy by the use of lookahead buffers,
vector prefetching hardware, etc.  The effects of caching and network
conflicts must also be accounted for.  No numbers are available on the
effective bandwidth yet.

>>5. Synchronization:

>>What hardware mechanisms does the system supply to support
>>synchronization?

   Lots and lots.  At the cluster level, Alliant provides several mechanisms
   for synchronization, including loop self-scheduling, and a variant of
   the usual test and set instructions.  The global memory network will
   have, at each memory module, a simple processor which will allow a
   variety of fetch/test/and op instructions to be performed on globally
   shared data.  These instructions are not performed, as in the combining
   network of the ultracomputer, in the switches of the network, but
   rather at the memory-module-side of the network.  See papers by
   Prof. Pen Yew (Illinois) for a description of the various instructions
   so implemented.

  >>6. Pragmatics

>>Is your system geared towards either specific programming languages or
>>applications?

   At present, the system is used almost exclusively by numerical analysts.
   But there is hope.  As I've mentioned, we are planning to implement lisp
   on the system.

>>Is the hardware of your system generally available or, if custom-made,
>>likely to be available in future?

   Alliant machines are generally available.  The modifications to the machine
   for the purposes of building CEDAR, as well as the global network boards
   and gate arrays, are, as yet, not generally available.

>>If you're now using a multiprocessor system, how does the architecture
>>of your development system correspond to your research target
>>architecture?

   Very closely.  Some differences have occurred because the Alliant
   machines were not designed to be themselves the components of a larger
   machine.

>>What are the advantages of your hardware setup, particularly for the
>>problems you are investigating?  What are the disadvantages?  Is there
>>a system that would better meet your needs?

   The biggest advantage is that someone else designed and built the
   processors; this is the usual situation when composing a machine
   of microprocessors, but it is even more appreciable when the processors
   are themselves high performance machines.  It means that we can concentrate
   on the central problem of the  design, namely, the global memory and
   omega network.  This principle advantage is also the principle disadvantage:
   we must incorporate our design into an existing architecture, which, I
   am told by those doing the hardware design, can be a painful process.

>>------------------------------

-Luddy Harrison

Center for Supercomputer Research and Development
302D Talbot Laboratory
104 South Wright Street
Urbana IL 61801-2932

(217) 333 4168

------------------------------

Date: Tue, 3 Dec 1985  18:01 PST
From: DAVIES@SUMEX-AIM.ARPA
Subject: SEAI Parallel Computing Survey

An advertising blurb showed up in the mail that PARSYM readers may
find interesting.  I include the text for informational purposes, not
as an advertisement, though I have included pointers for those who
want to check further.  I have no connection with SEAI Technical
Publications.

If anyone has seen the report itself, please pass along your opinions
to PARSYM.  -- BD

                           ----------------

Parallel Processing: The Technology of Fifth Generation Computers

Publication Date: September 1985
Project Manger: Richard K. Miller
ISBN: 0-89671-067-X
Price: $525

Preface: A new era in computing is beginning in which very
high-performance systems will be dominated by parallel architectural
systems.  The high cost of today's supercomputers prevent their use
for all but the most important problems, and the traditional von
Neumann computer architecture has been pushed to its limits.
Computers with parallel architectures have already been commercially
introduced.  These machines will provide advanced capabilities for
users in areas such as artificial intelligence, signal processing,
engineering/scientific simulation, and voice recognition.

Content: The report explains the workings of several SIMD, MIMD, and
dataflow architectures in non-theoretical terminology.  The impact of
parallel processing computers on business, industry, national defense,
science, engineering, and the quality of life is examined.  The
parallel processing projects and products of 35 international research
groups and 19 leading corporations are presented.  A survey of experts
in the field explores opinions and forecasts on general architecture,
problem solving strategies, and applications.  Views of experts in the
United States, Japan, and Europe are compared.  The international
markets for parallel processing computers are examined for 1986, 1988,
and 1990.

Intended audience: Engineers and scientists applying simulation,
modeling, and advanced problem solving with an interest in using
leading edge computatational tools; professionals applying artificial
intelligence technologies who are interested in next-generation
hardware; strategic planners involved in electronics, computer and
semiconductor businesses.


                               Contents

Executive Summary
1 - Introduction
2 - The need for a new generation of computing
    2.1 The emergence of AI
    2.2 Limitations in vision technology
    2.3 Limitations in speech recognition
    2.4 Limitations in natural language understanding
    2.5 Limitations in expert systems
    2.6 Limitations in supercomputers
3 - The technology of parallel procesing
    3.1 The von Neumann bottleneck
    3.2 Exploring new architectures
    3.3 What is parallel processing?
    3.4 SIMD architectures
    3.5 Pipeline processing
    3.6 MIMD architectures and dataflow
    3.7 Examples of parallel architectural design
	3.7.1 CEDAR at the University of Illinois
	3.7.2 PASM at Purdue University
	3.7.3 The Columbia Non-Von Machine
    3.8 Pioneer commercial parallel machines
	3.8.1 Parallelism in supercomputers
	3.8.2 Commercial massively parallel machines
	3.8.3 Governmental support of parallel processing
    3.9 Research in Japan
	3.9.1 The Fifth Generation Research Project
	3.9.2 National Superspeed Computer Project
    3.10 The MCC Advanced Generation Computer
    3.11 How much parallelism
    3.12 Applications for parallel processing
    3.13 Specialized machines
    3.14 Signal processing
    3.15 Symbolic processing
    3.16 Multifunction machines
4 - A new breed of software
    4.1 Software for AI
    4.2 Software for the non-von Neumann environment
    4.3 LISP vs. Prolog
    4.4 New languages for parallel processing
    4.5 Restructuring existing Fortran programs
    4.6 Software developement: an iterative process
    4.7 Task dedicated software
    4.8 Automatic programming
5 - The impact of Fifth Generation computers
    5.1 A new breed of computers
    5.2 What will Fifth Generation computers do?
    5.3 User interfaces
    5.4 A potential shift in world economics
    5.5 The critical importance of parallel computing
    5.6 Impact on business and office management
    5.7 Impact on factory automation
    5.8 Impact on national defense
    5.9 Impact on science and engineering
    5.10 Impact on the quality of life
    5.11 Summary
6 - Research groups to watch
    6.1  The Alvey Directorate
    6.2  Argonne National Laboratory
    6.3  California Institute of Technology
    6.4  Carnegie-Mellon University
    6.5  The Centre de Recherche Information de Montreal
    6.6  Columbia University
    6.7  Electrotechnical Laboratory
    6.8  ESPRIT
    6.9  University of Illinois
    6.10 Institute for New Generation Computer Technology
    6.11 Lawrence Livermore Laboratory
    6.12 Los Alamos National Laboratory
    6.13 University of Manchester
    6.14 Massachusetts Institute of Technology
    6.15 University of Michigan
    6.16 Microelectronics and Computer Technology Corp.
    6.17 University of Minnesota
    6.18 National Security Agency
    6.19 New York University
    6.20 University of North Carolina
    6.21 North Carolina State University
    6.22 Purdue University
    6.23 University of Reading
    6.24 Semiconductor Research Corporation
    6.25 Simon Fraser/Queen's University
    6.26 Stanford University
    6.27 System Development Foundation
    6.28 Technical University of Berlin
    6.29 University of Texas at Austin
    6.30 The University of Toronto
    6.31 University of Utah
    6.32 U. S. Department of Defense
    6.33 University of Washington
    6.34 University of Waterloo
    6.35 Yale University
7 - Companies to watch
    7.1  Alliant Computer Systems Corporation
    7.2  Automatix, Inc.
    7.3  Bolt Beranek and Newman, Inc.
    7.4  Control Data Corporation
    7.5  Cray Research, Inc.
    7.6  Denelcor, Inc.
    7.7  Digital Equipment Corporation
    7.8  Eastman Kodak
    7.9  ELXSI
    7.10 Encore Computer Corporation
    7.11 ETA Systems
    7.12 Floating Point Systems, Inc.
    7.13 General Electric
    7.14 Goodyear Aerospace
    7.15 IBM
    7.16 Intel Scientific Computers
    7.17 Kurzweil Applied Intelligence
    7.18 Schlumberger
    7.19 Thinking Machines Corporation
8 - Delphi survey
    8.1 General architecture
    8.2 Connections
    8.3 Memory
    8.4 Problem solving
    8.5 Programming languages
    8.6 Applications
    8.7 Processing limit for AI applications
    8.8 Comparison: United States, Japan, Europe
    8.9 Summary
9 - Market forecast
    9.1 Artificial intelligence potential market
    9.2 Scientific and engineering potential market
    9.3 Market forecast: United States
    9.4 Market forecast: Japan
    9.5 Market forecast: Europe
    9.6 International market leadership
    9.7 Summary
10 - For further reading
     10.1 Advanced generation computing periodicals
     10.2 Books
     10.3 Conference proceedings
     10.4 Technical papers and articles

(For information call (404) 342-9638 or write to:
     SEAI Technical Publications
     P.O. Box 590
     Madison, GA 30650)

------------------------------

End of PARSYM Digest
********************

∂03-Dec-85  1934	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #42  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  19:34:38 PST
Date:  3 Dec 85 1824-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #42
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest           Wednesday, 4 Dec 1985      Volume 1 : Issue 42

Today's Topics:

         Response to PARSYM Hardware Survey: CEDAR (Illinois)
                    SEAI Parallel Computing Survey

----------------------------------------------------------------------

Date: Fri, 29 Nov 85 12:31:36 cst
From: harrison@uicsrd.CSRD.UIUC.EDU (Luddy Harrison)

>>        The Third PARSYM Survey: Computer Hardware
>>
>>             Part I (Easy)
>>What computer hardware are you using or developing?
   The CEDAR machine will be constructed from eight-processor multiprocessors
   designed and built by Alliant computers, and modified by us.  The principle
   piece of hardware we are designing is the omega network which will connect
   the processors to a large, interleaved shared memory.  In addition, we
   are designing vector prefectching boards, "processors" which sit between
   each memory bank and the network (I'll describe their function later)
   and possibly, but not certainly, a cache system for the global memory.

>>What are its components (processors, memory) and how many; how do the
>>processors and memory communicate?
   We will double the machine size periodically.  Currently,
   we have 16 processors (2 Alliants); given the machine room size and
   the price of the Alliant machines, ~128 is the practical maximum.

>>SIMD/MIMD?  Shared memory or distributed memory?
   MIMD.  Well, ... Each of the eight processors in an Alliant machine
   has a full complement of vector instructions, so that the machine looks
   like a pure MIMD machine when the processors are in scalar mode, but
   a hybrid when both inter-processor concurrency and vector concurrency
   are being used.  Shared memory.  Each "cluster" (Alliant machine) has
   a shared memory, to which the processors are connected via a crossbar.
   There is a cache between each processor and its corresponding cluster
   memory, which is kept coherent by hardware.  In addition, there is
   a large "global" memory, added by us, to which every processor is connected
   via an omega network (packet switching) composed of 8x8 crossbars.
   The data path of the network is 64 bits wide.

>>What model of concurrency are you investigating with this system?
   ?? Is MIMD a model?  If so, then MIMD.

>>What would you like to see in your next hardware system?
   Not currently planning a next system.  (Still at work on this one!)



>>      Part II (Optional, for extra credit)


>>0. Project summary:

>>What model of concurrency (e.g., dataflow, communicating sequential
>>processes, actors, futures, etc.) is your work based on?  What
>>language, if any, are you using in your work?
   I hope that it does not reflect an essential inelegance that I can't
   come up with a model name.  We have a big, fast, shared memory machine
   and we will be running big, fast programs on it!

   Fortran is the first language, for better or worse, to appear on
   the machine.  We are planning a compiler / run-time environment
   for lisp.  Alliant is talking about making a parallel C available,
   but no specifics have come forth yet.  Probably, many langauges
   will be supported in the end.

  >>1. Architectural classification:

>>SIMD or MIMD?
  MIMD + vector instructions in each processor.

>>2. Processing elements:

>>Describe each processor in the system in terms of type (e.g., 68000
>>equivalent, Lisp machine, custom), "power" (e.g., in MIPS or relative
>>to a Vax or 68000), and memory size.  If the processors are simulated
>>then also describe the uniprocessor and mechanism for simulation.  If
>>the processors in the system are not of uniform type then describe
>>each type and their relationship to each other.

   Each cluster has eight so-called "CE's" (computational elements), which
   runs a 68020 instruction set, in addition to the vector instructions
   which are sped along by some fast floating point hardware.  While
   the 68020 instruction set is used, the processors are not 68020s: they
   are built from gate arrays, and each is roughly 2-3 times as fast as a VAX
   785.   There are also a number of "IP's" (interactive processors) in
   each cluster, which are used primarily to handle sequential jobs, such
   as editing.


>>How many processors are there in the minimum and maximum configurations?
>>How many are you currently using?

   We have 16 processors (two clusters) at present.  The idea is for the
   machine to be easily scalable.  No "new" hardware is required to
   double the machine size, just more of the old.  And money.
   Price (and machine room size) aside, 1K+ processors are feasible.


>>3. Memory system:

>>Describe how memory is distributed and accessed in the system.  Shared
>>memory or distributed memory?  Is there a single address space or does
>>each processor have its own address space?  What sort of caching is
>>employed, if any: snoopy, write-through, or something else?  What
>>sizes are the caches?

   One great big address space.  Data is labeled by the compiler or
   by the user as shared or private, at the processor, cluster, or global
   level.  There are 64Kbytes of cache between the processors of a cluster
   and a cluster memory.  Snoopy cache.  The cache is "a complex beast,"
   as one of our hardware people says.


>>What are the minimum and maximum amounts of memory?  How much are you
>>currently working with?

   Each module of global memory will have 8Mbytes (1Mword - 64 bit words).
   There will be one module per processor.  For a 32 processor machine, this
   means 256Mbytes of global memory.  If a 1Mbit chip becomes available
   soon, the figure will quadruple.   The cluster comes with up to 8Mbytes
   of cluster memory; extensions to the backplane to allow more than this
   are planned.


>>4. Communication system:

>>How are processors connected to each other and to memory?  Describe in
>>terms of topology (e.g., tree, systolic array, pipe, grid), switching
>>type (e.g., LAN, bus, Butterfly switch), and throughput (e.g., bus
>>transmission rates, time to access memory on remote nodes, time to
>>broadcast a message to all nodes).

Each stage in the interconnection network causes a delay of 170ns;
this gives a peak bandwidth, for a 32 processor machine, of 12Gbits/s
for the interconnection network. To realize such a bandwidth, each processor
will have to keep its memory port busy by the use of lookahead buffers,
vector prefetching hardware, etc.  The effects of caching and network
conflicts must also be accounted for.  No numbers are available on the
effective bandwidth yet.

>>5. Synchronization:

>>What hardware mechanisms does the system supply to support
>>synchronization?

   Lots and lots.  At the cluster level, Alliant provides several mechanisms
   for synchronization, including loop self-scheduling, and a variant of
   the usual test and set instructions.  The global memory network will
   have, at each memory module, a simple processor which will allow a
   variety of fetch/test/and op instructions to be performed on globally
   shared data.  These instructions are not performed, as in the combining
   network of the ultracomputer, in the switches of the network, but
   rather at the memory-module-side of the network.  See papers by
   Prof. Pen Yew (Illinois) for a description of the various instructions
   so implemented.

  >>6. Pragmatics

>>Is your system geared towards either specific programming languages or
>>applications?

   At present, the system is used almost exclusively by numerical analysts.
   But there is hope.  As I've mentioned, we are planning to implement lisp
   on the system.

>>Is the hardware of your system generally available or, if custom-made,
>>likely to be available in future?

   Alliant machines are generally available.  The modifications to the machine
   for the purposes of building CEDAR, as well as the global network boards
   and gate arrays, are, as yet, not generally available.

>>If you're now using a multiprocessor system, how does the architecture
>>of your development system correspond to your research target
>>architecture?

   Very closely.  Some differences have occurred because the Alliant
   machines were not designed to be themselves the components of a larger
   machine.

>>What are the advantages of your hardware setup, particularly for the
>>problems you are investigating?  What are the disadvantages?  Is there
>>a system that would better meet your needs?

   The biggest advantage is that someone else designed and built the
   processors; this is the usual situation when composing a machine
   of microprocessors, but it is even more appreciable when the processors
   are themselves high performance machines.  It means that we can concentrate
   on the central problem of the  design, namely, the global memory and
   omega network.  This principle advantage is also the principle disadvantage:
   we must incorporate our design into an existing architecture, which, I
   am told by those doing the hardware design, can be a painful process.

>>------------------------------

-Luddy Harrison

Center for Supercomputer Research and Development
302D Talbot Laboratory
104 South Wright Street
Urbana IL 61801-2932

(217) 333 4168

------------------------------

Date: Tue, 3 Dec 1985  18:01 PST
From: DAVIES@SUMEX-AIM.ARPA
Subject: SEAI Parallel Computing Survey

An advertising blurb showed up in the mail that PARSYM readers may
find interesting.  I include the text for informational purposes, not
as an advertisement, though I have included pointers for those who
want to check further.  I have no connection with SEAI Technical
Publications.

If anyone has seen the report itself, please pass along your opinions
to PARSYM.  -- BD

                           ----------------

Parallel Processing: The Technology of Fifth Generation Computers

Publication Date: September 1985
Project Manger: Richard K. Miller
ISBN: 0-89671-067-X
Price: $525

Preface: A new era in computing is beginning in which very
high-performance systems will be dominated by parallel architectural
systems.  The high cost of today's supercomputers prevent their use
for all but the most important problems, and the traditional von
Neumann computer architecture has been pushed to its limits.
Computers with parallel architectures have already been commercially
introduced.  These machines will provide advanced capabilities for
users in areas such as artificial intelligence, signal processing,
engineering/scientific simulation, and voice recognition.

Content: The report explains the workings of several SIMD, MIMD, and
dataflow architectures in non-theoretical terminology.  The impact of
parallel processing computers on business, industry, national defense,
science, engineering, and the quality of life is examined.  The
parallel processing projects and products of 35 international research
groups and 19 leading corporations are presented.  A survey of experts
in the field explores opinions and forecasts on general architecture,
problem solving strategies, and applications.  Views of experts in the
United States, Japan, and Europe are compared.  The international
markets for parallel processing computers are examined for 1986, 1988,
and 1990.

Intended audience: Engineers and scientists applying simulation,
modeling, and advanced problem solving with an interest in using
leading edge computatational tools; professionals applying artificial
intelligence technologies who are interested in next-generation
hardware; strategic planners involved in electronics, computer and
semiconductor businesses.


                               Contents

Executive Summary
1 - Introduction
2 - The need for a new generation of computing
    2.1 The emergence of AI
    2.2 Limitations in vision technology
    2.3 Limitations in speech recognition
    2.4 Limitations in natural language understanding
    2.5 Limitations in expert systems
    2.6 Limitations in supercomputers
3 - The technology of parallel procesing
    3.1 The von Neumann bottleneck
    3.2 Exploring new architectures
    3.3 What is parallel processing?
    3.4 SIMD architectures
    3.5 Pipeline processing
    3.6 MIMD architectures and dataflow
    3.7 Examples of parallel architectural design
	3.7.1 CEDAR at the University of Illinois
	3.7.2 PASM at Purdue University
	3.7.3 The Columbia Non-Von Machine
    3.8 Pioneer commercial parallel machines
	3.8.1 Parallelism in supercomputers
	3.8.2 Commercial massively parallel machines
	3.8.3 Governmental support of parallel processing
    3.9 Research in Japan
	3.9.1 The Fifth Generation Research Project
	3.9.2 National Superspeed Computer Project
    3.10 The MCC Advanced Generation Computer
    3.11 How much parallelism
    3.12 Applications for parallel processing
    3.13 Specialized machines
    3.14 Signal processing
    3.15 Symbolic processing
    3.16 Multifunction machines
4 - A new breed of software
    4.1 Software for AI
    4.2 Software for the non-von Neumann environment
    4.3 LISP vs. Prolog
    4.4 New languages for parallel processing
    4.5 Restructuring existing Fortran programs
    4.6 Software developement: an iterative process
    4.7 Task dedicated software
    4.8 Automatic programming
5 - The impact of Fifth Generation computers
    5.1 A new breed of computers
    5.2 What will Fifth Generation computers do?
    5.3 User interfaces
    5.4 A potential shift in world economics
    5.5 The critical importance of parallel computing
    5.6 Impact on business and office management
    5.7 Impact on factory automation
    5.8 Impact on national defense
    5.9 Impact on science and engineering
    5.10 Impact on the quality of life
    5.11 Summary
6 - Research groups to watch
    6.1  The Alvey Directorate
    6.2  Argonne National Laboratory
    6.3  California Institute of Technology
    6.4  Carnegie-Mellon University
    6.5  The Centre de Recherche Information de Montreal
    6.6  Columbia University
    6.7  Electrotechnical Laboratory
    6.8  ESPRIT
    6.9  University of Illinois
    6.10 Institute for New Generation Computer Technology
    6.11 Lawrence Livermore Laboratory
    6.12 Los Alamos National Laboratory
    6.13 University of Manchester
    6.14 Massachusetts Institute of Technology
    6.15 University of Michigan
    6.16 Microelectronics and Computer Technology Corp.
    6.17 University of Minnesota
    6.18 National Security Agency
    6.19 New York University
    6.20 University of North Carolina
    6.21 North Carolina State University
    6.22 Purdue University
    6.23 University of Reading
    6.24 Semiconductor Research Corporation
    6.25 Simon Fraser/Queen's University
    6.26 Stanford University
    6.27 System Development Foundation
    6.28 Technical University of Berlin
    6.29 University of Texas at Austin
    6.30 The University of Toronto
    6.31 University of Utah
    6.32 U. S. Department of Defense
    6.33 University of Washington
    6.34 University of Waterloo
    6.35 Yale University
7 - Companies to watch
    7.1  Alliant Computer Systems Corporation
    7.2  Automatix, Inc.
    7.3  Bolt Beranek and Newman, Inc.
    7.4  Control Data Corporation
    7.5  Cray Research, Inc.
    7.6  Denelcor, Inc.
    7.7  Digital Equipment Corporation
    7.8  Eastman Kodak
    7.9  ELXSI
    7.10 Encore Computer Corporation
    7.11 ETA Systems
    7.12 Floating Point Systems, Inc.
    7.13 General Electric
    7.14 Goodyear Aerospace
    7.15 IBM
    7.16 Intel Scientific Computers
    7.17 Kurzweil Applied Intelligence
    7.18 Schlumberger
    7.19 Thinking Machines Corporation
8 - Delphi survey
    8.1 General architecture
    8.2 Connections
    8.3 Memory
    8.4 Problem solving
    8.5 Programming languages
    8.6 Applications
    8.7 Processing limit for AI applications
    8.8 Comparison: United States, Japan, Europe
    8.9 Summary
9 - Market forecast
    9.1 Artificial intelligence potential market
    9.2 Scientific and engineering potential market
    9.3 Market forecast: United States
    9.4 Market forecast: Japan
    9.5 Market forecast: Europe
    9.6 International market leadership
    9.7 Summary
10 - For further reading
     10.1 Advanced generation computing periodicals
     10.2 Books
     10.3 Conference proceedings
     10.4 Technical papers and articles

(For information call (404) 342-9638 or write to:
     SEAI Technical Publications
     P.O. Box 590
     Madison, GA 30650)

------------------------------

End of PARSYM Digest
********************

∂03-Dec-85  2038	BRESNAN@SU-CSLI.ARPA 	interactions of morphology, syntax, and discourse    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 3 Dec 85  20:38:15 PST
Date: Tue 3 Dec 85 20:33:41-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: interactions of morphology, syntax, and discourse
To: folks@SU-CSLI.ARPA
26-Nov-85 22:20:05-PST,2835;000000000001
Mail-From: BRESNAN created at 26-Nov-85 22:19:29
Date: Tue 26 Nov 85 22:19:29-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>

←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
This Thursday, December 5, Draga Zec will present her work on
Obligatory Control in Clausal Complements in the CSLI conference
room at 10:00 a.m.  She derives the phenomenon of obligatory
anaphoric control from the theory of obviation together with the
semantics of control verbs.  Serbo-Croatian, from which much of
her evidence is drawn, beautifully illuminates the two separate
factors in obligatory control, which are obscured in English.
Everyone interested in the syntax and semantics of control
constructions and their implications for linguistic theory is
invited.  Written copies of the paper will be available on
Wednesday morning at the CSLI Receptionists' desk.

              INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                ``Obligatory Control in Clausal Complements''

	It is generally held that obligatory control correlates with
	the non-finiteness of the complement. Both syntactic and
	semantic theories of control have crucially depended on this
	particular assumption.  My intention is to present a case of
	obligatory control into clausal complements, develop an
	analysis within the LFG framework, and then explore the
	implications of this case for an adequate treatment of
	control.

	Serbo-Croatian has two types of clausal complements, Type 1
	which is generally uncontrolled, and Type 2 which allows
	obligatory control with predicates like 'try', 'intend',
	'persuade', 'force', etc. It will be shown that Type 2
	complements cannot be dealt with in terms of the LFG theory
	of functional control, or any other syntactic theory of
	control. Rather, it will be argued that these complements
	are a clear case of what in LFG is referred to as anaphoric
	control. Certain differences in anaphoric binding properties
	between the two complement types are attributed to the
	phenomenon of obviation which is found with Type 2 but not
	with Type 1 complements.

	Since anaphoric control cannot capture the systematic
	controller/controllee relation characteristic of obligatory
	control, one will have to make reference to the semantic or
	more precisely, thematic properties of control-inducing
	predicates. This may have implications for syntactic
	theories of obligatory control, whose aim is to make
	predictions about controller/controllee relations solely in
	syntactic terms. This case will also be relevant for the
	semantic analyses that account for control solely in terms
	of entailment.  --Draga Zec
	-------
-------
 
-------

∂04-Dec-85  0735	PATASHNIK@SU-SUSHI.ARPA 	next AFLBs
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  07:35:04 PST
Date: Wed 4 Dec 85 06:47:45-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: next AFLBs
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12164429642.9.PATASHNIK@SU-SUSHI.ARPA>

These are the last two AFLB talks this quarter:
		***********************************

5-Dec-85  :  John Cannon (University of Sydney)

	Finding the order of a permutation group of degree 10,000

An important goal in computational group theory is the development of
fast algorithms for determining the abstract structure (composition
factors) of a finite group.  This goal is now within our reach in the
case of permutation groups having degree less than 10,000.  Let G be a
permutation group acting on the finite set Omega.  A base for G is a
sequence B = {B←1, ..., B←k} of points from Omega, such that the identity
is the only element of G that fixes B pointwise.  Let G↑(i) denote the
stabilizer G←{B←1 ... B←i}.  A base for G defines the chain
		G = G↑(0) > G↑(1) > ... > G↑(k) = <1>.
A strong generating set S for G, relative to the base B, is a set of
elements which contains generators for each subgroup in the above
chain.  The construction of a base and strong generating set (BSCS) is
the first step in any investigation of the structure of G.  In
particular, knowledge of a BSCS enables us to read off the order of G.

In this talk I shall review the Sims-Schreier algorithm for computing
a BSCS.  I shall then describe the more powerful Todd-Coxeter-Schreier
algorithm.  Recent studies at Sydney have demonstrated that the latter
algorithm is very unstable.  As a consequence, Sims and I have
undertaken the development of a new algorithm which is based on ideas
used in the construction of very large sporadic simple groups.  The
new algorithm will be described and I shall present some information
about the performance of a first implementation of this algorithm.

***** Time and place: December 5, 12:30 pm in MJ352 (Bldg. 460) ******

12-Dec-85  :  Joan Hutchinson (MSRI)

	Recent developements in topological graph theory

(Approximate title; official title and abstract, forthcoming)

***** Time and place: December 12, 12:30 pm in MJ352 (Bldg. 460) ******

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.  If you
have a topic you'd like to talk about please let me know.  (Electronic
mail: patashnik@su-sushi.arpa, phone: (415) 497-1787). Contributions
are wanted and welcome.  Not all time slots for this academic year
have been filled.  The file [SUSHI]<patashnik.aflb>aflb.bboard contains
more information about future and past AFLB meetings and topics.
	--Oren Patashnik
-------

∂04-Dec-85  0919	ullman@su-aimvax.arpa 	meeting
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  09:19:30 PST
Received: by su-aimvax.arpa with Sendmail; Wed, 4 Dec 85 09:17:11 pst
Date: Wed, 4 Dec 85 09:17:11 pst
From: Jeff Ullman <ullman@diablo>
Subject: meeting
To: nail@diablo

Today, i think I'll cover the (small) ideas about P, NC, and
polynomial fringes that I didn't get to last week.
Meet at 11AM?

∂04-Dec-85  0949	@SU-SCORE.ARPA:MCCALL@SU-SUSHI.ARPA 	Announcing: CS523 - Survey of Artificial Intelligence
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  09:48:44 PST
Received: from SU-SUSHI.ARPA by SU-SCORE.ARPA with TCP; Wed 4 Dec 85 09:46:08-PST
Date: Wed 4 Dec 85 09:45:51-PST
From: Kim McCall <MCCALL@SU-SUSHI.ARPA>
Subject: Announcing: CS523 - Survey of Artificial Intelligence
To: csd@SU-SCORE.ARPA
Message-ID: <12164462065.20.MCCALL@SU-SUSHI.ARPA>


                    !!! ANNOUNCING !!!
   CS523 - An Intensive Survey of Artificial Intelligence
     (acutal title: Readings in Artificial Intelligence)

Under the sponsorship of Prof. Bruce Buchannan, we once again
present the ever popular "AI Qual Prep Seminar".   In this class
we will read and discuss sections of several of the main AI
reference texts as well as a fair proportion of the seminal and
most influential papers in the field.

The primary purpose of the course is to prepare PhD candidates
for the rigors of the AI qual, but other hardy souls are also
invited to join.

The anticipated format for the course includes two one-hour
discussion sessions and one one-hour invited lecture per week.
The primary formal responsibility of class members will be to
take turns leading or recording and publishing our discussions.

The formal workload is, thus, fairly low, but the reading load
is very high.

We will benefit greatly from the efforts of last year's seminar
leader, Devika Subramanian, who has done yeowoman's service in
preparing both an excellent annotated reading list for the course
and a well-edited compilation of notes on last year's discussions.

Due to the length of the quarter, the study of AI will be seen to
fall neatly into 10 distinct areas.  We expect to cover the following
topics:
   Week  1: Introduction to AI
   Week  2: Search and Heuristics
   Week  3: Knowledge Representation
   Week  4: Planning, Problem Solving, and Automatic Programming
   Week  5: Deduction, Inference, and Theorem Proving
   Week  6: Expert Systems
   Week  7: Learning
   Week  8: Natural Language Understanding
   Week  9: Vision and Robotics
   Week 10: Advanced Topics
              including: advanced reasoning techniques
                         software architectures
                         hardware architectures
                         philosophical issues
                         future directions

The main problem now is: When shall we meet?
Three ways to decide:
  (a) Fiat
  (b) Take a survey now and try to decide before next quarter
  (c) Take a survey now of when we could get together at the
        beginning of next quarter to decide the time.

Choice (a) is terrible.  If you are interested in this course,
please:
Reply to this message (or some other how get a message to this
year's TA, Kim McCall <McCall@Sushi>) to say:
  (1) how great the chances are that you will want to take the
        course
  (2) whether or not you ever plan to take the AI qual
  (3) whether you know when you will be free for the class next
        quarter
    (3a) if so, when
  (4) when you could make a planning meeting on Monday January 6
        or Tuesday January 7

Thank you.

Any Q's?

Send in those cards before it's too late.

Kim
-------

∂04-Dec-85  1037	DALRYMPLE@SU-CSLI.ARPA 	fun on Friday   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  10:36:28 PST
Date: Wed 4 Dec 85 10:26:52-PST
From: Mary Dalrymple <DALRYMPLE@SU-CSLI.ARPA>
Subject: fun on Friday
To: folks@SU-CSLI.ARPA, linguists@SU-CSLI.ARPA

Don't miss the last Linguistics Soiree of the semester, this Friday
at 4:45 in the Greenberg Room.
-------

∂04-Dec-85  1320	MODICA@SU-SCORE.ARPA 	Grade Sheets 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  13:20:30 PST
Date: Wed 4 Dec 85 13:16:38-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: Grade Sheets
To: instructors@SU-SCORE.ARPA
cc: su-bboard@SU-SCORE.ARPA
Message-ID: <12164500436.11.MODICA@SU-SCORE.ARPA>

This is reminder that ALL grade sheets (also known as End Quarter Reports)
should be brought to me (Gina, MJH 0303), NOT VICTORIA.
I know you've all grown very attached to Victoria, and that her office
is very conveniently located, but she is no longer the person handling
grades, so please don't take them to her any more.
Thank You.
-Gina
-------

∂04-Dec-85  1341	CAROL@SU-CSLI.ARPA 	Parking man    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  13:37:41 PST
Date: Wed 4 Dec 85 13:32:41-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: Parking man
To: folks@SU-CSLI.ARPA

is out there giving tix
-------

∂04-Dec-85  1702	EMMA@SU-CSLI.ARPA 	Newsletter December 5, No. 4   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  17:02:28 PST
Date: Wed 4 Dec 85 16:30:33-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter December 5, No. 4
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
December 5, 1985                Stanford                        Vol. 3, No. 4
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, December 5, 1985

   12 noon		TINLunch
     Ventura Hall       A Humanistic Rationale for Technical Writing
     Conference Room    by Carolyn R. Miller
			Discussion led by Geoff Nunberg (Nunberg@csli)
			(Abstract on page 2)

   3:30 p.m.		Tea
     Ventura Hall		

                              ←←←←←←←←←←←←
         CSLI ACTIVITIES FOR *THIS* THURSDAY, December 12, 1985

   12 noon		TINLunch
     Ventura Hall       Title to be announced
     Conference Room    by Werner Frey and Hans Kamp
			Discussion led by Werner Frey
			(Abstract will appear next week)
			
   2:15 p.m.		CSLI Seminar
			No seminar this week

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
			Title to be announced
			Lynn Bloom
                              ←←←←←←←←←←←←
                              ANNOUNCEMENT

   Please note that the seminar and colloquium are no longer in Redwood
   Hall room G-19.  The new room will be announced in next week's
   newsletter. 

!
Page 2                     CSLI Newsletter                   December 5, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                          THIS WEEK'S TINLUNCH
              A Humanistic Rationale for Technical Writing
                          by Carolyn R. Miller
                     Discussion led by Geoff Nunberg

      This paper is typical of a number of recent articles by
   sociologists, rhetoricians, and humanistically-trained writing
   specialists, that insist that scientific writing is no less
   rhetorical in its means and effects than is writing of an explicitly
   belletristic sort. Whether or not we find their arguments compelling,
   these articles raise interesting questions for producers and consumers
   of technical prose, especially in intellectually self-conscious
   disciplines like philosophy, AI, and linguistics. For example: What is
   the common understanding of the research enterprise that underlies the
   linguistic conventions characteristic of scientific prose, such as the
   avoidance of ``I'' and the unusual uses of ``we'', the frequent use of
   impersonal constructions, the numbering of paragraphs, and so on? Can
   we apply the apparatus of traditional rhetoric to the evaluation of
   the expository usefulness of particular formal languages and
   notational conventions? Is there grounds for distinguishing between a
   ``rhetoric of information'', concerned with the selection and
   arrangement of factual observations, and a ``rhetoric of description,''
   concerned with the linguistic means used to report such observations?
                               ----------
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                Obligatory Control in Clausal Complements
                          Draga Zec (Zec@csli)
          Thursday, December 5, 10:00, Ventura Conference Room

      It is generally held that obligatory control correlates with the
   non-finiteness of the complement. Both syntactic and semantic theories
   of control have crucially depended on this particular assumption.  My
   intention is to present a case of obligatory control into clausal
   complements, develop an analysis within the LFG framework, and then
   explore the implications of this case for an adequate treatment of control.
      Serbo-Croatian has two types of clausal complements, Type 1 which
   is generally uncontrolled, and Type 2 which allows obligatory control
   with predicates like `try', `intend', `persuade', `force', etc. It
   will be shown that Type 2 complements cannot be dealt with in terms of
   the LFG theory of functional control, or any other syntactic theory of
   control. Rather, it will be argued that these complements are a clear
   case of what in LFG is referred to as anaphoric control. Certain
   differences in anaphoric binding properties between the two complement
   types are attributed to the phenomenon of obviation which is found
   with Type 2 but not with Type 1 complements.
      Since anaphoric control cannot capture the systematic
   controller/controllee relation characteristic of obligatory control,
   one will have to make reference to the semantic or, more precisely,
   thematic properties of control-inducing predicates. This may have
   implications for syntactic theories of obligatory control, whose aim
   is to make predictions about controller/controllee relations solely in
   syntactic terms. This case will also be relevant for the semantic
   analyses that account for control solely in terms of entailment.
						--Draga Zec
   Everyone interested in the syntax and semantics of control constructions 
   and their implications for linguistic theory is invited.  Written
   copies of the paper are available at the CSLI Receptionist's desk.
!
Page 3                     CSLI Newsletter                  December 5, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                          PIXELS AND PREDICATES
                  Spatial Parsing for Visual Languages
                         Fred Lakin (Lakin@csli)
             1:00 pm, Wednesday, December 11, CSLI trailers

      Graphics are very important in human/computer interaction.  To
   participate effectively in this kind of interaction, computers must be
   able to understand how humans use graphics to communicate.  When a
   person employs a text and graphic object in communication, that object
   has meaning under a system of interpretation, or ``visual language''.
   A first step toward computer understanding of the visual communication
   objects used by humans is computer parsing of such objects, recovering
   their underlying syntactic structure.  The research described in this
   paper combines computer graphics, symbolic computation and textual
   linguistics to accomplish ``spatial parsing'' for visual languages.
      Parsing has been investigated in four visual languages: VennLISP (a
   visual programming language based on LISP), VIC (a visual
   communication system for aphasics), FSA (finite state automaton)
   diagrams, and SIBTRAN (graphic devices for organizing textual sentence
   fragments). A parser has been written which can recover the structure
   for graphic communication objects in the different visual languages.
   In addition, interactive visual communication assistants utilize
   modules from the parser to actively assist humans using two of the
   visual languages.
                               ----------
                              LOGIC SEMINAR
         Proving Properties of Destructive LISP Programs (cont.)
                  Ian Mason, Philosophy Dept., Stanford
    Friday, December 6, 12:00-1:00, Math. Faculty Lounge, Room 383-N
                               ----------
                       ENVIRONMENTS GROUP MEETING
                       Grammar-writer's Workbench
              Ronald Kaplan, Xerox and CSLI (Kaplan@xerox)
             Monday, December 9, noon, Ventura Seminar Room
                               ----------
                       PHONOLOGY/PHONETICS MEETING
           Adequacy in Intonation Analysis: The Case of Dutch
                           Carlos Gussenhoven
          Wednesday, December 11, 3:30, Ventura Conference Room
!
Page 4                     CSLI Newsletter                  December 5, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                              AFT MEETINGS

      In the winter term, I shall start the regular meetings of the AFT
   (Aitiational Frame Theory) Project on lexical representation.  The
   meetings will be once a week on Tuesdays at 11, and the first one will
   be on January 14. The room will be announced later.
      The project will be concerned with the construction of an adequate
   theory of lexical representation and lexical meaning.  While my
   interests center around the AFT proposal, the main aim will be to
   compare available theories of lexical meaning and come up with what
   will seem to the group to be the best one.
      In addition to the weekly meetings, there will be guest
   presentations by speakers such as Joseph Almog (UCLA), Nathan Salmon
   (UCSB), and Scott Soames (Princeton).

   The following is a partial list of topics to be discussed.

   a. Lexical meaning and the needed input for compositional semantics.
   b. Lexical meaning and the needed input for syntax.
   c. Lexical meaning and non-monotonic reasoning.
   d. AFT and doubts about the claim that natural languages are formal
      languages.
   e. Lexical meaning and ``psychological reality.''
   f. Lexical meaning, AFT, and morphology.

   I would appreciate it if those who are interested in joining this
   group would contact me via computer mail (Julius@csli) or phone
   ((415)497-2130) during the next couple of weeks.  --Julius Moravcsik

-------

∂04-Dec-85  2141	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Call for papers - Logic and Applications 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 4 Dec 85  21:41:11 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 4 Dec 85 21:26:33-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Wed 4 Dec 85 21:23:34-PST
Received: by rsch.wisc.edu; Wed, 4 Dec 85 23:11:52 CST
Date: Wed, 4 Dec 85 12:20:34 CST
From: udi@rsch.wisc.edu (Udi Manber)
Message-Id: <8512041820.AA06739@rsch.wisc.edu>
Received: by rsch.wisc.edu; Wed, 4 Dec 85 12:20:34 CST
To: udi@rsch.wisc.edu
Subject: Call for papers - Logic and Applications
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 04 Dec 85 23:11:15 CST (Wed)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

Sent by request of Ljubomir Ivanov


                         CALL FOR PAPERS


              INTERNATIONAL SUMMER SCHOOL AND CONFERENCE

                              on

               MATHEMATICAL LOGIC AND ITS APPLICATIONS

            Dedicated to the 80th Anniversary of Kurt Godel
                   September 24 - October 3, 1986
                         Druzhba, Bulgaria



The meeting is organized by the Bulgarian Academy of Sciences, Sofia
University and the International Foundation `Ljudmila Zhivkova'.

The topics of the school and the conference are: mathematical logic and
the foundations of mathematics (set-theoretical mathematics,
constructive mathematics, intuitionism, proof theory); mathematical
logic and computer science (abstract recursion theory, theory of
computability, semantics of programming languages, logical programming,
logics of programs, logical problems of information retrieval);
mathematical logic, philosophy and the study of language (model and
other intensional logics, non-monotonic logic and other logical issues
in artificial intelligence).

The list of invited lecturers includes: J.W. deBakker, J.F.A.K. van
Benthem, A.G. Dragalin, Yu.L. Ershov, S.S. Goncharov, H. Jervell, Y.N.
Moschovakis, N.M. Nagornij, V.A. Nepomnjashchij, H. Rasiowa, G. Sambin,
K. Segerberg, N.A. Shanin, A. Skowron, V. Stoltenberg-Hansen, B.A.
Trakhtenbrot, V.A. Uspenskij.

Authors are invited to send three copies of a draft full paper by March
31, 1986 to the program committee chairman:

		Dimitar G. Skordev
		Section of Logic
		Faculty of Mathematics
		Sofia University
		bul. Anton Ivanov 5
		1126 Sofia, Bulgaria
		(003592) 62561 (583)

The paper should be at most fifteen double-spaced typed pages, and
should present original contributions in the abovementioned areas. The
papers will be considered by the program committee (now under formation)
and authors will be notified of the acceptance or rejection by June 30,
1986. Accepted papers in final form are due to August 15, 1986 for
inclusion in the conference proceedings to be published by Plenum Press.

--------------
TN Message #12
--------------

∂05-Dec-85  0733	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #43  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  07:30:25 PST
Date:  5 Dec 85 0719-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #43
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Thursday, 5 Dec 1985      Volume 1 : Issue 43

Today's Topic:

           Seminar: Limits on the Power of Concurrent-Write
                    Parallel Machines (MIT)

----------------------------------------------------------------------

Date: 12/04/85 18:17:14
From: LISA at MIT-MC.ARPA
Re:   Beame's talk

[Forwarded from the MIT-MC BBoard by SASW@MIT-MC]

                                    SEMINAR

DATE:  Thursday, December 5, 1985
TIME:  3:45 p.m......Refreshments
       4:00 p.m......Lecture
PLACE: MIT NE43 - 453

     "LIMITS ON THE POWER OF CONCURRENT-WRITE PARALLEL MACHINES"


                              PAUL BEAME
                        University of Toronto

                              Abstract:

Concurrent-read concurrent-write parallel RAM's (CRCW PRAM'S) are very
powerful machines.  They can compute any boolean function in constant
time given sufficiently many processors and sufficiently large memory.
Also they can compute simple functions like the boolean OR and the sum
of two n-bit numbers in constant time using a polynomial number of
processors and cells.

We consider powerful generalizationns of CRCW PRAM's which are
non-uniform and allow the individual processors an unlimited
instruction set.  We show that for such machines if the number of
processors and cells is bounded by a polynomial in n then:

     Computing the sum of  n  integers of  n  bits each requires
     time  THETA (log↑n).

     Computing the parity of, sorting, or adding  n  input bits
     require time  OMEGA (sqrt  {↑log↑n}).

The latter result uses a recent result of Hastad concerning unbounded
fan-in circuits.

HOST:  David Shmoys


------------------------------

End of PARSYM Digest
********************

∂05-Dec-85  0958	REULING@SU-SCORE.ARPA 	CSD Mailing Lists
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  09:58:19 PST
Date: Thu 5 Dec 85 09:54:26-PST
From: John Reuling <Reuling@SU-SCORE.ARPA>
Subject: CSD Mailing Lists
To: faculty@SU-SCORE.ARPA, staff@SU-SCORE.ARPA
cc: Mailing-Lists@SU-SCORE.ARPA
Stanford: 206 Margaret Jacks Hall, +1 (415) 497-3549
Message-ID: <12164725771.23.REULING@SU-SCORE.ARPA>

[Summary: Send mailing-list errors to MAILING-LISTS@Score]


Please help us keep CSD's electronic mailing lists up to date.

If you send mail to a mailing list at Score and receive back an
error message from Mailer, please REMAIL or FORWARD the entire
error message to MAILING-LISTS@Score.  

Here are a few of the larger mailing lists maintained on Score:

        AC (Academic-Council)   PhD-Adm
        Admin                   Post-Docs
        Colloq                  Programmers
        CSD                     RAs
        Faculty                 Sec (Secretaries)
        Instructors             SrStaff
        Masters                 Staff
        MSAI                    TAs
        MSAI-Adm                Trainers
        PhD                     Visitors

Thank you.

-John Reuling
-------

∂05-Dec-85  1223	EMMA@SU-CSLI.ARPA 	Newsletter correction
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  12:23:01 PST
Date: Thu 5 Dec 85 11:40:27-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter correction
To: friends@SU-CSLI.ARPA
Tel:  497-3479


   The name of the speaker for the December 12 colloquium is Bjorn
Lindblom not Lynn Bloom as stated in the newsletter.  The abstract for
the colloquium follows.

                           CSLI COLLOQUIUM
                  Thursday, December 12, 1985, 4:15
            Turing Auditorium (not Redwood Hall room G-19)

            THEMES IN THE EVOLUTIONARY BIOLOGY OF LANGUAGE
                         A three-ring circus

      Bjorn Lindblom, Peter MacNeilage, Michael Studdert-Kennedy
                           CASBS, Stanford

The goal of our research is summarized by the phrase: DERIVE LANGUAGE
FROM NON-LANGUAGE! We are exploring an approach to the biology of
language that is deliberately distinct from that pursued within
Chomskyan autonomous linguistics. We take as our first priority an
explicit search for precursors to all aspects of language structure
and speech behavior. By precursors we mean either evolutionary
precursors, traceable to lower animals, or those human but
non-linguistic, cognitive, perceptual and motor capacities that both
constrain language and make it possible. Only when a search for such
precursors has failed can we justly term some characteristic
unique---either to language or to man---and attribute it to some
species-specific bioprogram for language learning and use (cf.
universal grammar). In short, while we acknowledge and respect the
discoveries of formal linguistics, we believe that a sound approach to
the biology of language must go beyond form and structure to ask:
``How did language get that way?''

A major language universal for which any phylogenetic or ontogenetic
theory must account is LA DOUBLE ARTICULATION, or DUALITY of
patterning.  We view the emergence of duality---that is, the use of
discrete elements and combinatorial rules at the two levels of
phonology and syntax---as the key to the communicative power of
language: duality provides a kind of impedance match between the
essentially unlimited semantics of cognition and a decidedly limited
set of signaling devices and processes.

Our central concern is with phonology: with the origin of discrete
phonological elements---phonemes and features---and with the processes
by which these intrinsically amodal elements are instantiated in the
modalities of production and perception. We shall review typological
facts about sound structure leading us to conclude that phonological
form adapts to the performance constraints of language use. How do we
choose our theoretical formalism for describing sound patterns?
Markedness theory or contemporary developments of generative phonology
and formal linguistics? No, since (i) spoken language is a product of
biological and cultural evolution; and (ii) there is considerable
empirical justification for viewing phonologies as adaptations to
biological and social selectional ⊂pressures the correct choice
appears to be some variant of the theoretical framework currently
explored by many students of biological and cultural evolution, viz.,
a Darwinian VARIATION*SELECTION model. In our talk we will present a
computational implementation of such a model. We will illustrate some
of its explanatory power by means of simulations indicating how a
number of typological facts can be predicted quantitatively and how
the emergence of ``a featural and segmental'' organization of lexical
systems can be derived in a self-organizing manner and deductively
(rather than just axiomatically). Drawing on corpora of speech error
data we describe the process by which discrete elements are organized
and sequenced in an actual utterance (phonologically and
syntactically) as one of inserting elements into a structured frame,
and, in our talk, we will consider the evolutionary relation between
this FRAME-CONTENT mode of organization and bimanual coordination.
Finally we will consider behavioral and neurophysiological evidence,
from both adults and children, consistent with a view of the
phonological element as an AMODAL structure linking production and
perception.

*Turing Auditorium is near Ventura Hall.
-------

∂05-Dec-85  1339	@SU-SCORE.ARPA:LES@SU-AI.ARPA 	Facilities action items 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  13:39:19 PST
Received: from SU-AI.ARPA by SU-SCORE.ARPA with TCP; Thu 5 Dec 85 13:12:49-PST
Date: 05 Dec 85  1315 PST
From: Les Earnest <LES@SU-AI.ARPA>
Subject: Facilities action items 
To:   faculty@SU-SCORE.ARPA 

Here is background on two items that will be taken up at the Faculty
meeting on Tuesday, December 10.

			APS Typesetter Aquisition

The Facilities Committee recommends that the department purchase the APS
typsetter from the TeX Project at depreciated cost (about $30,000) and
operate it as a cost center.  The TeX Project has been operating the
typesetter as a facility for its own use but has made it available to
others in the department and elsewhere for books and other publications
(e.g. Star Labs reports, A.I. Magazine, Pat Winston's book).  Outside
revenues were about $8,000 in 1984 and would have been much higher if Don
Knuth's typsetting had been charged for.

Two independent cost analyses by committee members indicate that there
would be little financial risk in taking on this responsibility.  The
committee approved the proposal to acquire it with the understanding that
it will be operated as a low-overhead facility with minimal hand-holding
and little marketing.  Faculty comments on this plan are invited.

			Courtesy Computer Accounts

A report on problems with computer courtesy accounts is attached.  The
Facilities Committee recommends that CSD adopt the policy shown at the
end, which is designed to deal with these problems.

	Les Earnest
	Facilities Committee Chair


-----------------------------------------------------------------------------

			Courtesy Computer Accounts

				Background

During the 22 years in which CSD has had its own computer facilities,
courtesy accounts have been extended to many outside users, often with
little serious justification.  In fact, limited public access to SAIL was
permitted for people without *any* account up to the early '70s.  While
the use of computer resources by guests sometimes inconvenienced other
users a bit, this permissiveness resulted in no direct economic losses to
the research programs or the department until recently.  Given that some
of the department's computer facilities now run as cost centers, courtesy
accounts on these machines now consume some of our limited and precious
unrestricted funds.

The use of some courtesy accounts has persisted long past anyone's memory
of why they were created.  There is also evidence of substantial waste and
abuse in the use of some accounts.  It seems desirable to continue to
offer courtesy accounts, but we should ensure that the financial burden
does not become excessive and that the resources devoted to this activity
benefit the department.  Toward this end, we need a clear policy and
administrative enforcement.

			Current Use and Abuse

A scan of recent courtesy account billings reveals the following
distribution of users:
     #    Category
    44  Former PhD students
     8  Former CSD staff
     1  Former CSD faculty
     9  Friends of the Department
    25  Unknown
    --
    93    Total
The "Friends of the Department" are faculty or researchers elsewhere
with close ties to the department.

Review of recent billings (May through August 1985) reveals that the
department has been paying about $18,000 a year for these 93 accounts.
These charges have been much larger than necessary for five main reasons:
  (1) inactive accounts have not been purged and accrue substantial disk
      storage charges;
  (2) some hyperactive accounts have incurred inappropriately large charges;
  (3) some courtesy users have been running on SCORE, which is a cost center,
      when they could have been on SUSHI, which is not;
  (4) once someone gets a courtesy account, there has been no systematic
      record-keeping of why the account was set up nor any review of the
      appropriateness of continuing it,
  (5) the existing accounting system does not tell how much is being spent
      on courtesy accounts.
The staff has done its best to cope with these problems but has been
hampered by a lack of clear policies.  Working on point (3), for example,
the courtesy accounts of outside users on SCORE were shifted to SUSHI late
last summer and the accounts of Masters students will be shifted to SUSHI
soon.  Though a partial purge was conducted this summer, there still
appear to be a number of people who have these accounts for no good reason.

37 courtesy accounts with no logins in 1985 have been incurring charges at
a rate over $5,000 per year.  Some of these accounts have not been used
since 1980.  Nevertheless, some are on distribution lists that cause their
mail files to keep growing.
    Ignacio Zabala's account is a rather horrendous example.  He received
    his PhD three years ago, returned to Spain and has not logged in since.
    In this time his mail file grew to the point where he alone has been
    accounting for over $2000 a year in disk charges.  (This account was
    purged in early November at my request.)
It seems sensible to purge accounts that have been inactive for a long
time -- say, one year.

Of the courtesy accounts that are active, some are incurring very large
charges -- one person is currently running at a rate over $2000 per year,
another is just over $1000 and another 5 are each costing the department
over $500 a year.  While Score software enforces account limitations,
SAIL does not.  If we enforced a limit of, say, $120 per year per courtesy
account, we would save about $9,000 annually.

One of the reasons that the problems cited above were not noticed was that
the accounting system does not yield cost information on courtesy accounts
as a separate category -- these charges are intermixed with those of all
other users who are charged to the CSD Gift Account.  In order to get the
data above it was necessary to extract records from detailed account
listings based on a line-by-line search, then total them separately.

Even if we set up a review procedure for courtesy accounts, there will
likely still be some abuses; e.g. some accounts may be "inherited" by
unauthorized users.  We do not propose to expend effort on preventing this
kind of abuse, but it should be dealt with if evidence appears
fortuitously.

When a sponsored research project participant decides that a courtesy
account is needed for someone, he should normally arrange for the account
on the machine used by the project.  If that machine is a cost center,
then the project should be expected to cover the cost of the account.  In
any case, the project administration should keep track of the courtesy
account and ensure that it is closed when no longer appropriate.

It appears, then, that by adopting a few simple measures, the cost of
courtesy accounts can be reduced from $18,000 per year to $4,000 or less
while retaining the principal benefits of the existing program.

			Recommendations

1. A policy such as the one given below should be adopted by the Department.
After adoption, all existing courtesy accounts should be reviewed for
consistency with the policy and exceptions should be dropped, with adequate
notice to the users.

2. The cost center accounting system should be modified to give separate
listings and totals on "subaccounts" -- subclasses of users who are
charged to the same University account.  This will facilitate periodic
review of the costs of courtesy accounts and other user subgroups.

-----------------------------------------------------------------------

		Policy on Courtesy Computer Accounts

PURPOSES.  It is the policy of CSD to provide limited access to
departmental computer resources at no cost to certain people with close
ties to the department.  The purpose of this access is to facilitate
communication with department members and to provide access to information
resources at Stanford.  The departmental purpose is to stimulate exchanges
of information with outside members of computer research and teaching
communities and to engender goodwill.

INITIATION.  Departing CSD students, faculty, and staff members will
normally be granted a grace period of up to 90 days to move or purge their
computer files if they request it in advance.  Otherwise, files will be
purged on termination.

Individuals may be granted ongoing courtesy accounts upon approval by a
principal investigator or other person with account commitment authority.
Any cost of such courtesy accounts shall be charged to the University
accounts of the approving authority.  Courtesy accounts on Department-
supported computers (e.g. SUSHI) or that are to be charged against
Departmental unrestricted funds shall be approved by the Department
Chairman.  Individual charges against Departmental unrestricted accounts
shall not be permitted to exceed $10 per month.

Courtesy account users shall be informed that the account is for their
personal use and should not be made available to others.  They shall
also be informed of any cost or other constraints on the use of the
account.

RECORDS AND REVIEW.  For each courtesy account, whether it involves direct
charges or not, a record shall be kept of the reason the account was set
up, who approved it, and the termination date or conditions, if any.
Principal Investigators and the Department Chairman shall establish
procedures for periodically reviewing courtesy accounts to ensure that the
cost is consistent with their purposes and that they are terminated when
appropriate.  Where possible, charge or resource constraints shall be
enforced automatically.  Otherwise, accounts shall be reviewed by someone
monthly and appropriate action shall be taken to curb any abuses.

Accounts that are supported with Departmental funds that have not been
used for a year shall be purged whether or not there are direct charges
involved.

∂05-Dec-85  1402	TAJNAI@SU-SCORE.ARPA 	nominations for IBM fellowship   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  14:02:10 PST
Date: Thu 5 Dec 85 13:58:03-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: nominations for IBM fellowship
To: faculty@SU-SCORE.ARPA
Message-ID: <12164770121.30.TAJNAI@SU-SCORE.ARPA>


Nominations are now open for the IBM fellowship.  I am only
accepting nominations for CSD students (with the exception of
Ted Shortliffe's student); Sharon Gerlach will handle the
CSL/EE nominations.

Students must have passed their qualifying examination.
IBM does not disqualify non-citizens.

Nominations will close 5 p.m., Friday December 13.

Carolyn 
-------

∂05-Dec-85  1514	NILSSON@SU-SCORE.ARPA 	Awards 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  15:13:56 PST
Date: Thu 5 Dec 85 15:06:58-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Awards
To: csd@SU-SCORE.ARPA
Message-ID: <12164782666.16.NILSSON@SU-SCORE.ARPA>

I recently received a letter regarding awards from Stanley Winker,
Chairman AFIPS awards committee.  Attached are the main paragraphs from
that letter.  Anyone interested in further info about this should stop
by my office and ask Anne Richardson for the referenced "enclosed"
information. -Nils

----

"As you may know, the deadline for nominations for the annual awards
of the American Federation of Information Processing Societies is
quickly approaching.  In addition to AFIPS annual Education Award
and Harry Goode Memorial Award, this year AFIPS is pleased to announce
a new award, the Student History Best Paper Prize, which will honor
superior writing achievements of graduate and undergraduate students
in the area of computer history.

As Chairman of the Awards Committee I am writing to remind you that we
are accepting nominations through December for the Harry Goode and 
Education Awards and through March 1st for the Student History Best
Paper Prize.  I am certain there are members of your department
who will want to make nominations for these AFIPS awards, as well as
students who will wish to submit papers for the new Student History
Best Paper Prize.  Enclosed is further information on the awards
and several nomination forms which you are encouraged to distribute
to all interested parties in your organization."

-----
-------

∂05-Dec-85  1718	ACUFF@SUMEX-AIM.ARPA 	I'll be gone for 6 days
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  17:18:35 PST
Date: Thu 5 Dec 85 17:16:27-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: I'll be gone for 6 days
To: ksl-lispm@SUMEX-AIM.ARPA
cc: ryalls@SUMEX-AIM.ARPA, sbarnes@SUMEX-AIM.ARPA, forbes@SUMEX-AIM.ARPA,
    VELAZQUEZ@SUMEX-AIM.ARPA
Message-ID: <12164806238.48.ACUFF@SUMEX-AIM.ARPA>

   I'll be away at a Common Lisp meeting until next Thursday.
Please refer emergencies having to do with Symbolics 36xx's or TI
Explorers to James Rice or an SSRG staff member, and send me mail
about others.

	Thanks,
	-- Rich
-------

∂05-Dec-85  2102	BRESNAN@SU-CSLI.ARPA 	morphology/syntax/discourse interactions   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  21:02:11 PST
Date: Thu 5 Dec 85 20:57:19-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: morphology/syntax/discourse interactions
To: folks@SU-CSLI.ARPA

On Thursday, December 12, at 10:00 a.m. we will meet in the CSLI
Conference Room to hear the following exciting presentation.  All
those interested are welcome.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Reflexivization Variation: Relations between Syntax, Semantics, and Lexical
Structure

Peter Sells, Annie Zaenen, Draga Zec

In this paper we examine the distinction between so-called transitive and
intransitive reflexive construction in several languages (English, Finnish,
German, Dutch, Chichewa, Warlpiri, Serbo-Croatian and Japanese). We argue
that three types of distinctions have to be made: transitivity versus
intransitivity in the lexicon, synthetic versus analytic forms in the
constituent structure and open versus closed predicates in the semantics;
thus there are three relevant levels of possible cross-linguistic
variation.  While there is a one-way implication between lexical
intransitivity and closed predication, there are in general no direct
correlations between either the lexical forms or the semantic forms and
their constituent structure representation.

We give an account of the different types of reflexive that we discuss in
Lexical-Functional Grammar augmented with Discourse Representation
Structures.

Copies of the paper will be available at the front desk on Monday December
9th.
-------
-------

∂05-Dec-85  2158	JF@SU-SUSHI.ARPA 	help   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  21:58:19 PST
Date: Thu 5 Dec 85 21:56:13-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: help
To: aflb.su@SU-SUSHI.ARPA
Message-ID: <12164857168.34.JF@SU-SUSHI.ARPA>

I need to send a message to theorynet.  I tried my best guess based on
manber's last message telling us how to use the list, but I failed.
How are we supposed to use the theorynet list from stanford machines?
Thanks,
Joan
-------

∂05-Dec-85  2247	CAROL@SU-CSLI.ARPA 	New office
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 5 Dec 85  22:47:18 PST
Date: Thu 5 Dec 85 22:43:49-PST
From: Carol Kiparsky <CAROL@SU-CSLI.ARPA>
Subject: New office
To: Folks@SU-CSLI.ARPA
cc: Carol@SU-CSLI.ARPA




*********************************************************************
**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DL
						
				Dec 5

Your itinerant dandelion consultant has yet another new office.
I'm now in the trailers, Room H-3.  My phone numbers here are
723-5565 and 497-9030.  The latter also rings in Marjorie's 
office in case you need to leave a message.
 
Feel free to call or drop in. I'm usually here between 9:30 and 
3:30.  
 

                             			-Carol     

ION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION**NEWS**DLION*
*********************************************************************
-------

∂06-Dec-85  1003	CONTRERAS@SU-SCORE.ARPA 	Package   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  10:03:47 PST
Date: Fri 6 Dec 85 09:48:01-PST
From: Tina Contreras <CONTRERAS@SU-SCORE.ARPA>
Subject: Package
To: CSD@SU-SCORE.ARPA
Message-ID: <12164986745.20.CONTRERAS@SU-SCORE.ARPA>



There is a package here that someone tried to send to The Recruiting Office
of The Athletic Department, please come and claim this package at the 
Reception desk.  I.D. Mail will not take it.



Tina
-------

∂06-Dec-85  1300	YAMARONE@SU-CSLI.ARPA 	HOLIDAY PARTY NEXT FRIDAY(THE 13TH!) 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  13:00:20 PST
Date: Fri 6 Dec 85 12:55:43-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: HOLIDAY PARTY NEXT FRIDAY(THE 13TH!)
To: FOLKS@SU-CSLI.ARPA




     *         *              *             *              *
*        *           *             *            *        *
   *         *     *     *       *      *             *
*       *      *       *      *       *       *     *      *
   *       *      *       *      *       *       *     *
*       *     *       *      *       *           *
   *                      *                  *         *
        *    *       *           *       *       *         *
  *  *         *           *        *        *           *
*                   *                *                    *
 *             *                 *            **     *
  *                *         *           *                 *
 * *       *                 *          *    *          *
       *       <↑>        *         *                *
 *              ↑ *            *            ←←←←←      *
   *      *    /↑\        *                 |   |   *
              /   \                         |   |
     *    *  /~~~~'\    *      *     *    ---------    *     *
 *          /'-/`'  \       *              / # # \    *   *
    *      /  ↑↑↑ *  \           *         (  ↑  )   
          /           \    *          *      `='
      *  / '~~~~~#~~`~~\      *    *       /     \    *
*       /   *           \                 /       \       *
   *   /       ````++~~~~\    *     *  >--(  CSLI )--<      *
      /  #~~~~~~~~~~~`....\     *         (       )    *
  *  /                     \        *      \=====/         *
    /  +===`~~~~``~~~~~"''* \  *      *    /     \    *   
   <  &      $     """"`=-`  >     *      (       )      *
*   ↑↑↑↑↑↑↑↑↑↑↑||||↑↑↑↑↑↑↑↑↑↑             (       )   * 
               ||||                    *  (       )       *
               ||||              *         \     /  *        *
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

   'TIS THE SEASON TO BE JOLLY,  FA LA LA LA LA LA LA LA LA

   SINGING SONGS BY BUDDY HOLLY, FA LA LA LA LA LA LA LA LA

   DON WE NOW OUR POT LUCK DISHES, FA LA LA LA LA LA LA LA LA

   PREPARE DAVE HIS FAREWELL WISHES, FA LA LA LA LA LA LA LA LA

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

   NEXT FRIDAY , THE LUCKY 13TH, PREPARE YOURSELVES FOR THE 
                                                                
        CCCCCCCC     SSSSSS       LLLLL               IIIII
      C         C   S      S        L                   I
     C             S                L                   I
    C              S                L                   I
    C               S               L                   I
    C                SSSSS          L                   I
    C                     S         L                   I
    C                      S        L                   I
     C                     S        L                   I
      C         C  S      S         L         L         I
        CCCCCCCC    SSSSSS        LLLLLLLLLLLLL       IIIII

            HOLIDAY POT-LUCK AND FAREWELL PARTY.

WE WILL BE CELEBRATING THE HOLIDAY SEASON AND SENDING DAVE
BROWN OFF TO PARTS-UNKNOWN IN THE SUBURBAN WASTELAND OF 
GREATER LOS ANGELES........DRINKS AND THE GREATER PART OF
A MEAL WILL BE PROVIDED AND WE ASK THAT YOU BRING YOUR FAVORITE
SIDE DISH TO COMPLETE THIS POT-LUCK EXTRAVAGANZA.

LIVE ENTERTAINMENT WILL BE PROVIDED BY YOURS TRULY...

AND A FUN TIME IS GUARANTEED FOR ALL!! SO PLAN ON IT!!

SEE YOU THERE!!


CONSULT PARTY PLANNERS - SUSI, TOM OR JAMIE - FOR DETAILS!



-------

∂06-Dec-85  1309	YAMARONE@SU-CSLI.ARPA 	WHAT TIME'S THE PARTY BEGIN?    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  13:08:18 PST
Date: Fri 6 Dec 85 13:04:39-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: WHAT TIME'S THE PARTY BEGIN?
To: FOLKS@SU-CSLI.ARPA


HEY, CAN YOU BELIEVE I DIDN'T MENTION A TIME FOR SUCH A LAVISH EVENT?

WHAT DO YOU SAY WE'LL BEGIN THE FESTIVITIES AT 3:00 AND BEGIN TO EAT
AND ENTERTAIN CLOSER TO 4:00?

OK, THAT'S THE TIME TABLE....REMEMBER, DON'T MAKE ANY DINNER PLANS
'CAUSE YOU'LL BE STUFFED AND STUMBLIN' WHEN YOU LEAVE HERE.
(THOSE WHO PRACTICE MODERATION, DON'T LET THIS STATEMENT UPSET YOU.)

BYE.

-------

∂06-Dec-85  1408	RICE@SUMEX-AIM.ARPA 	Explorers have new MIBs 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  14:07:57 PST
Date: Fri 6 Dec 85 14:05:47-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Explorers have new MIBs
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12165033670.72.RICE@SUMEX-AIM.ARPA>


New monitor interface boards have been put into explorers numbers 6, 8, 10, 11, 13 and 16.  This should cure the problems with the spurious characters and the
vigorous auto repeat mode.  If you see any more problems of this nature please
could you contact me or Mr. Acuff.

Rice.
-------

∂06-Dec-85  1423	@SU-CSLI.ARPA:PCOHEN@SRI-AI.ARPA 	Seminar by N. S. Sridharan, BBN Labs
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  14:23:47 PST
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Fri 6 Dec 85 14:18:52-PST
Date: Fri 6 Dec 85 14:19:06-PST
From: Phil Cohen <PCOHEN@SRI-AI.ARPA>
Subject: Seminar by N. S. Sridharan, BBN Labs
To: folks@SU-CSLI.ARPA

N. S. Sridharan, of BBN Labs, will give a seminar next Tues, the 10th,
in EJ228,   title and abstract below.

Title:  Semi-applicative Programming

Abstract:

 Most current parallel programming languages are designed
with a sequential programming language as the base language and have
added constructs that allow parallel execution.  We are experimenting
with an applicative base language that has implicit parallelism
everywhere, and then we introduce constructs that inhibit parallelism.
The base language uses pure LISP as a foundation and blends in
interesting features of Prolog and FP.  Proper utilization of
available machine resources is a crucial concern of programmers.  We
advocate several techniques of controlling the behavior of functional
programs without changing their meaning or functionality: program
annotation with constructs that have benign side-effects, program
transformation and adaptive scheduling.  This combination yields us a
SEMI-APPLICATIVE programming language and an interesting programming
methodology.

Starting with the specification of a context-free recognizer, we have
been successful in deriving variants of the recognition algorithm of
Cocke-Kasami-Younger.  One version is the CKY algorithm in parallel.
The second version includes a top-down predictor to limit the work
done by the bottom-up recognizer.  The third version uses a cost
measure over derivations and produces minimal cost parses using a
dynamic programming technique.  In another line of development, we
arrive at a parallel version of the Earley algorithm.
-------

∂06-Dec-85  1629	@SU-CSLI.ARPA:PCOHEN@SRI-AI.ARPA 	time for seminar by N. S. Sridharan 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  16:29:21 PST
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Fri 6 Dec 85 16:25:45-PST
Date: Fri 6 Dec 85 16:25:04-PST
From: Phil Cohen <PCOHEN@SRI-AI.ARPA>
Subject: time for seminar by N. S. Sridharan
To: folks@SU-CSLI.ARPA,
    AIC-PStaff: ;

is 10:30 AM, in EJ228, SRI.  

Sorry 'bout that!

Phil
-------

∂06-Dec-85  1634	PHILOSOPHY@SU-CSLI.ARPA 	colloquium
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  16:34:47 PST
Date: Fri 6 Dec 85 16:30:28-PST
From: Eve Wasmer <PHILOSOPHY@SU-CSLI.ARPA>
Subject: colloquium
To: bboard@SU-CSLI.ARPA
cc: folks@SU-CSLI.ARPA



                    Philosophy Department





                    Professor Susan Wolf

                    Philosophy Department
                    University of Maryland



               

          Above and Below the Line of Duty



                  Monday, December l6, 3:15 p.m


                  Philosophy Seminar Room 92Q
-------

∂06-Dec-85  2050	@SU-SUSHI.ARPA:karp@ernie.berkeley.edu  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 6 Dec 85  20:49:57 PST
Received: from ucbvax.berkeley.edu by SU-SUSHI.ARPA with TCP; Fri 6 Dec 85 20:48:08-PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA01902; Fri, 6 Dec 85 09:38:06 PST
Received: by ernie (5.31/5.16)
	id AA28232; Fri, 6 Dec 85 09:37:26 PST
Date: Fri, 6 Dec 85 09:37:26 PST
From: karp@ernie.berkeley.edu (Richard Karp)
Message-Id: <8512061737.AA28232@ernie>
To: aflb.local@su-sushi.arpa, allmsgs@ernie.berkeley.edu,
        csfaculty@ernie.berkeley.edu, csmsgs@ernie.berkeley.edu,
        theory-b@ernie.berkeley.edu, theory-f@ernie.berkeley.edu


MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley, Ca. 94720  (415) 642-0143

Seminars in Computational Complexity
Tuesday, December 10  2:00  MSRI Lecture Hall
Jeff Lagarias "The Nonlinear Geometry of Linear Programming"

Tuesday, December 10  4:00  MSRI Lecture Hall
Peter Shor  " Tight Bounds for Up-Right Matching, with Applications to Packing"

Thursday, December 12  4:00  MSRI Lecture Hall
George Lueker  "Linear Programming with Two Variables per Inequality in Poly-Log
Time"

∂07-Dec-85  0846	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #44  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 7 Dec 85  08:46:53 PST
Date:  7 Dec 85 0824-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #44
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Saturday, 7 Dec 1985      Volume 1 : Issue 44

Response to PARSYM Hardware Survey:

        SPUR (Berkeley)

[Responses are still welcome to the PARSYM Hardware Survey (V1 #33).
I don't expect you to neglect your Christmas shopping, but if you have
a moment, I'm sure the PARSYM readership would appreciate your
efforts.  Please feel free to fill out only the first part of the
survey.  To save you the trouble of looking up the back issue, the
"short form" is reproduced below. -- BD

	       The Third PARSYM Survey: Computer Hardware

                (Reply to PARSYM-Survey@SUMEX-AIM.ARPA)

			        Part I

    What computer hardware are you using or developing?

    What are its components (processors, memory) and how many; how
    do the processors and memory communicate?

    SIMD/MIMD?  Shared memory or distributed memory?

    What model of concurrency are you investigating with this system?

    What would you like to see in your next hardware system?
]

----------------------------------------------------------------------

From: larus@dali.berkeley.edu (James Larus)
Subject: PARSYM Survey, better late than ...
Date: 06 Dec 85 17:47:17 PST (Fri)

This response describes the SPUR project at UC Berkeley.

>> 			      Part I (Easy)
>> What computer hardware are you using or developing?

We are developing a multiprocessor workstation called SPUR (Symbolic
Processing Using RISCs (Reduced Instruction Set Computers)).


>> What are its components (processors, memory) and how many; how do the
>> processors and memory communicate?

The processors are custom VLSI, similar to RISC II processors, with an
on-chip instruction buffer and some additional instructions to support
Lisp.  Each processor has a coprocessor that supports IEEE standard
floating point and a large (128 Kbyte) cache.

We expect 6-10 processors will be the maximum that our current bus can
support.


>> SIMD/MIMD?  Shared memory or distributed memory?

MIMD.  There is a single, shared memory with a snooping cache scheme
to keep the per-processor caches consistent.


>> What model of concurrency are you investigating with this system?

We are investigating multiprocessor Lisp systems.


>> What would you like to see in your next hardware system?

It is a bit early to say, but a better bus would probably help a lot.



>> 		  Part II (Optional, for extra credit)


>> 0. Project summary:

>> What model of concurrency (e.g., dataflow, communicating sequential
>> processes, actors, futures, etc.) is your work based on?  What
>> language, if any, are you using in your work?

As mentioned above, we are implementing a multiprocess Lisp system.
yWe are looking at both the Multilisp and Qlambda approaches and are
trying to provide a set of primitives that will allow both techniques
to be implemented.


>> 1. Architectural classification:

>> SIMD or MIMD?

MIMD.


>> 2. Processing elements:

>> Describe each processor in the system in terms of type (e.g., 68000
>> equivalent, Lisp machine, custom), "power" (e.g., in MIPS or relative
>> to a Vax or 68000), and memory size.  If the processors are simulated
>> then also describe the uniprocessor and mechanism for simulation.  If
>> the processors in the system are not of uniform type then describe
>> each type and their relationship to each other.

The processors are all custom VLSI.  They have run various Lisp
benchmarks from Gabriel's book roughly twice as fast as a Symbolics
3600 or VAX 8600.  Their performance on C, Pascal, or Fortran should
be roughly comparable to a VAX 8600.


>> How many processors are there in the minimum and maximum configurations?
>> How many are you currently using?

Mimimum = 1.  Maximum = 10 - 12.  Currently we are simulating 1
processor.


>> 3. Memory system:

>> Describe how memory is distributed and accessed in the system.  Shared
>> memory or distributed memory?  Is there a single address space or does
>> each processor have its own address space?  What sort of caching is
>> employed, if any: snoopy, write-through, or something else?  What
>> sizes are the caches?

There is a single, shared memory.  It has 256 1-gigabyte segments.
A processor can access 4 segments at once.  Each processor has a 128
Kbyte cache that is kept consistent using a Snooping cache scheme.


>> What are the minimum and maximum amounts of memory?  How much are you
>> currently working with?

No real limit on the size of the main memory, except cost and
packaging.


>> 4. Communication system:

>> How are processors connected to each other and to memory?  Describe in
>> terms of topology (e.g., tree, systolic array, pipe, grid), switching
>> type (e.g., LAN, bus, Butterfly switch), and throughput (e.g., bus
>> transmission rates, time to access memory on remote nodes, time to
>> broadcast a message to all nodes).

The processors are connected to the main memory by a NuBus.  They
communicate through the memory and asynchronous interrupts.


>> 5. Synchronization:

>> What hardware mechanisms does the system supply to support
>> synchronization?

Test-and-set instruction and inter-processor interrupts.


>> 6. Pragmatics

>> Is your system geared towards either specific programming languages or
>> applications?

The processors have some support for Lisp, though they should run
conventional languages (C, Pascal, etc.) as well.  The applications
at which we are looking are "symbolic processing," i.e., compilation,
hardware simulation, etc.


>> Is the hardware of your system generally available or, if custom-made,
>> likely to be available in future?

The first set of workstations will be prototypes for use at Berkeley.
It is hoped that there will be a second design that will be done in
conjunction with a commercial vendor.


>> If you're now using a multiprocessor system, how does the architecture
>> of your development system correspond to your research target
>> architecture?

Not applicable.


>> What are the advantages of your hardware setup, particularly for the
>> problems you are investigating?  What are the disadvantages?  Is there
>> a system that would better meet your needs?

Ask again in 6 months or a year.


/Jim

------------------------------

End of PARSYM Digest
********************

∂08-Dec-85  2355	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #45  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 8 Dec 85  23:55:00 PST
Date:  8 Dec 85 2349-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #45
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest             Monday, 9 Dec 1985       Volume 1 : Issue 45

Today's Topics:

 Seminar: Massively Parallel Networks That Learn Representations (MIT)

----------------------------------------------------------------------

Date: 12/06/85 10:35:37
From: ELIZABETH%MIT-OZ@MIT-MC.ARPA
Re:   9.382 --- Seminar in Visual Information Processing

[Forwarded from the MIT-BBoard by SASW@MIT-MC]

        MASSIVELY PARALLEL NETWORKS THAT LEARN REPRESENTATIONS

                           Geoffrey Hinton
                      Carnegie-Mellon University

                          Monday, December 9
                    Eighth Floor Playroom, AI Lab
                               4:00 pm

I shall describe a new learning procedure for massively parallel
networks of neuron-like processing elements.  The procedure adjusts
the connection strengths in a multi-layered network so as to make it
give the correct output vector when given an input vector.  The units
in the intermediate layers come to represent important implicit
features of the task domain that generates the pairs of input/output
vectors.  As a result, the network can generalise appropriately to new
cases.  I shall describe a pattern recognition example in which the
network constructs a balanced ecology of feature detectors, and a
higher level task in which the network learns a set of relationships.
The second example illustrates the ability of the network to recognize
isomorphisms and make use of them in encoding and generalizing
knowledge.

------------------------------

End of PARSYM Digest
********************

∂09-Dec-85  1032	RICHARDSON@SU-SCORE.ARPA 	General Faculty Meeting 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Dec 85  10:32:19 PST
Date: Mon 9 Dec 85 10:28:01-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: General Faculty Meeting
To: faculty@SU-SCORE.ARPA
Message-ID: <12165780460.17.RICHARDSON@SU-SCORE.ARPA>

Here is the agenda for tomorrow's general faculty meeting (Dec. 10) to be
held in MJH 146 from 1:15 to 3:45.
-------------------------------------------------------------------------


1.  Reports
    @  Financial, Betty Scott
          (items: salary research offsets, ...)
    @  Facilities, Lester Earnest
          (items:  Courtesy Computer Accts Policy, APS, Long-range Plans, ...)
    @  PhD Program Committee Progress Report, Terry Winograd
    @  Forum, Carolyn Tajnai
    @  Others as may be appropriate

2.  Report on Long-Range Planning Process, Nils Nilsson

3.  Report on Undergraduate Major Planning
    @  Curriculum Matters, Jeff Ullman and Stuart Reges
    @  Summary of Curriculum approval process with UG Council and C-US, 
                                   Jeff Ullman and Nils Nilsson
    @  Summary of discussions with Dean Gibbons regarding resources, Nils Nilsson
    @  Discussion

4.  Report on Progress on Searches, Nils Nilsson
    @  Robotics
    @  Systems
    @  Possible New Searches

5.  Proposed Consulting Professors (Tenenbaum and Hayes), Nils Nilsson
    @ (curriculum vitaes distributed to faculty prior to meeting)

6.  Status of New Patent and Copyrights Policy, Nils Nilsson

7.  Proposal on "Visiting Professorship,"  John McCarthy and Nils Nilsson
    
8.  Suggestion for a Survey of CSD Graduates, Bob Floyd
-------

∂09-Dec-85  1045	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Dec 85  10:44:31 PST
Date: Mon 9 Dec 85 10:40:54-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
cc: library@SU-SCORE.ARPA, maslin@SUMEX-AIM.ARPA
Message-ID: <12165782805.17.RICHARDSON@SU-SCORE.ARPA>


Tomorrow is the final CSD Tuesday Lunch for the Fall Quarter in MJH 146 at
12:15. There will be general discussion of your choosing.
-------

∂09-Dec-85  1333	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books--CS   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 9 Dec 85  13:33:27 PST
Date: Mon 9 Dec 85 13:25:10-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books--CS
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA, cs%Playfair@SU-SCORE.ARPA
Message-ID: <12165812708.18.LIBRARY@SU-SCORE.ARPA>

Computer Speech Processing. by Frank Fallside and William Woods.
TK7895.S65F35 1985.

IEEE Computer Society. Proceedings. Trends and Applications 1985. Utilizing
Computer Graphics. May 1985.   T385.T74 1985.

Self-Validating Numerics For Function Space Problems.  Notes and Reports
in Computer Science and Applied Mathematics. by E. Kaucher and W. Miranker.
QA323.K38 1984.

Smart Robots; A Handbook Of Intelligent Robotic Systems. by Daniel Hunt.
TJ211.H86 1985 c.2

Applied Numerical Analysis. 3rd ed. by C. Gerald and P. Wheatley.
QA297.G47 1984 c.2

New Topics In Learning Automata Theory and Applications. by N. Baba.
Lecture Notes In Control and Information Sciences. Q335.B27 1984.

H. Llull
-------

∂09-Dec-85  1720	@SU-CSLI.ARPA:WALDINGER@SRI-AI.ARPA 	seminar on program transformation wednes, 3:45  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 9 Dec 85  17:20:01 PST
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Mon 9 Dec 85 17:16:55-PST
Date: Mon 9 Dec 85 17:09:43-PST
From: WALDINGER@SRI-AI.ARPA
Subject: seminar on program transformation wednes, 3:45
To: AIC-Associates: ;,
    CSL: ;, su-bboards@SU-AI.ARPA, friends@SU-CSLI.ARPA, bboard@SRI-AI.ARPA

Title: A Closer Look at the Tupling Strategy for Program Transformation

Speaker:  Alberto Pettorossi, IASI-CNR, Rome, Italy

Place: EK242 (Old AIC Conference Room), SRI International, 
   Ravenswood Avenue and Pine Street

Time:  3:45 pm Wednesday, 11 December

   Coffee in Waldinger office at 3:15

Abstract:
Tupling is a strategy for transforming programs expressed as recursive
equations. We see how it applies to some challenging "little"
problems: the tower of Hanoi, the Chinese rings problem, drawing
Hilbert curves, and computing recurrence relations. We characterize
the power of the tupling strategy in terms of the structure of the
graphs we obtain by unfolding the functions of the programs.
-------

∂09-Dec-85  2250	DAVIES@SUMEX-AIM.ARPA 	AAP meeting Wednesday 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 9 Dec 85  22:49:53 PST
Date: Mon, 9 Dec 1985  22:49 PST
Message-ID: <DAVIES.12165915422.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA, BHayes-Roth@SUMEX-AIM.ARPA
Subject: AAP meeting Wednesday
cc:   Davies@SUMEX-AIM.ARPA

There will be a meeting this Wednesday at 9:30 am.  One likely topic
is L100, James Rice's blackboard programming language.  Also, we will
welcome our newest member, Hiroshi "Gitchang" Okuno.

	-- Byron

∂10-Dec-85  1029	RICHARDSON@SU-SCORE.ARPA 	CSD Tuesday Lunch  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  10:29:31 PST
Date: Tue 10 Dec 85 10:25:48-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD Tuesday Lunch
To: faculty@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12166042201.14.RICHARDSON@SU-SCORE.ARPA>

Last lunch of the quarter today at 12:15 in MJH 146.

-------

∂10-Dec-85  1214	RICHARDSON@SU-SCORE.ARPA 	CSD 500 Colloquium Series    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  12:14:16 PST
Date: Tue 10 Dec 85 12:07:54-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: CSD 500 Colloquium Series
To: su-bboards@SU-SCORE.ARPA, colloq@SU-SCORE.ARPA, csd@SU-SCORE.ARPA
Message-ID: <12166060788.14.RICHARDSON@SU-SCORE.ARPA>


PLEASE NOTE THE NEW LOCATION FOR THE CS500 COLLOQUIUM SERIES TODAY (DEC.10)
DUE TO A SCHEDULING CONFLICT WITH A FINAL.


---------------------------------------------------------------------------
DAY           December 10, 1985
EVENT         Computer Science Colloquium
PLACE         Terman Auditorium
TIME          4:15
TITLE         A Knowledge-Based Approach to High-Level Program Optimization
PERSON        Allen Goldberg
FROM          Kestrel Institute and Univ. of California, Santa Cruz

      A KNOWLEDGE-BASED APPROACH TO HIGH-LEVEL PROGRAM OPTIMIZATION

Our concern is the compilation of high-level declarative programs into
efficient Pascal-level implementations. Specifications are written as
equational assertions over a set-theoretic data domain.  Compilation
is  achieved by semi-automatic application of source-to-source
transformations.  These transformations formalize knowledge about
high-level optimization techniques required for the task.  These
techniques include finite differencing, loop fusion, algebraic
simplification, symbolic evaluation, data structure selection, and
store-vs-compute.  The relevance of this work to the optimization of
Prolog and functional programs will be discussed.
-------

∂10-Dec-85  1235	DAVIES@SUMEX-AIM.ARPA 	Wednesday AAP meeting 9:30 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  12:35:01 PST
Date: Tue, 10 Dec 1985  12:34 PST
Message-ID: <DAVIES.12166065567.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: Wednesday AAP meeting 9:30

The topic of the Wednesday meeting is the concurrent blackboard
architecture TINA.  People interested in implementing applications are
especially encouraged to attend.

	-- Byron

∂10-Dec-85  1237	BROWN@SU-CSLI.ARPA 	Message for Sandwich
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  12:36:37 PST
Date: Tue 10 Dec 85 12:32:35-PST
From: Dave <BROWN@SU-CSLI.ARPA>
Subject: Message for Sandwich
To: folks@SU-CSLI.ARPA

There are 2 extra sandwiches today.....

1 Turkey

1 Turkey and avacado

-The Ventura Sandwich Corp.
-------

∂10-Dec-85  1319	LIBRARY@SU-SCORE.ARPA 	The Connection Machine by Hillis
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  13:19:35 PST
Date: Tue 10 Dec 85 13:11:58-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: The Connection Machine by Hillis
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA
Message-ID: <12166072449.21.LIBRARY@SU-SCORE.ARPA>

The Math/CS Library has The Connection Machine as a technical report
number 41646 and in the book collection under call number QA267.H487 
1985.  Enigneering Library also has a copy of the book.  Our copy
is currently checked out (the book).  If you wish to see it, send
a message and we will place you on the waiting list.
H. Llull
-------

∂10-Dec-85  1834	ullman@su-aimvax.arpa 	tomorrow's meeting    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  18:34:52 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 10 Dec 85 18:30:41 pst
Date: Tue, 10 Dec 85 18:30:41 pst
From: Jeff Ullman <ullman@diablo>
Subject: tomorrow's meeting
To: nail@diablo

As usual, we have no official program.
However, several people have been talking to me about what looks
like almost the same problem, and I'd like to get a discussion
going (you know who you are).
The general area is what happens when we expand a rule or rules
to get an infinite sequence of relational algebra expressions,
and we want to prove some property holds for all (or almost
all, or for an infinite number) of these expressions.

Meet at 11AM.
				---Jeff

∂10-Dec-85  2226	vardi@su-aimvax.arpa 	Database Seminar  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 10 Dec 85  22:26:18 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 10 Dec 85 22:23:15 pst
Date: Tue, 10 Dec 85 22:23:15 pst
From: Moshe Vardi <vardi@diablo>
Subject: Database Seminar
To: nail@diablo

From @SU-SCORE.ARPA:MILTON@SRI-AI.ARPA Tue Dec 10 00:57:03 1985
Received: from SU-SCORE.ARPA by su-aimvax.arpa with Sendmail; Tue, 10 Dec 85 00:56:59 pst
Received: from SRI-AI.ARPA by SU-SCORE.ARPA with TCP; Tue 10 Dec 85 00:53:39-PST
Date: Tue 10 Dec 85 00:55:18-PST
From: MILTON@SRI-AI.ARPA
Subject: Database Research Seminar for 12/13
To: CS545:;

We will meet this Friday at 3:15 in room 352 MJH for the last seminar of the
quarter --


     Integrating Databases and Logic Based AI Representation Systems

                          Richard Weyhrauch
                         Stanford University


The recognition that relational data bases can be directly viewed as a
model for a first order theory is now well understood.  We will give a
formal presentation of FOL, a reformulation of first order logic, that
will make data base access, reasoning about data and using meta theoretic
reasoning available in the same system.  One example of the power of this
system is the ability to distinguish explicitly between making the closed
world assumption and the case where the data base is simply incomplete.
-------

∂11-Dec-85  0617	PATASHNIK@SU-SUSHI.ARPA 	Last AFLB of the quarter 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 11 Dec 85  06:15:38 PST
Date: Wed 11 Dec 85 06:12:01-PST
From: Oren Patashnik <PATASHNIK@SU-SUSHI.ARPA>
Subject: Last AFLB of the quarter
To: aflb.all@SU-SUSHI.ARPA
Message-ID: <12166258145.8.PATASHNIK@SU-SUSHI.ARPA>

Here's the abstract for the last AFLB of the quarter.  Also, later in
day there's another talk that might be of interest; details follow
the AFLB abstract.

	  **************************************************

12-Dec-85  :  Joan Hutchinson (MSRI & Smith College)

	  Results and Algorithms in Topological Graph Theory

In this seminar I will present results on "separating" algorithms and
"planarizing" algorithms for graphs of bounded genus.  Namely, given a
graph with n vertices and genus g we look for a set of O(sqrt(gn))
vertices whose removal leaves all components of size at most 2n/3 and
a set of O(sqrt(gn)) vertices whose removal leaves a planar graph.  As
time permits I will also discuss Robertson and Seymour's recent work
that provides polynomial time algorithms for these problems:
i) for fixed k, upon input G and k pairs of vertices (s←i,t←i) of G,
determine whether there are k vertex-disjoint paths joining s←i and
t←i for i=1..k.
ii) for a fixed graph H, determine whether G contains H as a minor.

***** Time and place: December 12, 12:30 pm in MJ352 (Bldg. 460) ******

Additional talk:
	Department of Mathematics Colloquium
	Prof. Wolfgang Maass - MSRI & Univ. of Illinois, Chicago Circle
	"Lower Bound Arguments for Turing Machines"
	Thursday, December 12, 4:15 p.m.
	Bldg 380 (Math), Room 380-W

	  **************************************************

Regular AFLB meetings are on Thursdays, at 12:30pm, in MJ352.  If you
have a topic you'd like to talk about please let me know.  (Electronic
mail: patashnik@su-sushi.arpa, phone: (415) 497-1787). Contributions
are wanted and welcome.  Not all time slots for this academic year
have been filled.  The file [SUSHI]<patashnik.aflb>aflb.bboard contains
more information about future and past AFLB meetings and topics.
	--Oren Patashnik
-------

∂11-Dec-85  1424	RICE@SUMEX-AIM.ARPA 	Today's talk. 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 11 Dec 85  14:23:33 PST
Date: Wed 11 Dec 85 14:22:46-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Today's talk.
To: aap@SUMEX-AIM.ARPA
Message-ID: <12166347483.17.RICE@SUMEX-AIM.ARPA>


If anyone want Xeroces of the slides of the proto worked examples please
come and see me.  I also have a few sets of the manuals left over for
anyone who failed to get them today.

Rice.
-------

∂11-Dec-85  1436	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:udi@rsch.wisc.edu 	Technical Debate at Stanford on SDI, December 19   
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 11 Dec 85  14:36:23 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Wed 11 Dec 85 14:32:00-PST
Received: from rsch.wisc.edu by SU-SCORE.ARPA with TCP; Wed 11 Dec 85 14:32:00-PST
Received: by rsch.wisc.edu; Wed, 11 Dec 85 16:13:50 CST
Received: from SU-SUSHI.ARPA by rsch.wisc.edu; Wed, 4 Dec 85 21:05:08 CST
Date: Wed 4 Dec 85 15:50:28-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: Technical Debate at Stanford on SDI, December 19
To: theory@rsch.wisc.edu
Message-Id: <12164528442.63.JF@SU-SUSHI.ARPA>
Status: R
Resent-To: TheoryNet-list@rsch.wisc.edu
Resent-Date: 11 Dec 85 16:12:45 CST (Wed)
Resent-From: Udi Manber <udi@rsch.wisc.edu>

		``SDI: How Feasible, How Useful, How Robust?''

This will be a technical debate, covering both hardware and software aspects 
of SDI.
The debate will be videotaped by the Stanford Instructional Television Network.


Sponsor: Stanford Computer Science Department

Date: December 19, 1985

Time: 8:00 p.m.

Place: Terman Auditorium

Organizer: Barbara Simons, IBM-SJ

Moderator:  Dr. Marvin L. Goldberger, President of Cal Tech.
Former member of President's Science Advisory Committee
and Consultant on Arms Control and International Security.

Panelists:

Opponents:
Dr. Richard L. Garwin, IBM Fellow and Adjunct Professor of Physics at
Columbia University, Physicist and Defense Consultant.

Professor David Parnas, Lansdown Professor of Computer Science at the 
University of Victoria, Former member of the SDI Organization's
Panel on Computing and Support of Battle Management.

Advocates:
Professor Richard Lipton, Professor of Computer Science at Princeton
University, Current member of SDIO's Panel on Computing and Support of Battle
Management.

Major Simon Peter Warden, the Special Assistant to the Director of the SDIO
and Technical Advisor to the Nuclear and Space Arms Talk with the USSR
in Geneva.


--------------
TN Message #13
--------------

∂11-Dec-85  1752	EMMA@SU-CSLI.ARPA 	Newsletter December 12, No. 5  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 11 Dec 85  17:52:20 PST
Date: Wed 11 Dec 85 17:07:17-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter December 12, No. 5
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
December 12, 1985               Stanford                        Vol. 3, No. 5
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←

          CSLI ACTIVITIES FOR *THIS* THURSDAY, December 12, 1985

   12 noon		TINLunch
     Ventura Hall       Plural Determiners and Plural Anaphora
     Conference Room    by Werner Frey and Hans Kamp
			Discussion led by Werner Frey
			(Abstract on page 1)
			
   2:15 p.m.		CSLI Seminar
			No seminar this week

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
			Themes in the Evolutionary Biology of Language:
                        A three-ring circus
			Bjorn Lindblom, CASBS
			(Abstract on page 2)

                             --------------
                              ANNOUNCEMENT

   Please note that the seminar and colloquium are no longer in Redwood
   Hall room G-19.  The new room is Turing Auditorium in Jordan Quad.
                             --------------
                          THIS WEEK'S TINLUNCH
                 Plural Determiners and Plural Anaphora
                    Werner Frey, University of Texas

      Werner Frey will discuss his and Hans Kamp's work on plural noun
   phrases, focusing on:

     a) The interpretation of anaphoric plural pronouns, with special
   attention to the ways in which it differs from that of singular
   pronouns.
     b) The difference between definite plural NP's, such as `the boys'
   and `indefinite' plurals, such as e.g., `many boys'.
     c) The nature of the definite article `the', both in its plural and
   its singular uses.

   Copies of a longer abstract will be available at the Ventura desk.

!
Page 2                     CSLI Newsletter                  December 12, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                         THIS WEEK'S COLLOQUIUM
   Themes in the Evolutionary Biology of Language: A three-ring circus
    Bjorn Lindblom, Peter MacNeilage, Michael Studdert-Kennedy, CASBS

      The goal of our research is summarized by the phrase: DERIVE
   LANGUAGE FROM NON-LANGUAGE! We are exploring an approach to the
   biology of language that is deliberately distinct from that pursued
   within Chomskyan autonomous linguistics. We take as our first priority
   an explicit search for precursors to all aspects of language structure
   and speech behavior. By precursors we mean either evolutionary
   precursors, traceable to lower animals, or those human but
   non-linguistic, cognitive, perceptual and motor capacities that both
   constrain language and make it possible. Only when a search for such
   precursors has failed can we justly term some characteristic
   unique---either to language or to man---and attribute it to some
   species-specific bioprogram for language learning and use (cf.
   universal grammar). In short, while we acknowledge and respect the
   discoveries of formal linguistics, we believe that a sound approach to
   the biology of language must go beyond form and structure to ask:
   ``How did language get that way?''
      A major language universal for which any phylogenetic or
   ontogenetic theory must account is LA DOUBLE ARTICULATION, or DUALITY
   of patterning.  We view the emergence of duality---that is, the use of
   discrete elements and combinatorial rules at the two levels of
   phonology and syntax---as the key to the communicative power of
   language: duality provides a kind of impedance match between the
   essentially unlimited semantics of cognition and a decidedly limited
   set of signaling devices and processes.
      Our central concern is with phonology: with the origin of discrete
   phonological elements---phonemes and features---and with the processes
   by which these intrinsically amodal elements are instantiated in the
   modalities of production and perception. We shall review typological
   facts about sound structure leading us to conclude that phonological
   form adapts to the performance constraints of language use. How do we
   choose our theoretical formalism for describing sound patterns?
   Markedness theory or contemporary developments of generative phonology
   and formal linguistics? No, since (i) spoken language is a product of
   biological and cultural evolution; and (ii) there is considerable
   empirical justification for viewing phonologies as adaptations to
   biological and social selectional pressures the correct choice appears
   to be some variant of the theoretical framework currently explored by
   many students of biological and cultural evolution, viz., a Darwinian
   VARIATION*SELECTION model. In our talk we will present a computational
   implementation of such a model. We will illustrate some of its
   explanatory power by means of simulations indicating how a number of
   typological facts can be predicted quantitatively and how the
   emergence of ``a featural and segmental'' organization of lexical
   systems can be derived in a self-organizing manner and deductively
   (rather than just axiomatically). Drawing on corpora of speech error
   data we describe the process by which discrete elements are organized
   and sequenced in an actual utterance (phonologically and syntactically)
   as one of inserting elements into a structured frame, and, in our
   talk, we will consider the evolutionary relation between this
   FRAME-CONTENT mode of organization and bimanual coordination.  Finally
   we will consider behavioral and neurophysiological evidence, from both
   adults and children, consistent with a view of the phonological
   element as an AMODAL structure linking production and perception.
!
Page 3                     CSLI Newsletter                  December 12, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
                       Reflexivization Variation:
       Relations between Syntax, Semantics, and Lexical Structure
                  Peter Sells, Annie Zaenen, Draga Zec
          Thursday, December 12, 10:00, Ventura Conference Room

      In this paper we examine the distinction between so-called
   transitive and intransitive reflexive construction in several
   languages (English, Finnish, German, Dutch, Chichewa, Warlpiri,
   Serbo-Croatian and Japanese).  We argue that three types of
   distinctions have to be made: transitivity versus intransitivity in
   the lexicon, synthetic versus analytic forms in the constituent
   structure and open versus closed predicates in the semantics; thus
   there are three relevant levels of possible cross-linguistic
   variation.  While there is a one-way implication between lexical
   intransitivity and closed predication, there are in general no direct
   correlations between either the lexical forms or the semantic forms
   and their constituent structure representation.
      We give an account of the different types of reflexive that we
   discuss in Lexical-Functional Grammar augmented with Discourse
   Representation Structures.
      Copies of the paper are available at the front desk.
                               ----------
                              CSLI SEMINAR
         NETTALK:  Teaching a Massively-Parallel Network to Talk
                  Terrence J. Sejnowski, Johns Hopkins
              1:00pm, Wednesday, December 18, CSLI trailers
           A special seminar in place of Pixels and Predicates

      Text to speech is a difficult problem for rule-based systems
   because English pronunciation is highly context dependent and there
   are many exceptions to phonological rules.  An alternative knowledge
   representation for correspondences between letters and phonemes will
   be described in which rules and exceptions are treated uniformly and
   can be determined with a learning algorithm in a connectionist model.
   The architecture is a layered network of 400 simple processing units
   with 9,000 weights on the connections between the units.  The training
   corpus is continuous informal speech transcribed from tape recordings.
   Following training on 1000 words from this corpus the network can
   generalize to novel text.  Even though this network was not designed
   to mimic human learning, the development of the network in some
   respects resembles the early stages in human language acquisition.
   Following damage of the network by either removal of units or addition
   of random values to the weights the performance of the network
   degraded gracefully.  Issues which will be addressed include scaling
   of the learning algorithm with the size of the problem, robustness of
   learning to predicate order of the problem, and universality of
   learning in connectionist models.
!
Page 4                     CSLI Newsletter                  December 12, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                       ENVIRONMENTS GROUP MEETING
         DefinitionGroups: Organizing Programs in Time and Space
                       Daniel Bobrow, Xerox PARC
              Monday, 12:00, December 16, Ventura Trailers

      Most current systems use files for long term storage, and to
   organize the conceptual structure of the system.  The definition
   groups project (with Daniel Bobrow, David Fogelsong and Mark Miller)
   is exploring an object oriented approach to the organization of a
   system, and to the maintenance of sequential incremental changes.  It
   will also support the exploration of alternative development paths.
   Extensive use of browsers allows alternative views and interaction
   with the program structure.
      DEFGROUPS uses current file system as a base, but is set up to move
   to a database system.  A prototype of the system is currently working,
   and supporting its own development.

-------

∂12-Dec-85  0902	LIBRARY@SU-SCORE.ARPA 	USENIX and UNIFORUM Conference Proceedings
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Dec 85  09:01:59 PST
Date: Thu 12 Dec 85 08:58:09-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: USENIX and UNIFORUM Conference Proceedings
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA, mogul@SU-SCORE.ARPA
Message-ID: <12166550534.28.LIBRARY@SU-SCORE.ARPA>

I received a message today that a number of people are interested in the
library having USENIX Association conference proceedings.  I have order
these conferences along with the UniForum Conferences from the same
association.  I appreciate receiving requests like this whether or not
I have already order it or not.  It helps me monitor the level of interest
and change in that interest in the various research areas.  Please feel
free to make suggestions for new purchases by sending me messages 
Library@Score.

Harry Llull
-------

∂12-Dec-85  0943	YAMARONE@SU-CSLI.ARPA 	HOLIDAY CELEBRATION REMINDER    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 12 Dec 85  09:43:47 PST
Date: Thu 12 Dec 85 09:34:02-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: HOLIDAY CELEBRATION REMINDER
To: FOLKS@SU-CSLI.ARPA


TOMORROW, TOMORROW, WE'LL PARTY TOMORROW,
IT'S ONLY A DAY AWAY...........

     *         *              *             *              *
*        *           *             *            *        *
   *         *     *     *       *      *             *
*       *      *       *      *       *       *     *      *
   *       *      *       *      *       *       *     *
*       *     *       *      *       *           *
   *                      *                  *         *
        *    *       *           *       *       *         *
  *  *         *           *        *        *           *
*                   *                *                    *
 *             *                 *            **     *
  *                *         *           *                 *
 * *       *                 *          *    *          *
       *       <↑>        *         *                *
 *              ↑ *            *            ←←←←←      *
   *      *    /↑\        *                 |   |   *
              /   \                         |   |
     *    *  /~~~~'\    *      *     *    ---------    *     *
 *          /'-/`'  \       *              / # # \    *   *
    *      /  ↑↑↑ *  \           *         (  ↑  )   
          /           \    *          *      `='
      *  / '~~~~~#~~`~~\      *    *       /     \    *
*       /   *           \                 /       \       *
   *   /       ````++~~~~\    *     *  >--(  CSLI )--<      *
      /  #~~~~~~~~~~~`....\     *         (       )    *
  *  /                     \        *      \=====/         *
    /  +===`~~~~``~~~~~"''* \  *      *    /     \    *   
   <  &      $     """"`=-`  >     *      (       )      *
*   ↑↑↑↑↑↑↑↑↑↑↑||||↑↑↑↑↑↑↑↑↑↑             (       )   * 
               ||||                    *  (       )       *
               ||||              *         \     /  *        *
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

   'TIS THE SEASON TO BE JOLLY,  FA LA LA LA LA LA LA LA LA

   SINGING SONGS BY BUDDY HOLLY, FA LA LA LA LA LA LA LA LA

   DON WE NOW OUR POT LUCK DISHES, FA LA LA LA LA LA LA LA LA

   PREPARE DAVE HIS FAREWELL WISHES, FA LA LA LA LA LA LA LA LA

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

   TOMORROW, FRIDAY , THE LUCKY 13TH, PREPARE YOURSELVES FOR THE 
                                                                
        CCCCCCCC     SSSSSS       LLLLL               IIIII
      C         C   S      S        L                   I
     C             S                L                   I
    C              S                L                   I
    C               S               L                   I
    C                SSSSS          L                   I
    C                     S         L                   I
    C                      S        L                   I
     C                     S        L                   I
      C         C  S      S         L         L         I
        CCCCCCCC    SSSSSS        LLLLLLLLLLLLL       IIIII

            HOLIDAY POT-LUCK AND FAREWELL PARTY.

WE WILL BE CELEBRATING THE HOLIDAY SEASON AND SENDING DAVE
BROWN OFF TO PARTS-UNKNOWN IN THE SUBURBAN WASTELAND OF 
GREATER LOS ANGELES........DRINKS AND THE GREATER PART OF
A MEAL WILL BE PROVIDED AND WE ASK THAT YOU BRING YOUR FAVORITE
SIDE DISH TO COMPLETE THIS POT-LUCK EXTRAVAGANZA.

LIVE ENTERTAINMENT WILL BE PROVIDED BY THE FOLK/ROCK/
CHRISTMAS-SONG DUO OF JOHN JOSEPH AND TOM YAMARONE..

AND A FUN TIME IS GUARANTEED FOR ALL!! SO PLAN ON IT!!

FESTIVITIES BEGIN AT 4:00 PM
POT-LUCK "DINNER" BEGINS AT 5:00 PM.

HOPE TO SEE YOU THERE!!!
-------

∂12-Dec-85  1116	GOTELLI@SU-SCORE.ARPA 	Correction  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Dec 85  11:16:32 PST
Date: Thu 12 Dec 85 11:08:04-PST
From: Lynn Gotelli <GOTELLI@SU-SCORE.ARPA>
Subject: Correction
To: csd@SU-SCORE.ARPA
Message-ID: <12166574183.10.GOTELLI@SU-SCORE.ARPA>

Dan Kolkowitz will be moving into MJH030B sometime after the
first of the year NOT MJH030C.  
-------

∂12-Dec-85  1136	GOTELLI@SU-SCORE.ARPA 	New Staff Member 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 12 Dec 85  11:36:47 PST
Date: Thu 12 Dec 85 10:52:33-PST
From: Lynn Gotelli <GOTELLI@SU-SCORE.ARPA>
Subject: New Staff Member
To: csd@SU-SCORE.ARPA
Message-ID: <12166571357.10.GOTELLI@SU-SCORE.ARPA>

Please welcome Dan Kolkowitz who joins us in Computer Facilities
and will be responsible for the operation and software support
of the CF unix systems.  Dan can be found temporarily in MJH040D
until the end of the year when he will move into MJH030C.  We
are very happy to have Dan as a member of our staff.
-------

∂12-Dec-85  1238	ullman@su-aimvax.arpa 	Recent developments   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Dec 85  12:38:13 PST
Received: by su-aimvax.arpa with Sendmail; Thu, 12 Dec 85 12:31:26 pst
Date: Thu, 12 Dec 85 12:31:26 pst
From: Jeff Ullman <ullman@diablo>
Subject: Recent developments
To: nail@diablo

Those who left early (i.e., before 2PM) missed a good NAIL! meeting.
I'm going to try to sumarize the things that Jeff N., Allen, Shuky,
and I worked on.  We focussed on fixed points of monotone functions.
That is, let f be a function from k-ary relations to k-ary relations,
such that if X C← Y, then f(X) C← f(Y). [I'll use C← for "contained in"]
The principal source we have in mind for such function is SyRipS,
e.g, a rule like T(X,Y) :- E(X,Z) & T(Z,Y) defines the function
from T to T: f(T) = PROJ←{1,3}(E*T), i.e, the projection onto
columns 1 and 3 of the natural join of E and T.
However, we might like to use some f that does not come from a single rule.

Let f↑i be the composition of f with itself i times, and f* be the union
of f↑i for all i >= 1.
In what follows, it is important to remember that we are using the
"sigma" (Shuky's term) or "all-models" interpretation of functions.
That is, to say f C← g is to say that f(X) C← g(X) for all relations X.
f = g means the "containment" goes both ways.

We also assume that what we really want to know about is
the result of applying the function until a fixed point is reached.
This interpretation is equivalent to saying that for two rules to be
"equal" they must produce the same value for their predicate p when
started with p equal to an arbitrary relation,
and the rules are then applied until convergence occurs.
In comparison, the more conventional, but less tractable definition of
equivalence is to say that rules are equal if their least fixed points,
i.e., their results started with p empty, are the same.

The problem we would like to address is: given a monotone function f,
what is its "minimal" equivalent function?  In order to attack that
question, we at least have to answer the question whether two functions,
f and g are equivalent in the all-models sense, i.e., is f*=g*?

Shuky pointed out a while ago that if f and g come from full Datalog
rules ("full" in this sense means that a variable appearing in
the head also appears somewhere in the body.) then equivalence
of f* and g* is the same as testing whether f and g are equivalent
as dependencies, and thus f* = g* is decidable.
However, we still need efficient decision procedures for subcases.

LEMMA: f* C← g* if and only if f C← g↑i for some i.

COROLLARY: f* = g* iff f C← g↑i and g C← f↑j for some i and j.

Say f is containment-free if for no i>1 is f C← f↑i.
The test for f* = g* appears to depend significantly on whether
f or g are containment-free.

THEOREM: If f is containment-free then f* = g* iff f = g.

Note that for sufficiently general f and g, even testing f = g
might not be decidable.  However, for reasonable classes of functions,
f = g is easy to decide, e.g., if f and g are tableau mappings.

A similar issue comes up for deciding whether f (or g) is containment-free.
I think a lot of what Jeff N. has done regarding data-independent
recursion can be applied here.

The above theorem has an interesting partial converse.

THEOREM: if there is some i such that f is *properly* contained
in f↑i, then there is always some g (namely g= f↑i) such
that f* = g*, yet f != g.

The interesting case that remains is: what happens if f
is not containment-free, but there is no proper containment, i.e.,
f = f↑i for some i.
In this case, f is bounded, in the sense that f*
is equal to finite unions of powers of f.  However, some f's
have a g for which f* = g*, but f != g, and some do not.
The function PROJ←{2,3,1}, i.e., the function that cycles the
columns of a ternary relation, is an example of an f for which
g exists; g = PROJ←{3,1,2}.
On the other hand, the identity function f has no such g.

OPEN PROBLEMS:

It is interesting to suppose that functions f could be divided
into two classes: those that are containment-free and those that
are bounded, i.e., f = f↑i for some i>1.
If we can pick f's from a sufficiently rich class, that isn't true.
For example, Allen suggested the following function:
f(R) = I union PROJ←{1,3}(E*R)
where I is the set of pairs (x,x) such that x is a "node", and
f↑i is the set of (x,y) such that there is a path of edges E
from x to y, of length <= i.
Then f↑i C← f↑{i+1} for some i, in fact for any i, yet for no
i is f = f↑i.

Jeff N. also contributes the following example.  The recursion
is data dependent (therefore not bounded), there is only
one rule, yet the function f defined by the rule is not
containment-free.
	t(X,Y) :- t(X,W) & e(W,W) & e(W,Y)
If t←0 is the initial value of t, then
f(t←0) is defined by t(X,Y) :- t←0(X,W←0) & e(W←0,W←0) & e(W←0,Y)
while f↑2(t←0) is given by
	t(X,Y) :- t←0(X,W←1) & e(W←1,W←1) & e(W←1,W←0) & e(W←0,W←0) & e(W←0,Y)
(i.e., the project-join formulas for f and f↑2 are the ones
implied by these rules)
	It is easy to check that f C← f↑2 but not vice-versa, using
the idea of tableau mappings.

We conjecture that the dichotomy containment-free/bounded appears
when f is chosen from a sufficiently simple class, perhaps even
for all single-rule, linear logic programs with no repeated predicates.
Note Allen's example requires two rules to define R, and Jeff's
uses two occurrences of E and also uses variable repetition.

We also would like to see classes for which there is an efficient
test for the containment-free property, and classes for which
f = g (as mappings, not fixpoints) is easy.

Finally, we are not really especially interested in the equivalence
problems; we really want efficient procedures for deciding, given
f, what is the "smallest" g such that f* = g*.
Perhaps the theory suggested above can help with this optimization
problem, e.g., by showing that in many cases, the opt. question
reduces to "find a smallest g such that f = g."

∂12-Dec-85  1537	ullman@su-aimvax.arpa 	oops   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 12 Dec 85  15:36:18 PST
Received: by su-aimvax.arpa with Sendmail; Thu, 12 Dec 85 15:32:01 pst
Date: Thu, 12 Dec 85 15:32:01 pst
From: Jeff Ullman <ullman@diablo>
Subject: oops
To: nail@diablo

That message I sent out has at least two flaws, pointed out by Vlad
Lifshitz.

1. Minor point: I didn't mean to say that f* C← g* iff f C← g↑i
for some i applied for any f and g, but only for the case of
tableau mappings.  The result depends on the Sagiv-Yannakakis
theorem about when unions of tableau mappings are equivalent,
and is just one of the ways in which tableaux are special in useful ways.

2. More importantly, while I glibly talked about logic programs
as "functions," really they are parametized functions.
The "parameter" is the database.
Thus, given the simple transitive closure rules:
	t(X,Y) :- t(X,Z) & e(Z,Y)
we really have one function f←e for each database e.

The theorem I stated is still true about any f←e; that is,
if f←e is never a subset of (f←e)↑i, for i>1, then
(f←e)* = (g←e)* iff f←e = g←e.

However, the worrisome thing is that there might be some way
to optimize program f, say by replacing it by program g,
such that (f←e)* = (g←e)* for any e, yet it was not true
that f←e = g←e for the following reason:
	There were some e for which f←e is containment-free.
In those cases, f←e = g←e.  There were other e for which
f←e C← (f←e)↑i, and in those cases, f←e != g←e.
Thus, we could not optimize f by looking for g's for which
f←e = g←e.

Unfortunately, we cannot avoid this problem by defining "containment-free"
to mean that f←e is containment-free for all e; the problem
is that almost any nontrivial program will have SOME e for
which that is false, e.g., e = EMPTY SET means that th transitive
closure program does not have the c.f. property.

I'm not sure what to do.  What comes to mind is that when
we ask whether f←e = g←e, we do not want to ask whether that
is true for all e, but only for those e's satisfying a certain
condition (that e doesn't induce the a containment on the program of f).
That sounds like the old idea of equivalence of tableau mappings
on only those databases that satisfy certain dependencies,
and we were able to figure out how to do that, so perhaps we
can crack this one, but it is harder than it looked.
				---Jeff

∂13-Dec-85  0843	TAJNAI@SU-SCORE.ARPA 	Contacts at Apple 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  08:42:57 PST
Date: Fri 13 Dec 85 08:27:13-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: Contacts at Apple
To: Faculty@SU-SCORE.ARPA
cc: Fullerton@SU-SIERRA.ARPA
Message-ID: <12166807044.9.TAJNAI@SU-SCORE.ARPA>


We would like to recruit Apple into the Forum.  Do any of you
have contacts at Apple?  

Carolyn
-------

∂13-Dec-85  0936	MODICA@SU-SCORE.ARPA 	End Quarter Reports    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  09:36:17 PST
Date: Fri 13 Dec 85 09:30:41-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: End Quarter Reports
To: instructors@SU-SCORE.ARPA
Message-ID: <12166818598.26.MODICA@SU-SCORE.ARPA>

Hi again.
This is to remind you that I have to have ALL End Quarter Reports
by Tuesday December 17th (that's next Tuesday). I will walk
them to the Registrar's Office late in the afternoon, so as to
give everyone ample time to get them to me. You can drop them off
at my desk (MJH 030) or leave them in my mailbox (MODICA, top row),
and if at all possible try to avoid using the id mail to send them
to me.
Once again, don't take them to Victoria.
She does not want them.
I do.
Bye.
-Gina
-------

∂13-Dec-85  1019	MODICA@SU-SCORE.ARPA 	important correctio    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  10:18:54 PST
Date: Fri 13 Dec 85 10:14:40-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: important correctio
To: instructors@SU-SCORE.ARPA
Message-ID: <12166826606.26.MODICA@SU-SCORE.ARPA>

Hi.
The End Quarter Reports have to be at the Registrar's Office by NOON on
Tuesday. Sorry, my mistake.
Anyway, that means that I have to have them as early as possible Tuesday
morning.
Please remember this deadline.
And forget the other one.
It is wrong.
-Gina
-------

∂13-Dec-85  1141	ullman@su-aimvax.arpa 	oops again  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  11:41:19 PST
Received: by su-aimvax.arpa with Sendmail; Fri, 13 Dec 85 11:36:23 pst
Date: Fri, 13 Dec 85 11:36:23 pst
From: Jeff Ullman <ullman@diablo>
Subject: oops again
To: nail@diablo

Boy do I feel stupid.  I *think* I now understand what Allen
was trying to say.  As long as the function f←e is expressible
by a tableau, then the condition (ALL e) (f←e)* C← (g←e)*
is equivalent to (EXISTS i>1)(ALL e) f←e C← (g←e)↑i
	[e stands for an arbitrary "database"]
I thought the quantification went the other way.

As a result, we can prove (ALL e) (f←e)* = (g←e)* implies (ALL e) f←e = g←e
as follows, in the case that (EXISTS e) f←e is containment free.
	(EXISTS i,j>1)(ALL e) f←e C← (g←e)↑i and g←e C← (f←e)↑j
	thus (ALL e) f←e C← (f←e)↑{ij}, violating the c.f. condition.

There may still be one nasty: in simple cases like the transitive
closure, the e for which f←e is containment-free is infinite;
there cannot be a finite such e.  That may bother some people.

∂13-Dec-85  1422	ullman@su-aimvax.arpa 	papers received  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  14:22:13 PST
Received: by su-aimvax.arpa with Sendmail; Fri, 13 Dec 85 14:14:17 pst
Date: Fri, 13 Dec 85 14:14:17 pst
From: Jeff Ullman <ullman@diablo>
Subject: papers received
To: nail@diablo

"A model and an architecture for relational knowledge base"
by H. Yokota and H. Itoh of ICOT

"Deductive database system based on unit resolution"
by Yokota, Itoh and K. Sakai

∂13-Dec-85  1441	LANSKY@SRI-AI.ARPA 	revised PLANLUNCH schedule    
Received: from SRI-AI.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  14:41:48 PST
Date: Fri 13 Dec 85 14:36:54-PST
From: LANSKY@SRI-AI.ARPA
Subject: revised PLANLUNCH schedule
To: planlunch.dis: ;
cc: ladkin@KESTREL.ARPA

Due to various scheduling rearrangements, there will be NO MORE planlunches
this quarter.  We will be startup up again in January, with the following
tentative schedule:

January 8 (Wed) : James Allen
January 15 (Wed): Marcel Schoppers
January 20 (Mon): Dave Smith
January 27 (Mon): Peter Ladkin
  :
  ?
Once again, those of you who would like to give planlunch talks, please
contact me.

Happy Holidays,
Amy Lansky (LANSKY@SRI-AI)
-------

∂13-Dec-85  1609	@SU-SUSHI.ARPA,@SU-SCORE.ARPA:karp@ernie.berkeley.edu  
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  16:09:14 PST
Received: from SU-SCORE.ARPA by SU-SUSHI.ARPA with TCP; Fri 13 Dec 85 16:03:13-PST
Received: from ucbvax.berkeley.edu by SU-SCORE.ARPA with TCP; Fri 13 Dec 85 15:40:46-PST
Received: by ucbvax.berkeley.edu (5.31/1.7)
	id AA28415; Fri, 13 Dec 85 15:35:25 PST
Received: by ernie (5.31/5.16)
	id AA12160; Fri, 13 Dec 85 14:20:20 PST
Date: Fri, 13 Dec 85 14:20:20 PST
From: karp@ernie.berkeley.edu (Richard Karp)
Message-Id: <8512132220.AA12160@ernie>
To: aflb.local@su-score.arpa, allmsgs@ernie.berkeley.edu,
        csfaculty@ernie.berkeley.edu, csmsgs@ernie.berkeley.edu,
        theory-b@ernie.berkeley.edu, theory-f@ernie.berkeley.edu

MATHEMATICAL SCIENCES RESEARCH INSTITUTE
1000 Centennial Drive, Berkeley, Ca. 94720 (415)642-0143
SEMINAR IN COMPLEXITY
Tuesday, December 17  2:00  MSRI Lecture Hall
Daniel D. Sleator  "Tree Rotations, Dissections of Polyhedra and Hyperbolic
Geometry"

∂13-Dec-85  1804	@MIT-MC.ARPA:JF@SU-SUSHI.ARPA 	SDI Debate at Stanford, 12/19/85  
Received: from MIT-MC.ARPA by SU-AI.ARPA with TCP; 13 Dec 85  18:04:02 PST
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 13 DEC 85  21:04:21 EST
Received: from MIT-MC.ARPA by MIT-OZ via Chaosnet; 13 Dec 85 20:59-EST
Received: from SU-SUSHI.ARPA by MIT-MC.ARPA 13 Dec 85 21:02:13 EST
Date: Fri 13 Dec 85 17:56:01-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: SDI Debate at Stanford, 12/19/85
To: AILIST@SRI-AI.ARPA, ARMS-D@MIT-MC.ARPA, ARPANET-BBOARDS@MIT-MC.ARPA,
    EVOLUTION@KESTREL.ARPA, MsgGroup@BRL.ARPA, NA@SU-SCORE.ARPA,
    PHIL-SCI@MIT-MC.ARPA, POLI-SCI@RED.RUTGERS.EDU, PROLOG@SU-SCORE.ARPA
Message-ID: <12166910591.26.JF@SU-SUSHI.ARPA>

     APOLOGIES TO THOSE WHOSE BBOARDS RECEIVED MULTIPLE COPIES OF THIS.

←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

		``SDI: How Feasible, How Useful, How Robust?''

This will be a technical debate, covering both hardware and software aspects 
of SDI.

Sponsor: Stanford Computer Science Department

Date: December 19, 1985

Time: 8:00 p.m.

Place: Terman Auditorium

Organizer: Barbara Simons, IBM-SJ

Moderator:  Dr. Marvin L. Goldberger, President of Cal Tech.
Former member of President's Science Advisory Committee
and Consultant on Arms Control and International Security.

Panelists:

Advocates:
Professor Richard Lipton, Professor of Computer Science at Princeton
University, Current member of SDIO's Panel on Computing and Support of Battle
Management.

Major Simon Peter Warden, the Special Assistant to the Director of the SDIO
and Technical Advisor to the Nuclear and Space Arms Talk with the USSR
in Geneva.

Opponents:
Dr. Richard L. Garwin, IBM Fellow and Adjunct Professor of Physics at
Columbia University, Physicist and Defense Consultant.

Professor David Parnas, Lansdown Professor of Computer Science at the 
University of Victoria, Former member of the SDI Organization's
Panel on Computing and Support of Battle Management.



-------

∂14-Dec-85 JF@SU-SUSHI.ARPA	14-Dec-85 JMC	Sleator talk at SRC, Monday, 12/16/85, 2 p.m. 
Received: from SU-SUSHI.ARPA by SU-AI.ARPA with TCP; 14 Dec 85  14:53:13 PST
Date: Sat 14 Dec 85 14:49:25-PST
From: Joan Feigenbaum <JF@SU-SUSHI.ARPA>
Subject: Sleator talk at SRC, Monday, 12/16/85, 2 p.m.
To: aflb.all@SU-SUSHI.ARPA
cc: su-bboards@SU-SUSHI.ARPA
Message-ID: <12167138766.16.JF@SU-SUSHI.ARPA>

Monday, Dec. 16, 1985, DECSRC, 2:00 p.m.

Speaker: Dan Sleator, CMU

Title: "Tree Rotations, Dissections of Polyhedra and Hyperbolic Geometry"

Editorial Comment (jf):  

If I've got this straight, Danny is going to
present his joint work with Tarjan and Thurston on the ``rotation distance''
between two trees, which is, vaguely, the number of local restructuring
operations needed to transform one tree into another.  They extend previously
known combinatorial results to prove that 2m-6 in an upper bound on
the distance between a pair of m-node trees and then use hyperbolic
geometry to construct an infinite family of pairs for which this bound
is tight.  This work provides a rare and beautiful example of the use of
continuous methods to resolve a long-standing open problem in discrete 
mathematics, and I'm sure the talk will be entertaining and well worth your
time.


Please arrive at DECSRC, 130 Lytton Avenue, Palo Alto, at least 10 minutes
early so that you can be ``signed in''.
-------

jmc - Isn't that "Bud" Sleator?
∂14-Dec-85  1627	BRESNAN@SU-CSLI.ARPA 	interactions of morphology, syntax, and discourse    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 14 Dec 85  16:27:24 PST
Date: Sat 14 Dec 85 16:24:29-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: interactions of morphology, syntax, and discourse
To: folks@SU-CSLI.ARPA

This Thursday, December 19, at 10:00 a.m., Jonni Kanerva will present
his analysis of pronominal incorporation in Finnish possessives,
which shows that syntactically functioning pronouns can occur
within the lexical morphology of words, a result with
far-reaching consequences for the organization of grammars.  Everyone is
invited.  Copies of the written version of the presentation are
available at the front desk at Ventura Hall.
                ---------------

     A class of five morphemes in Finnish, traditionally called
possessive suffixes (henceforeward Px), raises interesting questions
about the relationship of morphological structure to syntactic
functions.  Px's appear to be pronominal, anaphoric, or even
agreement-like elements that occur on nominals and nonfinite verbs
following case suffixes.  They are important syntactically:  among
other things, they occur as possessors of nouns and as subjects of
nonfinite clauses.  The very importance of Px's in the syntax tempts
one to analyse them as syntactic units -- clitics -- that are joined
phonologically to host words, as two recent analyses have done.
Nonetheless, a number of facts in Finnish indicate that these
syntactic functions are born by truly morphological units --
suffixes -- rather than clitics.

     I argue from phonological, morphological, and semantic evidence.
First, any allomorphy or phonological alternation in Finnish that is
sensitive to word boundaries treats the undisputed suffixes and the
Px's alike as being inside the word and treats a class of clitics as
being outside the word.  Second, the occurrence of Px's is sometimes
dependent on the specific morphological structure of the stem.  Third,
a large number of semantically idiosyncratic lexical items containing
Px's provide further support for a suffixal analysis of Px's, insofar
as suffixes are more susceptible to idiosyncratic lexicalization than
clitics.  I argue next against the possibility that Px's are lexical
level clitics (i.e., clitics that attach to words at the morphological
level) by showing that it is quite costly to the theory of lexical
phonology to have a lexical level in Finnish that contains all of the
undisputed suffixes yet excludes the Px's; hence Px's must occupy the
same lexical level as other suffixes.  Considering, then, all of the
evidence favoring a suffixal analysis for the Px's, especially the
morphological interactions between Px's and their stems, it is
extremely weak to set Px's apart from the other suffixes solely on
the basis of morpheme order.  This indicates that the Px's are indeed
suffixes and therefore that a syntactic analysis of Px's should be
consistent with this finding.

--Jonni Kanerva

-------
-------

∂14-Dec-85  1633	ACUFF@SUMEX-AIM.ARPA 	Common Lisp Meeting Summary 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 14 Dec 85  16:33:46 PST
Date: Sat 14 Dec 85 16:31:55-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Common Lisp Meeting Summary
To: ksl-lispm@SUMEX-AIM.ARPA
Message-ID: <12167157426.54.ACUFF@SUMEX-AIM.ARPA>

   I've sent a summary of the Common Lisp meeting held in Boston
recently to Common-Lisp@Sumex, so it will be part of bboard:common-lisp.txt
and can be read with "bboard common-lisp" from the Exec or MM on sumex.

	-- Rich
-------

∂14-Dec-85  2249	vardi@su-aimvax.arpa 	Preliminary PODS program    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 14 Dec 85  22:48:53 PST
Received: by su-aimvax.arpa with Sendmail; Sat, 14 Dec 85 22:44:41 pst
Date: Sat, 14 Dec 85 22:44:41 pst
From: Moshe Vardi <vardi@diablo>
Subject: Preliminary PODS program
To: fagin@ibm-sj, nail@diablo

                                ACM SYMPOSIUM ON
                         PRINCIPLES OF DATABASE SYSTEMS

                  Cambridge, Massachusetts, March 24-26, 1986

                                    Program


Session 1   Chaired by Avi Silberschatz, University of Texas at Austin

  Magic Sets and Other Strange Ways To Implement Logic Programs
      Francois Bancilhon (Microelectronics and Computer Technology Corporation)
      David Maier (Oregon Graduate Center)
      Yehoshua Sagiv (Stanford University)
      Jeffrey D. Ullman (Stanford University)

  Convergence of Sideways Query Evaluation
      Foto Afrati (National Technical University of Athens)
      Christos Papadimitriou (Stanford University)
      George Papageorgiou (National Technical University of Athens)
      Athena Roussou (National Technical University of Athens)
      Yehoshua Sagiv (Stanford University)
      Jeffrey D. Ullman (Stanford University)

  On the Implementation of a Simple Class of Logic Queries
      Domenico Sacca (Microelectronics and Computer Technology Corporation)
      Carlo Zaniolo (Microelectronics and Computer Technology Corporation)


Session 2   Chaired by Hector Garcia-Molina, Princeton University

  A Theoretical Foundation of Multi-Level Concurrency Control
      Gerhard Weikum (Technische Hochschule Darmstadt)

  Deleting Completed Transactions
      Thanasis Hadzilacos (National Technical University of Athens)
      Mihalis Yannakakis (AT&T Bell Labs)

  Safety of Non-well-locked Transaction Systems
      Jianwen Su (University of Lowell)


Session 3   Chaired by Meral Ozsoyoglu, Case Western Reserve University

  A Calculus for Complex Objects
      Francois Bancilhon (Microelectronics and Computer Technology Corporation)
      Setras Khoshafian (Microelectronics and Computer Technology Corporation)

  Some Classes of Multilevel Relational Structures
      Dirk Van Gucht (Indiana University at Bloomington)
      Patrick C. Fischer (Vanderbilt University)

  Weak Temporal Relations
      Shashi K. Gadia (Texas Tech University)


Session 4   Chaired by Per-Ake Larson, University of Waterloo

  Rearranging Data to Maximize the Efficiency of Compression
      Frank Olken (Lawrence Berkeley Laboratory)

  Order Preserving Linear Hashing Using Dynamic Key Statistics
      John T. Robinson (Thomas J. Watson Research Center)

  Balanced Multidimensional Extendible Hash Tree
      Ekow J. Otoo (Carleton University)


Session 5   Chaired by Moshe Vardi, IBM Research Lab

  Negation by Failure in First-Order Queries
      Shamim A. Naqvi (AT&T Bell Labs)

  Positivism vs. Minimalism in Deductive Databases
      Nicole Bidiot (I.N.R.I.A.)
      Richard Hull (University of Southern California)

  The Extended Closed World Assumption and Its Relationship to Parallel 
  Circumscription
      M. Gelfond (University of Texas at El Paso)
      H. Przymusinska (University of Texas at El Paso)
      T. Przymusinska (University of Texas at El Paso)


Session 6   Chaired by Alberto Mendelzon, University of Toronto

  On the Properties and Characterization of Connection-trap-free Schemes
      Edward P. F. Chan (University of Alberta)

  One Flavor Assumption and Gamma-acyclicity for Universal Relation Views
      J. Biskup (Universitat Dortmund)
      H.H. Bruggeman (Universitat Dortmund)
      L. Schnetgoke (Universitat Dortmund)
      M. Kramer (FernUniversitat Hagen)

  The Equivalence of Tree Projections and Solving Queries
      Yehoshua Sagiv (Stanford University)
      Oded Shmueli (Technion-Israel Institute of Technology)


Session 7   Chaired by Ed Sciore, Boston University

  Unifying Functional and Multivalued Dependencies for Relational Database 
  Design
      Meral Ozsoyoglu (Case Western Reserve University)
      Li-Yan Yuan (Case Western Reserve University)

  On Finite FD-Acyclicity
      Yehoshua Sagiv (Stanford University)
      Oded Shmueli (Technion-Israel Institute of Technology)

  Alpha-acyclic Decompositions of Relational Database Schemes
      Detlev Ruland (Universitat Wurzburg)
      Dietmar Seipel (Universitat Wurzburg)


Session 8   Chaired by Ronald Fagin, IBM Research Lab

  Constant Time Maintenance or the Triumph of the fd
      Marc Graham (Georgia Institute of Technology)
      Ke Wang (Georgia Institute of Technology)

  Test Data for Relational Queries
      Heikki Mannila (University of Helsinki)
      Kari-Jouko Raiha (University of Tampere)

  A Model-Theoretic Approach to Updating Logical Databases
      Marianne Winslett (Stanford University)


Session 9   Chaired by Paris Kanellakis, Brown University

  Properties of Transactional Schemas
      Serge Abiteboul (I.N.R.I.A.)
      Victor Vianu (University of California, San Diego)

  Availability in Partitioned Replicated Databases
      Amr El Abbadi (Cornell University)
      Sam Toueg (Cornell University)


Session 10  Chaired by Francois Bancilhon, Microelectronics and Computer 
            Technology Corporation

  Data Independent Recursion in Deductive Databases
      Jeff Naughton (Stanford University)

  Incomplete Information and Default Reasoning
      Moshe Vardi (IBM Research Lab)

  Parallel Evalution of Recursive Rule Queries
      S. Cosmadakis (IBM Thomas J. Watson Research Center)
      Paris Kanellakis (Brown University)
-------

∂15-Dec-85  0948	DELAGI@SUMEX-AIM.ARPA 	[HAL%MIT-OZ@MIT-MC.ARPA: intro computer science course at Brandeis]
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 15 Dec 85  09:48:18 PST
Date: Sun 15 Dec 85 09:47:48-PST
From: Bruce Delagi <DELAGI@SUMEX-AIM.ARPA>
Subject: [HAL%MIT-OZ@MIT-MC.ARPA: intro computer science course at Brandeis]
To: aap@SUMEX-AIM.ARPA
Message-ID: <12167346003.15.DELAGI@SUMEX-AIM.ARPA>

fyi......./b
                ---------------

Return-Path: <@MIT-MC.ARPA:HAL@MIT-OZ>
Received: from MIT-MC.ARPA by SUMEX-AIM.ARPA with TCP; Sat 14 Dec 85 06:04:39-PST
Received: from MIT-OZ by MIT-MC.ARPA via Chaosnet; 14 DEC 85  08:25:33 EST
Date: Sat, 14 Dec 1985  08:22 EST
Message-ID: <HAL.12167035500.BABYL@MIT-OZ>
From: HAL%MIT-OZ@MIT-MC.ARPA
To:   scheme@MIT-MC.ARPA
Subject: intro computer science course at Brandeis


This past semester, Brandeis University changed its introductory
computer science subject to use "Structure and Interpretation of
Computer Programs" and Scheme (running on Texas Instruments PCs).

The following essay is by Harry Mairson, who was in charge of the
subject.


\magnification = 1200
\centerline
{\bf Structure and Interpretation of Computer Programs}
\centerline{Autumn Semester, 1985}
\medskip
\centerline{\sl by Harry Mairson}
\bigskip
\bigskip
\parskip = 5pt plus1pt minus .5pt

This year, the Brandeis computer science department introduced a new
introductory course for its prospective majors.  {\sl Structure and
Interpretation of Computer Programs}, a textbook and novel curriculum
under development at MIT for several years, replaced a standard course
teaching Pascal and modular programming.  I taught the course in
cooperation with Harold Abelson, one of the MIT professors who
developed the curriculum.  As the course comes to a conclusion this
term, I would like to make a few comments about its contents, and why
I believe it has been and will continue to be a success at Brandeis.

First of all, ``Structure and Interpretation of Computer Programs'' is
a sophisticated and high-powered course, presenting to students with
no presumed background in computer science an intensive introduction
to the subject of computation.  Among the subjects presented in this
course are: induction and recursion, lambda calculus, actor models,
tail recursion, orders of growth, abstraction mechanisms, data-driven
and procedure-driven (i.e., ``message passing'') computation, data
structures, substitution and environment models of computation, stream
processing, applicative- and normal-order evaluation, how interpreters
and compilers work, and logic programming.  All this as an {\sl
introduction} to computer science! I did not learn about many of these
subjects until I was in graduate school, and some I learned only while
teaching this innovative course.

Such an ambitious curriculum cannot succeed without a substantial
commitment of resources from both students and the University.  Students
worked 15 to 20 hours a week on this course, sometimes more.  The
computer science department bought 18 Texas Instruments Professional
Computers, with the latest version of the Scheme programming language that
was practically hot off the press; there were fewer than three students per
personal computer.  Hal Abelson gave a two-hour master class each week.
I gave two one-hour recitation lectures, plus interminable tutorials.
(I had no fixed office hours, tried to be in as much as possible, and
students badgered me with questions constantly, sometimes until early
morning...)  Three teaching assistants, Brandeis graduate students
Larry Bookman, Xiru Zhang, and Brandeis senior Moises Lejter ran
one-hour tutorials every week in groups of four to five students,
and worked as tutors in the laboratory.  They were unflagging in their
dedication and patience.  In addition, they were joined by three MIT
undergraduates, David Alcocer, Jos\'e Capo, and Scott Heeschen, providing
on-the-spot counseling and tutoring in the midst of programming crises.
The MIT students were great, but as I joked to them on their arrival,
``I'm delighted to have you help, and I can't wait to throw you out.''
Next year, Brandeis undergraduates will take their places.

The kids in the class knew every resource we could conceivably provide
them was there.  Given that total commitment, they showed beyond any
doubt their own commitment and willingness to work real hard and absorb
the material.

This course has something new and profound to say about computation and
learning.  With no time wasted, it immediately plunges into a discusssion
of the {\sl semantics} of computation, and treats almost peripherally
the questions of syntax.  The importance of this approach can best be
explained by considering the difference between learning a foreign
language (say, French) and a computer language.

Before learning my first word of French, I already understood the semantics
of the language, since ``meaning'' means the same thing in English as it
does in French, despite any Gallic claims to the contrary: the French
talk about tables, chairs, families, taxes, good food, vacations; they use
the same verbs, adjectives, and adverbs as we do, and essentially the same
notions of time.  So ``learning French'' meant learning a new grammar to
express the same ideas and thoughts I already knew how to express
in English.

The same cannot be said for learning a computer language, because there the
overriding questions of the {\sl d\'ebutant} are: what is the computer
doing, what {\sl can} it do, and what does my program {\sl mean?}
The grammar of any programming language is far simpler than that of
French; given enough compulsiveness, anyone can learn to make sure that all
the semicolons are in the right place, that for every {\tt BEGIN} there
is an {\tt END}, that left and right parentheses match.  These grammatical
rules are annoying, but not conceptually deep.

It happens that this course is taught using the Scheme language, a dialect
of a more famous programming language called Lisp.  But as the authors of 
the text write, ``After a short time we forget about syntactic details
of the language (because there are none), and get on with the real issues.''
The Scheme language is used because it expresses easily a wide range of
computational processes, but these processes, and not the language itself,
are the centerpiece of this course.

Such an emphasis, as well as many other aspects of the course, is in
the best tradition of liberal education.  The course is not simply a
compendium of ``current'' tricks of the programming trade that will 
doubtless change in the next six months, but presents a foundation from
which today's and tomorrow's issues and controversies in computer science
can be understood and put into perspective.  I expect this course will 
teach students to think for themselves.

Computer science often attracts the ire of specialists in other academic
fields, principally physicists and mathematicians, but philosophers and
just about everyone else too, because it seems so narrow and self-referential
that it doesn't relate to other fields of study.  This course goes a long
way to healing the wounds of that misconception, as well as teaching
students a healthy appreciation for the respective fields that are
not simply the ancestors of computer science, but its much needed partners
in the pursuit of knowledge and understanding.  That means I want my
students to take lots of math, physics, and philosophy courses, and know
what these subjects are about!

For example, no discussion of the meaning of language can take place outside
the tradition of philosophy and its profound contribution to the understanding
of language.  In the Scheme language, for example, the meaning of a simple
expression like {\tt (cons x y)} can be explored on many levels, none of them
artificial.  The mechanism used by the computer to evaluate this expression
can be implemented in several ways that are substantially different:
{\tt (cons x y)} could be interpreted as a memory structure, an integer,
or a procedural abstraction.  On the other hand, the expression has
precise meaning simply in its fixed use with respect to the other
constructs of the Scheme language.

I spent one lecture describing these two radically different points of
view, showing how the former view of ``meaning as implementation'' is the
direct descendant of logical atomism in the tradition of Bertrand Russell
and his followers, while the latter idea of ``meaning as use'' is a
natural consequence of the philosophy of language as proposed in the 
later writings of Ludwig Wittgenstein.  Similarly, in a discussion of the
so-called ``environment model'' of computation, the subject of denotation
is explored: we understand how classical problems of referential
transparency in language are resolved, and how conceptions of meaning and
time relate to each other.  The semantics of computation becomes a
controlled laboratory where we can discover more about the complex
relationship of meaning and language.

Apart from the usual arithmetic programming examples, the course draws on
problems and examples from mathematics and physics.  In one two-week problem
set, for example, the students implemented a video game called ``Lunar
Lander,'' where a spaceship with a fixed amount of fuel must be landed
under a gravitational force without crashing.  In doing so, they experimented
with a variety of landing strategies.  The principal goal of the problem set
is to teach the students how to use something called lambda expressions,
a programming language construct borrowed from mathematical logic.  But the
problem set also makes the students think about models of physical reality,
where computation is not simply {\sl sui generis} but an analog of the
physical world.  They implemented an optimal fuel-efficient landing
strategy, and I made them understand the ideas of calculus and elementary
physics that justifies calling the strategy optimal.

In other material developed in the course, methods of data structuring are
used to implement symbolic differentiation as in the differential calculus.
Stream processing provides a method for understanding integration, and
shows how the arithmetic of infinite power series can be automated.  The
connections of these methods to the Macsyma system of ``symbolic
mathematics,'' a computer system of great use to researchers in science
and mathematics, are no mere coincidences.

Finally and most profoundly, the material presented in this course teaches
a wonder and respect for computation.  It is true, as often repeated in that
hackneyed phrase, that computers only do what we tell them to.  Stated
otherwise, a computational process evolves in a deterministic fashion from
an initial state, subject to fixed and mechanical constraints on the nature
of that evolution.  But the potential richness, depth, and variety of that
evolution, as this course teaches, is truly mind-boggling.  In the 
introduction to their book, Harold Abelson and his co-author Gerald Jay
Sussman, also of MIT, write the following:

\bigskip

\item{}
We are about to study the idea of a {\sl computational process.}
Computational processes are abstract beings that inhabit computers.  As
they evolve, processes manipulate other abstract things called {\sl data.}
The evolution of a process is directed by a pattern of rules called a
{\sl program.}  People create programs to direct processes.  In effect,
we conjure the spirits of the computer with our spells.

\item{}
A computational process is indeed much like a sorcerer's idea of a spirit.
It cannot be seen or touched.  It is not composed of matter at all.  
However, it is very real.  It can perform intellectual work.  It can
answer questions.  It can affect the world by disbursing money at a bank
or by controlling a robot arm in a factory.  The programs we use to conjure
processes are like a sorcerer's spells.  They are carefully composed from
symbolic expressions in arcane and esoteric {\sl programming languages}
that prescribe the tasks that we want our processes to perform.

\item{}
A computational process, in a correctly working computer, executes programs
precisely and accurately.  Thus, like the sorcerer's apprentice, the
novice programmer must learn to understand and to anticipate the
consequences of his conjuring.

\bigskip

In the same way that a biochemist marvels at DNA as a code that directs
the amazing growth and development of living organisms, I marvel as I watch
computer programs provide the code for the creation and interaction of
Abelson and Sussman's computational ``spirits.'' Just as
a chemist sees the elementary chemical building blocks 
of nature interact, combine,
and recombine as she pours solutions together, the computer scientist
sees the rich and seemingly unpredictible interaction of computational
processes as a result of her procedural incantations.

For all those who have looked askance at the existence of computer science
in the university, Abelson and Sussman have written, ``Underlying our
approach to this subject is our conviction that `computer science' is
not a science and its significance has little to do with computers.
The computer revolution is a revolution in the way we think and in the way
we express what we think.''

I believe that nothing could be more exciting or more important at a
university than to understand the way we think and how we express those
thoughts.  This new introductory course comes face-to-face with these profound
issues, and brings its students into the heart of the consequent
intellectual debate, sometimes pushing them to within striking distance
of the frontiers of research.  I believe it will provide them with tools
to strike out on frontiers of their own, frontiers of personal discovery,
creativity, understanding and synthesizing of knowledge, driven on by
the powerful metaphors that the study of computational processes provide,
and by the personal intellectual success that this course provides them.

This new course demands a great commitment from its students and
teaching staff.  Students have five official ``contact hours'' per
week with the teaching staff, and many more on an informal basis.
Because of that great contact, I believe that this course has another
thing to say that is also profound, though perhaps peripheral to the
issues of computation.  The University is a place for teaching and
learning, not simply for the transmission of information.  Teaching
and learning reinforce the respect, encouragement, and love that any
society or university needs to flourish.  I hope this class says
loudly and clearly that students have a place in the crucial function
of this University, not as passive receivers of knowledge that spouts
from the end of a pedagogical assembly line, but as partners in an
important dialogue that defines what the University is.

I want the students in this class to have learned two things.  First,
how much there is that they don't know.  And second, that they can
learn anything they want.  I hope that they will never forget the
former, and always be inspired by the latter.  Every conceivable thing
was done in this course to provide a fertile and inspiring environment
for learning and discovery.  What could students possibly do under
such circumstances but mature, grow, and learn?

The lesson for the teaching staff in this course is not so different.
While Newton said he saw further because he stood on the shoulders of
giants, I have always preferred Nietzsche's remark that a teacher is
poorly repaid if one only remains a student.  

What a tremendous debt we teachers owe to those who nurtured us!  We
honor those who taught us by nurturing students, and the way we
express that nurturing is to teach students to grow and think for
themselves, so they don't need us any longer.  A friend of mine who is
a physician once said that the responsibility of a doctor is to love
your patients.  I believe above all that the responsibility of a
teacher is to love your students, to show them what is known, and to
inspire them to confront the unknown.  The territory is vast, and
because computer science is still so new, largely unexplored.  The
rewards are great for those who simply press forward.

\bye

-------

∂15-Dec-85  1854	PARSYM-Request@SUMEX-AIM.ARPA 	PARSYM Digest   V1 #46  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 15 Dec 85  18:54:02 PST
Date: 15 Dec 85 1838-PST
From: Moderator Byron Davies <PARSYM-REQUEST@SUMEX-AIM.ARPA>
Reply-to: PARSYM@SUMEX-AIM.ARPA
Subject: PARSYM Digest   V1 #46
To: PARSYM@SUMEX-AIM.ARPA


PARSYM Digest            Monday, 16 Dec 1985       Volume 1 : Issue 46

Today's Topics:

                     SEAI Parallel Computing Survey
      Seminar: A Parallel Algorithm for Depth First Search (MIT)

----------------------------------------------------------------------

Date: Thursday, 12 December 1985  09:40-PST
From: Carolyn Talcott <CLT at SU-AI.ARPA>
Subject: SEAI Parallel Computing Survey

I called the SEAI folks and got the following additional information:

    It was compiled by R. Miller by asking `experts in the field'
      (of which he is not one).
    It is current - last survey results received in August 85
    Length = 123 pages
    Can get 15% discount for university use
    It can be returned within 30 days for full refund

[Has anybody on PARSYM purchased the SEAI survey?  Are there any
surprises in it? -- BD]

------------------------------

Date: 12/11/85 10:50:41
From: LISA at MIT-MC.ARPA
Re:   Talk by Anderson

[Forwarded from the MIT-MC BBoard by SASW@MIT-MC]

                       DATE:  Friday, December 13, 1985
                           TIME:  1:00.....Lecture
                              PLACE:  NE43 - 512A

                 "A PARALLEL ALGORITHM FOR DEPTH FIRST SEARCH"

                               Richard Anderson
                                     MSRI

                                   Abstract:

A new parallel algorithm for constructing a depth first search tree in an
undirected graph will be described.  The algorithm is a P-RAM algorithm and
uses several probabilistic algorithms as sub-routines.

The run time of the algorithm is  2 sq.rt. log n.  This makes it an almost RNC
                                             eps.
algorithm, since the run time is less than  n     for any  eps.>0.

The standard sequential algorithm for depth first search can be shown to be
"inherently sequential", so this shows that substantial speed up for depth
first search is possible when a different approach is taken.

                                 David Shmoys
                                     Host

------------------------------

End of PARSYM Digest
********************

∂16-Dec-85  0844	RICHARDSON@SU-SCORE.ARPA 	Near West Campus   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 16 Dec 85  08:44:13 PST
Date: Mon 16 Dec 85 08:41:11-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Near West Campus
To: ac@SU-SCORE.ARPA, srstaff@SU-SCORE.ARPA
Message-ID: <12167596021.20.RICHARDSON@SU-SCORE.ARPA>


Today is the Near West Campus review which will take place in Durand 450
from 4:00 - 6:00 with Dean Gibbons in attendance.
-------

∂16-Dec-85  1057	DELANEY@SUMEX-AIM.ARPA 	Re: [HAL%MIT-OZ@MIT-MC.ARPA: intro computer science course at Brandeis]
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 16 Dec 85  10:55:16 PST
Date: Mon 16 Dec 85 10:54:29-PST
From: John R Delaney <DELANEY@SUMEX-AIM.ARPA>
Subject: Re: [HAL%MIT-OZ@MIT-MC.ARPA: intro computer science course at Brandeis]
To: DELAGI@SUMEX-AIM.ARPA
cc: aap@SUMEX-AIM.ARPA, DELANEY@SUMEX-AIM.ARPA
In-Reply-To: <12167346003.15.DELAGI@SUMEX-AIM.ARPA>
Message-ID: <12167620285.57.DELANEY@SUMEX-AIM.ARPA>

Quite inspirational, but what did the students who had to spend 15 to 20 or
more hours a week feel about the course? How much did they actually learn
from this course? Did they have time left to learn anything in the other
courses they were taking? I would like to hear the rest of the story some
day.

John (the cynic)
-------

∂16-Dec-85  1151	DIKRAN@SU-CSLI.ARPA 	New CSLI Informal Notes 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 16 Dec 85  11:51:34 PST
Date: Mon 16 Dec 85 11:45:44-PST
From: Dikran Karagueuzian <DIKRAN@SU-CSLI.ARPA>
Subject: New CSLI Informal Notes
To: folks@SU-CSLI.ARPA


                CSLI Informal Notes

Three new CSLI Informal Notes have been published recently.
Copies may be obtained from Suzi Parker in Ventura Hall. The
Notes are

  A Weak Logic of Knowledge and Belief: Epistemic
  and Doxastic Logic for the Yuppie Generation
  by David Israel (No. IN-CSLI-3),

  Analogy
  by Todd Davies (No. IN-CSLI-4),

  Left-associative Grammar and the Parser NEWCAT
  by Roland Hausser (No. IN-CSLI-5).

-------

∂16-Dec-85  1528	MODICA@SU-SCORE.ARPA 	Grades  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 16 Dec 85  15:27:46 PST
Date: Mon 16 Dec 85 15:23:05-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: Grades
To: instructors@SU-SCORE.ARPA
Message-ID: <12167669184.24.MODICA@SU-SCORE.ARPA>

This is very important!!!!!
I need your end quarter reports by 11:00 am tomorrow, so that I can
sort through them and get them to the registrar's by noon.
Once again, you can leave them on my desk or in my mail box on the
second floor.
Thanks.
-Gina
-------

∂16-Dec-85  1614	BRESNAN@SU-CSLI.ARPA 	talk by Fassi Fehri    
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 16 Dec 85  16:12:28 PST
Mail-From: BRESNAN created at 16-Dec-85 11:19:11
Date: Mon 16 Dec 85 11:19:11-PST
From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
Subject: talk by Fassi Fehri
To: linguists@SU-CSLI.ARPA
ReSent-Date: Mon 16 Dec 85 16:08:29-PST
ReSent-From: Joan Bresnan <BRESNAN@SU-CSLI.ARPA>
ReSent-To: folks@SU-CSLI.ARPA

This Wednesday, December 16 at 10:00 a.m. Fassi Fehri will
present his work on agreement in Arabic in the Ventura Conference
Room.  

           Agreement in Arabic, Binding and Extended Coherence

    We provide a fragment of a conceptual framework in which agreement
phenomena can be naturally characterised in correlation with grammatical
functions, and the appropriate well-formedness constraints on functional
structres would have as an effect to rule out agreement relations that 
are unlikely to occur in natural languages. More specifically, we assume
a taxonomy of grammatical functions distinguishing three classes: lexical
( = L) and nuclear ( = N) GFs ( Subj, Obj, Obl, ...), non-L but N GFs 
( Adjunct, Modifier ), and non-L non-N ( Top, Foc ... ). Non-L non-N 
grammatical functions are some times called discourse functions or DFs.
We argue that Coherence, whose initial essential role in KB was to ensure 
the duplication in the syntax of L-government relations, should be redefined
to extend to non-L N as well as non-L non-N grammatical functions.
We furthermore argue for a typology of agreement distinguishing 'grammatical'
agreement from 'anaphoric' agreement. GAgr is with L GFs, AAgr with non-L
Gfs. Our proposal is that what appears to be an anaphoric agreement marker 
is in fact an incorporated pronoun. The different types of agreement fall 
out as effects of our Extended Coherence Condition plus other independently 
motivated well-formedness conditions on functional structures.
-------
-------

∂16-Dec-85  1637	ullman@su-aimvax.arpa 	Request for open problems  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 16 Dec 85  16:33:48 PST
Received: by su-aimvax.arpa with Sendmail; Mon, 16 Dec 85 16:25:34 pst
Date: Mon, 16 Dec 85 16:25:34 pst
From: Jeff Ullman <ullman@diablo>
Subject: Request for open problems
To: nail@diablo

I got a letter from Jan Paredaens asking for a list of
open problems in the area of database theory, for publication
in EATCS.  If anybody has some suggestions, I'll be happy to
forward them, or you can write to Jan at:
	Universitaire Instelling Antwerpen
	Departement Wiskunde en Informatica
	Universiteitsplein 1 B-2610 Wilrijk
	BELGIUM

∂17-Dec-85  0759	CHOMICKI@RED.RUTGERS.EDU 	recent debate on f*=g* and containment-free functions 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  07:59:05 PST
Received: from RED.RUTGERS.EDU by su-aimvax.arpa with Sendmail; Tue, 17 Dec 85 07:54:37 pst
Date: 17 Dec 85 10:54:12 EST
From: Jan Chomicki <CHOMICKI@RED.RUTGERS.EDU>
Subject: recent debate on f*=g* and containment-free functions
To: nail@SU-AIMVAX.ARPA
Message-Id: <12167849611.74.CHOMICKI@RED.RUTGERS.EDU>

I happened to follow the last debate documented on NAIL and I have some 
problems with putting the results mentioned there together. 

1. The fact:

(*) (ALL e) (f←e)* C← (g←e)* iff (EXISTS i>1) (ALL e) f←e C←(g←e)↑i

implies with g←e=f←e :

(EXISTS i>1) (ALL e) f←e C← (f←e)↑i

which implies that there are no containment-free functions satisfying (*).
Or am I missing something?

2. That makes
sense as strongly typed rules (where all the variables are typed)
satisfy (*) and none of them is containment-free.

3.It is not quite clear how to prove (*) for larger classes of rules
and exactly when it holds. (*) does not hold for the non-linear 
transitive closure:

t(X,Y) :- t(X,Z) & t(Z,Y) 

Again take f←e=g←e, although the database e does not matter here.

4.Non-linear transitive closure is containment-free.

					Jan Chomicki
					Dept.of Computer Science
					Rutgers University
					New Brunswick, NJ 08903

-------

∂17-Dec-85  1153	MODICA@SU-SCORE.ARPA 	grades  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  11:50:57 PST
Date: Tue 17 Dec 85 11:38:11-PST
From: Gina Modica <MODICA@SU-SCORE.ARPA>
Subject: grades
To: instructors@SU-SCORE.ARPA
Message-ID: <12167890385.24.MODICA@SU-SCORE.ARPA>

I you are not going to be able to hand in your grades on time,
please be sure and post grades so that students can see them.
The reason for this is that the registrar will give them all an
asterix (since grades were not on time), and unless they can find
the grades, they will come to me or you in a panick.
Thank You.
-Gina
-------

∂17-Dec-85  1416	DAVIES@SUMEX-AIM.ARPA 	Wednesday Meeting
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  14:16:11 PST
Date: Tue, 17 Dec 1985  13:57 PST
Message-ID: <DAVIES.12167915816.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: Wednesday Meeting
cc:   Davies@SUMEX-AIM.ARPA

There will be a meeting tomorrow at 9:30 am to continue the discussion
of TINA.

	-- Byron

∂17-Dec-85  1600	SCHMIDT@SUMEX-AIM.ARPA 	Alice's Lisp Machine 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  16:00:48 PST
Date: Tue 17 Dec 85 15:57:27-PST
From: Christopher Schmidt <SCHMIDT@SUMEX-AIM.ARPA>
Subject: Alice's Lisp Machine
To: KSL-LispM@SUMEX-AIM.ARPA, KSL-Dolphins@SUMEX-AIM.ARPA
Message-ID: <12167937584.12.SCHMIDT@SUMEX-AIM.ARPA>

	There's a 16243 byte cover of "Alice's Restaurant" set in the
MIT-AI/LISPM world that's posted as AIList Digest V3 #187.  It's
awfully long, but I found it amusing enough to read all the way
through.  --Christopher
-------

∂17-Dec-85  1601	EMMA@SU-CSLI.ARPA 	Newsletter 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  16:01:15 PST
Date: Tue 17 Dec 85 15:55:32-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter
To: folks@SU-CSLI.ARPA
Tel:  497-3479



     Tomorrow's newsletter is the last for the year and should contain 
announcements for all events upto and including Friday, January 10.

    Please remember that the deadline for submission is tomorrow (December
18) at noon.

Many thanks,

Emma
-------

∂17-Dec-85  2126	ACUFF@SUMEX-AIM.ARPA 	I'll be away 23-Dec-85 to 6-Jan-86    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  17:05:09 PST
Date: Tue 17 Dec 85 17:03:37-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: I'll be away 23-Dec-85 to 6-Jan-86
To: sbARNES@SUMEX-AIM.ARPA, JORDAN@SUMEX-AIM.ARPA, ksl-lispm@SUMEX-AIM.ARPA,
    ssrg-systems-staff@SUMEX-AIM.ARPA
Message-ID: <12167949628.56.ACUFF@SUMEX-AIM.ARPA>

   I'll be gone over Christmas and New Years.  I'll still read my
mail, so feel free to contact me that way.  If an emergency comes up,
please contact James Rice if a Symbolics or TI machine is down, or an
SSRG Systems Staff person if there is a power or air conditioning
problem.

	-- Rich
-------

∂17-Dec-85  2154	FEHRI@SU-CSLI.ARPA  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  21:27:50 PST
Date: Tue 17 Dec 85 20:13:15-PST
From: Fassi fehri <FEHRI@SU-CSLI.ARPA>
To: folks@SU-CSLI.ARPA


Has anyone seen the counter for the Xerox machine?
-------

∂17-Dec-85  2206	ullman@su-aimvax.arpa 	messages to nail 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  21:54:58 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Dec 85 17:04:01 pst
Date: Tue, 17 Dec 85 17:04:01 pst
From: Jeff Ullman <ullman@diablo>
Subject: messages to nail
To: nail@diablo

I know there are many people on the nail list who are not
at Stanford.  You may not know that you can mail to the
whole list by sending to nail@diablo.
(diablo is an arpanet host, if your mail system cares)
If you are curious or worried who is on the list, you
can mail to mailer@diablo a message with SUBJECT list nail
and no body.  You must then do the same with subject list nail2
because for technical reasons we had to divide the list in two.
				---Jeff Ullman

∂17-Dec-85  2206	@SU-CSLI.ARPA:PENTLAND@SRI-AI.ARPA 	MACHINES LEARNING TO TALK... 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  21:28:13 PST
Received: from SRI-AI.ARPA by SU-CSLI.ARPA with TCP; Tue 17 Dec 85 21:00:19-PST
Date: Tue 17 Dec 85 21:01:13-PST
From: PENTLAND@SRI-AI.ARPA
Subject: MACHINES LEARNING TO TALK...
To: folks@SU-CSLI.ARPA

a very highly recommended talk about machines learning
phonological rules --- automatically, from a limited corpus.

 
A special seminar in place of Pixels and Predicates:

	NETTALK:  Teaching a Massively-Parallel Network to Talk

Who:            Terrence J. Sejnowski, Johns Hopkins
Where:          CSLI trailers 
When:           1:00pm - Wednesday, December 18, 1985

Abstract:

	Text to speech is a difficult problem for rule-based systems
because English pronunciation is highly context dependent and there are
many exceptions to phonological rules.  An alternative knowledge
representation for correspondences between letters and phonemes will be
described in which rules and exceptions are treated uniformly and can
be determined with a learning algorithm in a connectionist model.  The
architecture is a layered network of 400 simple processing units with
9,000 weights on the connections between the units.  The training
corpus is continuous informal speech transcribed from tape recordings. 
Following training on 1000 words from this corpus the network can
generalize to novel text.  Even though this network was not designed to
mimic human learning, the development of the network in some respects
resembles the early stages in human language acquisition.  Following
damage of the network by either removal of units or addition of random
values to the weights the performance of the network degraded
gracefully.  Issues which will be addressed include scaling of the
learning algorithm with the size of the problem, robustness of
learning to predicate order of the problem, and universality of
learning in connectionist models.
-------

∂17-Dec-85  2208	ullman@su-aimvax.arpa 	tomorrow's meeting    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  21:56:48 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Dec 85 17:09:10 pst
Date: Tue, 17 Dec 85 17:09:10 pst
From: Jeff Ullman <ullman@diablo>
Subject: tomorrow's meeting
To: nail@diablo

I'm going to try to lead a discussion of how we build on
what Kathy has done to actually write some capture rules.
(sorry for that split infinitive)
With luck, I'll have completed my roadmap by tomorrow
and can distribute it.  If not, I'll whine in public and hope
Kathy can bail me out.

There may be some other points of interest as well.
Allen has just shown me a neat NC proof for certain hard-looking
rules, and we might want to go over the hash that I made of
the "containment-free" business.
				---Jeff

∂17-Dec-85  2214	ullman@su-aimvax.arpa 	addressing the nail list   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  21:52:00 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Dec 85 20:18:13 pst
Date: Tue, 17 Dec 85 20:18:13 pst
From: Jeff Ullman <ullman@diablo>
Subject: addressing the nail list
To: nail@diablo

Arthur Keller informs me that the "official" name of diablo
in ARPA addressing tables is su-aimvax, and that you can't be guaranteed
of reaching the nail list (or me for that matter) unless you
use that host name.

∂17-Dec-85  2215	ullman@su-aimvax.arpa 	paper received   
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  21:53:19 PST
Received: by su-aimvax.arpa with Sendmail; Tue, 17 Dec 85 17:10:00 pst
Date: Tue, 17 Dec 85 17:10:00 pst
From: Jeff Ullman <ullman@diablo>
Subject: paper received
To: nail@diablo

"An Algebraic Approach to Recursive Inference"
by Y. Ioannidis and E. Wong, UCB/ERL M85/93, Berkeley.

∂17-Dec-85  2300	ark@SALLY.UTEXAS.EDU 	Re:  addressing the nail list    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 17 Dec 85  23:00:25 PST
Received: from sally.UTEXAS.EDU by su-aimvax.arpa with Sendmail; Tue, 17 Dec 85 22:57:06 pst
Date: Wed, 18 Dec 85 00:48:06 cst
From: ark@SALLY.UTEXAS.EDU (Arthur M. Keller)
Posted-Date: Wed, 18 Dec 85 00:48:06 cst
Message-Id: <8512180648.AA08361@sally.UTEXAS.EDU>
Received: by sally.UTEXAS.EDU (4.22/4.22)
	id AA08361; Wed, 18 Dec 85 00:48:06 cst
To: ullman@SU-AIMVAX.ARPA
Subject: Re:  addressing the nail list
Cc: nail@SU-AIMVAX.ARPA

Actually, the official name of diablo is "SU-AIMVAX.ARPA" until
Stanford adopts the new domain addressing standard.  Some hosts accept
"diablo" and "SU-AIMVAX" but not all do.  "nail@SU-AIMVAX.ARPA" is the
current best way to reach the nail! mailing list and
"ullman@SU-AIMVAX.ARPA" is a good way to reach Jeff Ullman.  Note that
SU-AIMVAX.ARPA insists on using "diablo" in its mail headers (in
violation of the relevant DDN standard), which may cause confusion on
replies from sites not accepting accepting the nickname "diablo".

Arthur

∂18-Dec-85  1302	YAMARONE@SU-CSLI.ARPA 	extra sandwiches 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Dec 85  13:01:06 PST
Date: Wed 18 Dec 85 12:57:25-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: extra sandwiches
To: folks@SU-CSLI.ARPA


Two turkey on light ryes left.....only $2.50.....

Listen to your stomach...come and get them while the last!

the Ventura Sandwich Corp.

-------

∂18-Dec-85  1428	EPPLEY@SU-SCORE.ARPA 	Day Off 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 18 Dec 85  14:28:17 PST
Date: Wed 18 Dec 85 14:13:29-PST
From: LaDonna Eppley <EPPLEY@SU-SCORE.ARPA>
Subject: Day Off
To: CSD@SU-SCORE.ARPA
Message-ID: <12168180800.25.EPPLEY@SU-SCORE.ARPA>


The University has officially given all staff Monday, December 23, as
a paid holiday.  This is, in addition, to December 24 and 25.  Have a
wonderful holiday.

LaDonna
-------

∂18-Dec-85  1537	TOM@SU-SCORE.ARPA 	Re: Day Off
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 18 Dec 85  15:37:14 PST
Date: Wed 18 Dec 85 15:27:34-PST
From: Thomas Dienstbier <TOM@SU-SCORE.ARPA>
Subject: Re: Day Off
To: EPPLEY@SU-SCORE.ARPA
cc: CSD@SU-SCORE.ARPA, TOM@SU-SCORE.ARPA
In-Reply-To: <12168180800.25.EPPLEY@SU-SCORE.ARPA>
Message-ID: <12168194287.21.TOM@SU-SCORE.ARPA>

What about thursday and friday to make it complete...

tom
-------

∂18-Dec-85  1659	TREITEL@SUMEX-AIM.ARPA 	date on 3600s   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 18 Dec 85  16:59:28 PST
Date: Wed 18 Dec 85 16:56:14-PST
From: Richard Treitel <TREITEL@SUMEX-AIM.ARPA>
Subject: date on 3600s
To: mjh-lispm@SU-AI.ARPA
cc: treitel@SUMEX-AIM.ARPA
Message-ID: <12168210429.31.TREITEL@SUMEX-AIM.ARPA>

For some unknown reason, when we brought our 3600's up today, all five of them
believed it was November 18th 1985 (they had the time of day right, at least!).
I have manually reset four of them to December, but was unable to get to Coax
and change its date.   So whoever next has the chance, please ...

			- Richard
-------

∂18-Dec-85  1727	@SU-SCORE.ARPA:welch@ames-vmsb.ARPA 	SIGBIG  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 18 Dec 85  17:27:16 PST
Received: from ames-vmsb.ARPA by SU-SCORE.ARPA with TCP; Wed 18 Dec 85 17:21:47-PST
Date: 18 Dec 85 16:50:00 PST
From: welch@ames-vmsb.ARPA
Subject: SIGBIG
To: super@su-score.arpa
Reply-To: welch@ames-vmsb.ARPA


               ASSOCIATION FOR COMPUTING MACHINERY
                San Francisco Golden Gate Chapter
               "SIGBIG" Special Interest Committee
                 For Large High Speed Computers

 Meetings on  the first Wednesday of each month at 7:30 PM.   Speakers 
 who  can give insights to various aspects of  SUPERCOMPUTING are 
 featured each month.

 Next meeting:     Wednesday, January 8,1986,  7:30 PM

     Speaker:      Bob Hedges / Elxsi

     Topic:        Elxsi Systems and Supercomputing

     Location:     AXIOM Systems
		   1589 Centre Pointe Drive
		   Milpitas, CA 

     Directions: 17 South to Montague Expressway East. Left from
		 Montague onto Centre Point (before Capitol).
	or	 17 South to Capitol Expressway East. Right from
		 Capitol onto Centre Point (before Montague).
	or	 680 South to Montague Expressway West. Right from
		 Montague onto Centre Point (after Capitol).

 ---------------------------------------------------------------
 Tape-recordings  of  most of the previous  may  be obtained
 in exchange for a tape cassette or $5.00 by contacting: 
                Mary Fowler (415)261-4058 (rec)
                Supercomputing  #192, BOX 2787
                Alameda, CA. 94501-0787

 For information contact Mary Fowler, Chairperson (415) 839-6547
                     or  Mike Austin, Publ. Chair (415) 423-8446

------

∂18-Dec-85  2105	JODY@SU-CSLI.ARPA 	Burrito Bandito 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 18 Dec 85  21:05:39 PST
Date: Wed 18 Dec 85 21:01:59-PST
From: Joe Zingheim <JODY@SU-CSLI.ARPA>
Subject: Burrito Bandito
To: folks@SU-CSLI.ARPA

Friends of the Burrito Bandito will be happy to hear that he has once again
eluded the Border Patrol and the INS, and will once again visit the Center
for one (perhaps last) time.  Tacos, soft shelled, and the huge burritos will
be available, plus, perhaps, tamales (a south of the border season treat).

Sorry for the lack of advanced notice about this, but the sanctuary movement
is understandably unpredictable.  With these you may not need the pillow to
augment your Santa Claus costume.

Further notice tomorrow morning with options for the un-initiated.

-------

∂19-Dec-85  1129	JODY@SU-CSLI.ARPA 	Burrito Bandito's visit   
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  11:27:21 PST
Date: Thu 19 Dec 85 10:55:37-PST
From: Joe Zingheim <JODY@SU-CSLI.ARPA>
Subject: Burrito Bandito's visit
To: folks@SU-CSLI.ARPA

Burritos and tacos are available with:
	Chile Verde with rice
	Chile Colorado with beans
	Chicken with rice
	Tongue with rice
	Carnitas with beans
	Picadillo with beans
	Chile Relleno with rice
	Carne Asada with beans

Regular burritos cost $3.00
Special burritos (with sour cream, avacado, and cheese) $3.75
Tacos are $1.75

Send your request by replying to this, or send me mail -- Jody
-------

∂19-Dec-85  1204	@MCC.ARPA:francois%mcc-db.ARPA@mcc.ARPA 	list nail
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  12:04:42 PST
Received: from MCC.ARPA by su-aimvax.arpa with Sendmail; Thu, 19 Dec 85 11:59:40 pst
Received: from mcc-db.ARPA by MCC.ARPA with TCP; Thu 19 Dec 85 13:59:04-CST
Date: Thu, 19 Dec 85 13:59:36 cst
From: Francois Bancilhon <francois%mcc-db.ARPA@mcc.ARPA>
Posted-Date: Thu, 19 Dec 85 13:59:36 cst
Message-Id: <8512191959.AA07225@mcc-db.ARPA>
Received: by mcc-db.ARPA (4.12/4.22) 
	id AA07225; Thu, 19 Dec 85 13:59:36 cst
To: nail%diablo@mcc.ARPA
Subject: list nail


∂19-Dec-85  1205	@MCC.ARPA:francois%mcc-db.ARPA@mcc.ARPA 	list nail2    
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  12:05:31 PST
Received: from MCC.ARPA by su-aimvax.arpa with Sendmail; Thu, 19 Dec 85 12:01:20 pst
Received: from mcc-db.ARPA by MCC.ARPA with TCP; Thu 19 Dec 85 13:59:34-CST
Date: Thu, 19 Dec 85 14:00:04 cst
From: Francois Bancilhon <francois%mcc-db.ARPA@mcc.ARPA>
Posted-Date: Thu, 19 Dec 85 14:00:04 cst
Message-Id: <8512192000.AA07243@mcc-db.ARPA>
Received: by mcc-db.ARPA (4.12/4.22) 
	id AA07243; Thu, 19 Dec 85 14:00:04 cst
To: nail%diablo@mcc.ARPA
Subject: list nail2


∂19-Dec-85  1233	EM@SU-SCORE.ARPA 	Common Lisp - Graphics.    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  12:33:30 PST
Date: Thu 19 Dec 85 12:30:00-PST
From: Eric Muller <EM@SU-SCORE.ARPA>
Subject: Common Lisp - Graphics.
To: mjh-lispm@SU-AI.ARPA
Message-ID: <12168424106.18.EM@SU-SCORE.ARPA>


Two questions to our expert community.

1. Do I pay a penalty by writting Common Lisp code insteas of Zeta Lisp ?
  More precisely, are there CL constructs that do not translate well
in ZL, either because it is not possible or because the Symbolics CL
do not implement them efficiently right now ? Are there some typical
schemas that are well coded in CL style, do not execute efficiently
now, and can be coded in a more efficient way (even if this means
calling some ZL functions, provided that I can hide this and not
sacrifice the CL style) ?

2. I would like to make something similar to the data inspector;
instead of inspecting lisp objects, it would inspect logic formulas. In
this case, the mouse sensitive areas would be the subexpressions. The
user must have the ability to designate subexpressions, and then to
activate some functions that use these subexpressions and produce new
formulas. Of course, one need some form of scrolling because there are
more formulas than the screen can hold; also, it would be nice if the
set of formula to be displayed could be variable (i.e. the user
selects which formulas he wants to see in the window). Where must I
look to know if I can do something simple (I had a look at vol 7 of
the documentation, but it is not clear how I can implement the
selection mechanism I want) ? Is it possible to look at the code of
the data inspector (and if so, where is it) ? Am I trying to reinvent
the moon ?

thanks very much for your help,
eric.

p.s. Should I send this message to KSL-LispM as well ?
-------

∂19-Dec-85  1533	EMMA@SU-CSLI.ARPA 	Newsletter December 19, No. 6  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  15:33:26 PST
Date: Thu 19 Dec 85 15:11:49-PST
From: Emma Pease <Emma@SU-CSLI.ARPA>
Subject: Newsletter December 19, No. 6
To: friends@SU-CSLI.ARPA
Tel:  497-3479


!
                      C S L I   N E W S L E T T E R
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
December 19, 1985               Stanford                        Vol. 3, No. 6
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                                
     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ←←←←←←←←←←←←
              CSLI ACTIVITIES FOR THURSDAY, January 9, 1986

   12 noon		TINLunch
     Ventura Hall       Some Remarks on How Words Mean
     Conference Room    by Georgia Green
			Discussion led by Susan Stucky
			(Abstract will be in the next newsletter)
			
   2:15 p.m.		CSLI Seminar
			Whither CSLI? II
			John Perry, CSLI Director

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
			None planned
                             --------------
                              ANNOUNCEMENT

   Please note that the seminar and colloquium are no longer in Redwood
   Hall room G-19.  The new location will be announced in early January.
   Please also note that no activities are planned for December 19 and
   26.  Activities and the newsletter will resume January 9.
                             --------------
                            TALK ON JANUARY 2 

   On Thursday, January 2 at 2:15, Alexis Manaster-Ramer will discuss
   ``Finding Natural Languages a Home in Formal Language Theory''
   (coauthored with William C. Rounds and Joyce Friedman, abstract to be
   distributed later).  The talk will be in Ventura Hall.
                             --------------
                    TENTATIVE WINTER QUARTER SCHEDULE

   THURSDAY SEMINARS:

   Date			Speaker or Organizer

   January 9		John Perry
   January 16		Embedded Computation: Research on Situated Automata
   January 23		Embedded Computation: Semantically Rational Computer
			Languages
   January 30		Helene Kirchner
   February 6		Embedded Computation: Representation and Reasoning
   February 13		Semantics of Computational Languages
   February 20		Carol Cleland
   February 27		Linguistic Approaches to Computer Languages
   March 6		Mats Rooth

!
Page 2                     CSLI Newsletter                  December 19, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←

   THURSDAY COLLOQUIA: (one colloquium during each period)

   Time				Organizers

   January 9 to January 23:	Embedded Computation group and
   				Document Preparation group

   January 30 to February 13:	Embedded Computation group,
   				Computational Language Semantics group and
   				Linguistic Approaches to Computer Languages
				group

                             --------------
            INTERACTIONS OF MORPHOLOGY, SYNTAX, AND DISCOURSE
             Pronominal Incorporation in Finnish Possessives
                   Jonni Kanerva, (jkanerva@csli.arpa)
          Thursday, December 19, 10:00, Ventura Conference Room

      A class of five morphemes in Finnish, traditionally called
   possessive suffixes (henceforeward Px), raises interesting questions
   about the relationship of morphological structure to syntactic
   functions.  Px's appear to be pronominal, anaphoric, or even
   agreement-like elements that occur on nominals and nonfinite verbs
   following case suffixes.  They are important syntactically: among
   other things, they occur as possessors of nouns and as subjects of
   nonfinite clauses.  The very importance of Px's in the syntax tempts
   one to analyse them as syntactic units---clitics---that are joined
   phonologically to host words, as two recent analyses have done.
   Nonetheless, a number of facts in Finnish indicate that these
   syntactic functions are born by truly morphological
   units---suffixes---rather than clitics.
      I argue from phonological, morphological, and semantic evidence.
   First, any allomorphy or phonological alternation in Finnish that is
   sensitive to word boundaries treats the undisputed suffixes and the
   Px's alike as being inside the word and treats a class of clitics as
   being outside the word.  Second, the occurrence of Px's is sometimes
   dependent on the specific morphological structure of the stem.  Third,
   a large number of semantically idiosyncratic lexical items containing
   Px's provide further support for a suffixal analysis of Px's, insofar
   as suffixes are more susceptible to idiosyncratic lexicalization than
   clitics.  I argue next against the possibility that Px's are lexical
   level clitics (i.e., clitics that attach to words at the morphological
   level) by showing that it is quite costly to the theory of lexical
   phonology to have a lexical level in Finnish that contains all of the
   undisputed suffixes yet excludes the Px's; hence Px's must occupy the
   same lexical level as other suffixes.  Considering, then, all of the
   evidence favoring a suffixal analysis for the Px's, especially the
   morphological interactions between Px's and their stems, it is
   extremely weak to set Px's apart from the other suffixes solely on the
   basis of morpheme order.  This indicates that the Px's are indeed
   suffixes and therefore that a syntactic analysis of Px's should be
   consistent with this finding.
!
Page 3                     CSLI Newsletter                  December 19, 1985
←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←←
                         SUMMARY OF A CSLI TALK
           Agreement in Arabic, Binding and Extended Coherence
                Abdelkader Fassi Fehri, (Fehri@csli.arpa)

      We provide a fragment of a conceptual framework in which agreement
   phenomena can be naturally characterised in correlation with
   grammatical functions, and the appropriate well-formedness constraints
   on functional structres would have as an effect to rule out agreement
   relations that are unlikely to occur in natural languages. More
   specifically, we assume a taxonomy of grammatical functions
   distinguishing three classes: lexical and nuclear grammatical
   functions (Subj, Obj, Obl, ...), non-lexical but nuclear grammatical
   functions (Adjunct, Modifier), and non-lexical non-nuclear (Top, Foc,
   ...). Non-lexical non-nuclear grammatical functions are some times
   called discourse functions or DFs.  We argue that Coherence, whose
   initial essential role in KB was to ensure the duplication in the
   syntax of lexical-government relations, should be redefined to extend
   to non-lexical nuclear as well as non-lexical non-nuclear grammatical
   functions.  We furthermore argue for a typology of agreement
   distinguishing `grammatical' agreement (GAgr) from `anaphoric'
   agreement (AAgr).  GAgr is with lexical grammatical functions, AAgr
   with non-lexical grammatical functions.  Our proposal is that what
   appears to be an anaphoric agreement marker is in fact an incorporated
   pronoun. The different types of agreement fall out as effects of our
   Extended Coherence Condition plus other independently motivated
   well-formedness conditions on functional structures.
-------

∂19-Dec-85  1601	TAJNAI@SU-SCORE.ARPA 	Happy Holidays!   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  16:01:17 PST
Date: Thu 19 Dec 85 15:55:07-PST
From: Carolyn Tajnai <TAJNAI@SU-SCORE.ARPA>
Subject: Happy Holidays!
To: CSD@SU-SCORE.ARPA
Message-ID: <12168461448.44.TAJNAI@SU-SCORE.ARPA>


                  *	       
                  ↑	      	      With warm greetings and
		 ↑↑↑		      renewed hopes for
		↑*↑↑↑		      peace, contentment and joy in
 	      ↑*↑↑↑↑↑↑*		      the coming year.
	     ↑↑↑↑↑↑↑↑↑↑↑	 
	    *↑↑↑↑↑*↑↑↑↑↑↑	      
	   ↑↑↑*↑↑↑↑↑↑*↑↑↑↑	             Carolyn and Ann
	  ↑↑↑↑↑↑*↑↑↑↑↑↑↑↑↑↑
	 ↑↑↑↑↑*↑↑↑↑↑↑↑*↑↑↑↑↑			     and
	↑↑↑↑↑↑↑↑↑↑↑*↑↑↑↑↑↑*↑↑			  
       ↑↑↑*↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑*			  the Forum
      ↑↑↑↑↑↑↑↑↑↑*↑↑↑↑↑↑↑↑↑↑↑↑↑↑
     ↑*↑↑↑↑↑↑↑↑↑↑↑*↑↑↑↑*↑↑↑↑↑↑↑↑	
    ↑↑↑↑↑↑↑↑*↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑*↑↑↑↑
   ↑↑↑↑↑*↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑*↑↑↑↑↑↑↑↑
  ↑↑↑*↑↑↑↑↑↑↑↑*↑↑↑↑↑↑↑*↑↑↑↑↑↑↑↑*↑↑↑
 ↑*↑↑↑↑↑*↑↑↑↑↑↑↑↑*↑↑↑↑↑↑↑↑↑↑*↑↑↑↑↑↑↑    
	     /\/\/\/\/\			
	     /\/\/\/\/\
	     /\/\/\/\/\		    


-------

∂19-Dec-85  1730	YAMARONE@SU-CSLI.ARPA 	No Lunch Friday, 20th 
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  17:30:11 PST
Date: Thu 19 Dec 85 17:26:00-PST
From: Tom Yamarone <YAMARONE@SU-CSLI.ARPA>
Subject: No Lunch Friday, 20th
To: folks@SU-CSLI.ARPA

The Ventura Sandwich Corp. is on vacation......

Service will resume Monday, December 30th.

Happy Holidays and for those who use this term,

    MM           MM  EEEEEEEEE  RRRRRRR     RRRRRRR      Y            Y
    M  M       M  M  E          R      R    R      R       Y        Y
    M    M   M    M  E          R       R   R       R        Y    Y  
    M     M M     M  E          R      R    R      R           YY
    M      M      M  EEEEEE     RRRRRRR     RRRRRRR            YY
    M             M  E          R   R       R   R              YY
    M             M  E          R    R      R    R             YY
    M             M  E          R     R     R     R            YY
    M             M  EEEEEEEEE  R      R    R      R           YY


 CCCC  H      H  RRRRR   III    SSSS   TTTTTTTTT MM      MM      A      SSSS
C    C H      H  R    R   I   S     S      T     M M    M M     A A   S     S
C      H      H  R    R   I   S            T     M  M  M  M    A   A  S     
C      HHHHHHHH  RRRRR    I    SSSS        T     M   MM   M   AAAAAAA  SSSS
C      H      H  R R      I         S      T     M        M   A     A       S
C    C H      H  R  R     I   S     S      T     M        M   A     A S     S
 CCCC  H      H  R   R    I     SSSS       T     M        M   A     A   SSSS

-------

∂19-Dec-85  1835	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	intellectual property contract  
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 19 Dec 85  18:35:46 PST
Received: from su-aimvax.arpa by SU-SCORE.ARPA with TCP; Thu 19 Dec 85 18:31:45-PST
Received: by su-aimvax.arpa with Sendmail; Thu, 19 Dec 85 18:34:34 pst
Date: Thu, 19 Dec 85 18:34:34 pst
From: Jeff Ullman <ullman@diablo>
Subject: intellectual property contract
To: faculty@score

I've been talking with various people connected with the
Committee on Research, on which I sit, about the department's
concerns over the new contract.  I know a few people have mentioned
some concrete problems, either with the contract or with the
policy that it appears to enforce.  Can any of you who have
concerns please refresh my memory?  Chapter and verse where possible.
				---Jeff

∂20-Dec-85  1154	SCHOLZ@SU-SUSHI.ARPA 	database seminar  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 20 Dec 85  11:50:32 PST
Received: from SU-SUSHI.ARPA by su-aimvax.arpa with Sendmail; Fri, 20 Dec 85 11:25:30 pst
Date: Fri 20 Dec 85 11:22:21-PST
From: Karin Scholz <SCHOLZ@SU-SUSHI.ARPA>
Subject: database seminar
To: wiederhold@SUMEX-AIM.ARPA, nail@SU-AIMVAX.ARPA, mugs@SUMEX-AIM.ARPA,
        winslett@SU-SCORE.ARPA
Cc: milton@SRI-KL.ARPA
Message-Id: <12168673934.16.SCHOLZ@SU-SUSHI.ARPA>


since jack milton has graciously volunteered to maintain the mailing list
for the seminar, it should be a simple matter to schedule the speakers.
marianne winslett and i will be helping to coordinate the scheduling.
the current plan is to have each of the three involved research groups
deliver three seminars each.  it could make the seminar series more
logically continuous to have each group scheduled for three consecutive
weeks, but it might give people a greater understanding of each other's wor
earlier on if we interleave the groups.
i propose the following schedule:
jan 10  gio
jan 17  jeff
jan 24  mike
jan 31  gio
  .
  .
  .

the groups should decide among themselves who will be speaking on which
dates.
speaker and title should be sent to contreras@score by the monday morning
of the week before the seminar.
milton@sri needs an abstract 1 week in advance.

i  am far away and won't be back until mickey's big hand  is on the 86,
so if people have suggestions or objections to this schedule, please
work it out with marianne (sorry marianne).
the first week's speaker should send notice to contreras@score as soon
as she is committed.  likewise, the first nail group speaker (how bout
you, naughton?) will need to send notice by jan 6.

thanks everyone.
karin
-------

∂20-Dec-85  1158	@SU-SCORE.ARPA:FEIGENBAUM@SUMEX-AIM.ARPA 	Re: intellectual property contract    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 20 Dec 85  11:58:24 PST
Received: from SUMEX-AIM.ARPA by SU-SCORE.ARPA with TCP; Fri 20 Dec 85 08:43:25-PST
Date: Fri 20 Dec 85 08:46:09-PST
From: Edward Feigenbaum <FEIGENBAUM@SUMEX-AIM.ARPA>
Subject: Re: intellectual property contract
To: ullman@SU-AIMVAX.ARPA, faculty@SU-SCORE.ARPA
In-Reply-To: Message from "Jeff Ullman <ullman@diablo>" of Thu 19 Dec 85 19:29:08-PST
Message-ID: <12168645500.30.FEIGENBAUM@SUMEX-AIM.ARPA>

Thanks, Jeff. I've been oblivious to what has been going on re the
intellectual property contract issue. I would appreciate being copied
on the responses to Jeff's message (so I can be an informed signer or
non-signer).

Ed F.
-------

∂20-Dec-85  1200	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	GMD visit   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 20 Dec 85  11:59:53 PST
Received: from su-aimvax.arpa by SU-SCORE.ARPA with TCP; Fri 20 Dec 85 09:57:06-PST
Received: by su-aimvax.arpa with Sendmail; Fri, 20 Dec 85 09:59:56 pst
Date: Fri, 20 Dec 85 09:59:56 pst
From: Jeff Ullman <ullman@diablo>
Subject: GMD visit
To: faculty@score

We're still expecting a visit from the GMD folks on Jan. 8.
Our plan is to have brief (say 20-30 minute) presentations
from faculty who might be interested in participating in the
institute.  There will be technical people along on the
visit, so the presentation can be geared accordingly.

Apparently, they are most interested in new initiatives,
rather than supporting preexisting projects, and they
appear to be interested in interdisciplinary (or really inter-sub-disciplinary
work within CS) projects.  Thus, two or more people might wish
to team up on a proposal.

I have so far heard expressions of interest from Gio, Vaughan,
Nils,  Terry, and Ernst.  I hope I'm not forgetting people,
but if I am, please let me know.  Also, if you want a time slot,
please arrange one with Anne Richardson (richardson@score),
who is going to keep the schedule.
				---Jeff

∂20-Dec-85  1425	RICHARDSON@SU-SCORE.ARPA 	Closing Early 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 20 Dec 85  14:25:42 PST
Date: Fri 20 Dec 85 14:17:20-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Closing Early
To: csd@SU-SCORE.ARPA
Message-ID: <12168705791.10.RICHARDSON@SU-SCORE.ARPA>

Betty Scott asked me to let you know that CSD offices will close at 
4:00 p.m. today.
-------

∂20-Dec-85  1610	ullman@su-aimvax.arpa 	paper available  
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 20 Dec 85  16:10:17 PST
Received: by su-aimvax.arpa with Sendmail; Fri, 20 Dec 85 16:06:23 pst
Date: Fri, 20 Dec 85 16:06:23 pst
From: Jeff Ullman <ullman@diablo>
Subject: paper available
To: nail@diablo

Reprints of my recent TODS paper "Implementation of Logical
Query Languages for Databases" have been received.
If anybody at Stanford wants one, come by.
If someone off campus wants one, please write send mail to
rfn@sail, or as Arthur Keller would say: rfn@SU-AI.ARPA.

∂22-Dec-85  1505	ACUFF@SUMEX-AIM.ARPA 	Re: Common Lisp - Graphics. 
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 22 Dec 85  15:02:35 PST
Date: Sun 22 Dec 85 15:02:21-PST
From: Richard Acuff <Acuff@SUMEX-AIM.ARPA>
Subject: Re: Common Lisp - Graphics.
To: EM@SU-SCORE.ARPA, mjh-lispm@SU-AI.ARPA
In-Reply-To: <12168424106.18.EM@SU-SCORE.ARPA>
Message-ID: <12169238274.30.ACUFF@SUMEX-AIM.ARPA>

   I believe there are very few local expert users of Common Lisp.
However, you should be safe using Symbolics Common Lisp.  As always,
implement whatever in the cleanest, way you can, and then, if
necessary, try to hack it for efficiency.  Since there is no graphics
in Common Lisp, you will have to resort to ZetaLisp (or rather
"Symbolics extentions to Common Lisp") to accomplish your goals.

   You can look at the code of the Inspector by doing M-. on INSPECT.

	-- Rich
-------

∂23-Dec-85  1605	GINN@SUMEX-AIM.ARPA 	Please   
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 23 Dec 85  16:04:43 PST
Date: Mon 23 Dec 85 16:04:34-PST
From: Michael Ginn <GINN@SUMEX-AIM.ARPA>
Subject: Please
To: mjh-lispm@SU-AI.ARPA
Message-ID: <12169511743.11.GINN@SUMEX-AIM.ARPA>


     Please remove me from this list.  I do NOT want it to go to my
e-mail file.

---Michael
-------

∂24-Dec-85  0828	@SUMEX-AIM.ARPA:ihnp4!kddlab!nttlab!NTT-20!Okuno@seismo.CSS.GOV 	Greeting  
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 24 Dec 85  08:26:03 PST
Received: from seismo.CSS.GOV by SUMEX-AIM.ARPA with TCP; Tue 24 Dec 85 08:25:18-PST
Return-Path: <ihnp4!kddlab!nttlab!NTT-20!Okuno>
Received: from cbosgd.UUCP by seismo.CSS.GOV with UUCP; Tue, 24 Dec 85 11:11:22 EST
Received: from lc/garage/packard.DK 
	by cbosgd.ATT.UUCP (4.12/UUCP-Project/11.09.85)
	id AA15783; Tue, 24 Dec 85 10:55:34 est
Date: 24 Dec 1985 1835
From: Hiroshi G. Okuno <ihnp4!kddlab!nttlab!NTT-20!Okuno@seismo.CSS.GOV>
Subject: Greeting
Message-Id: <8512240935.AA26501@ntt.junet>
Received: by ihnp4.ATT.UUCP id AA12430; 24 Dec 85 08:41:08 CST (Tue)
Received: by kddlabs.junet (4.12/4.7)
	id AA23040; Tue, 24 Dec 85 18:37:11 jst
Received: by kddlabs.junet (4.12/4.7)
	id AA23037; Tue, 24 Dec 85 18:37:06 jst
Received: by ntt.junet (4.12/4.7JC-5) CHAOS with CHAOS-MAIL
	id AA26501; Tue, 24 Dec 85 18:35:02 jst
Received: from ihnp4.UUCP by lc/garage/packard.DK; 8512241556
To: hpp@sumex-aim.arpa
Cc: aap@sumex-aim.arpa


m       m      EEEEEE     RRRRRrr       RRRRRrr      Yy      yY
MM     MM      E	  R     R	R     R	       Yy  yY
M M   M M      EEEEEE	  RRRRRRr       RRRRRRr 	 YY
M  MmM  M      E	  R    R	R    R		 YY
M   	M      EEEEEE	  R     R	R     R		 YY
				

'  '  '   '  '  '  ANGEL' '  '  Xx   xX     m       m	    A	     SSSSSs
 '  '   '   '  '  '  ' ' ' '      X X       MM     MM	   A A	    S
'  'A '   'D  '  'V '  '* '  '     X	    M M   M M	  A   A	     SSSSSS
 '  '   '   '  '   '  ' * ' '     X X 	    M  MmM  M    AAAAAAA    s     S
'  ' A'   'N ' 'C   ' t*#* '    xX   Xx     M       M	A	A    SSSSSS
' E ' D '  '  '  '  'a**%**t '  
  '   '  'A  '  '  'o**%**$*a'  RCHI' '  '   '  ' T ' E'   'C  T  '   '  '
'  'U '    '  R' ' i**KSL*SU*o'   '  'E '  '   '   '  ' S'   'P  '  '  '  ' 
 '  '  R'  '   '  s**$**%*#*%*t '  '  '   '  'O  '   '  '  '  ' J' 'E  'CT '
'  '  '   '   '  s**%*#*%*$**%*h'   '  ''  '''  '  '  ' '  '  '   '   '  '  '
  '  ' T   '  ' i**%*#*%*#*%*#**a'  '  '' '  'I '  ''N'  '  '  '  '  '  ' 'A
'  '   '  C'   l**%*$*#*N*U*E**%*t  '  '  '   '   '  '  '  'A '  '  ' '  '  '
 '   '   '   'e**%*#*%*#*%H$**%*#*i' '  ' '  '  '  '  '  '  '   '   '  '  ' '
' O'  S'   ' n**N*T*T**%*#P%*E*C*L*s '' '   '  '   '  '  '  ' L' '  '  '  ' '
 '  O'   '  t**A*#*G**E*#*P**%*#*%**t '  'Q  'E  'L '   '  '  '   '   '  '  '
'  ' I  ' 'l**%*| |**%*#*%*#*%**$**%*a '  '   '    '  '   '  '  ' N  '  '  '
 '   '  ' a**%**| |*C**A**R**E**%*#*%*o  '  '   T'  L ' M '   '  '   '   '   '
'C 'R '  o**%**(←←|**%*#*%*$**%*#*%*#**e'  '  '   '   '   '   '  '   'Y  '   
  'P 'T '**%*#*%*#*$**%*#*%*#*%*#*$**%**d '  ' O'  C' L. '   ' o ' b'  '  '  '
 ' '  '  i   s  'n ' |     |' o'  't '  t  '  '  '   '   '   '   '  'r '   '
'  '  'u  '   ' t '  |     | '  '  '  '  ao'  '  '   '  '   '   '  '   '' g'

Happy New Year!

				- Gitchang -
-------

∂25-Dec-85  1159	avg@su-aimvax.arpa 	acyclicity in recursive rules 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 25 Dec 85  11:59:38 PST
Received: by su-aimvax.arpa with Sendmail; Wed, 25 Dec 85 11:56:44 pst
Date: Wed, 25 Dec 85 11:56:44 pst
From: Allen VanGelder <avg@diablo>
Subject: acyclicity in recursive rules
To: nail@diablo

Here are some random observations on the role of acyclicity in recursive
rules.
The evaluation hypergraph of an adorned rule is formed by making a node
for the head of the rule that contains only the bound arguments.  Nodes
for subgoals contain all arguments.  Assume no function symbols or
constants appear in the rules.  The evaluation hypergraph may or may not be
acyclic.  When it is, it has a qual tree, and information passing in
accordance with the edges of the qual tree looks attractive.

Now look at the common ancestor (ca) problem.
Read ca(A, B, C) as "A and B have common ancestor C".
Read a(A, B) as "A has ancestor B".
Read p(A, B) as "A has parent B".  This is the EDB, assumed to be a DAG.

One set of rules for ca (rules for  a  are standard):
ca(A,B,B) :- a(A,B).
ca(A,B,A) :- a(B,A).
ca(A,B,C) :- a(A,D), a(B,E), ca(D,E,C).

Look at the evaluation hypergraph for the recursive rule for ca↑bbf.
The nodes are:
ca↑b(A,B)    a←1(A,D)    a←2(B,E)    ca(D,E,C)
This is clearly cyclic.

Now look at the evaluation hypergraph for the recursive rule for ca↑bff.
The nodes are:
ca↑b(A)    a←1(A,D)    ca(D,E,C)    a←2(B,E)
This is acyclic, and has a qual tree that connects the nodes in the order
written.

Question:  Does this imply that evaluating ca↑bbf is about as hard as
evaluating ca↑bff?

Another set of rules for ca (Now we show  a  also):
ca(A,B,C) :- a(A,C), a(B,C).
a(A,A).
a(A,B) :- p(A,X), a(X,B).

Look at the evaluation hypergraph for the rule for ca↑bbf.
The nodes are:
ca↑b(A,B)    a←1(A,C)    a←2(B,C)
This is clearly cyclic.

Now look at the evaluation hypergraph for the recursive rule for ca↑bff.
The nodes are:
ca↑b(A)    a←1(A,C)    a←2(B,C)
This is acyclic, and has a qual tree that connects the nodes in the order
written.

Merry Christmas and Happy New Year.

∂26-Dec-85  1519	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books--Computer Science    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 26 Dec 85  15:19:08 PST
Date: Thu 26 Dec 85 15:14:06-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books--Computer Science
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA, cs%Playfair@SU-SCORE.ARPA
Message-ID: <12170288988.15.LIBRARY@SU-SCORE.ARPA>

Machine-Independent Organic Software Tools (MINT) 2nd edition. by Godfrey,
Hendry, Hermans, and Hessenberg.QA76.6.M319 1982.

Artificial Intelligence: Towards Practical Apllication. GDI Technology
Assessment and Management.  edited by Bernold and Albers. Q334.T43 1985 c.2

PORTAL Language Description. Lecture Notes In Computer Science. 198.
by Arnold Businger. QA76.73.P66.B87 1985.

IEEE Computer Society. Reliable Distributed System Software. by John
Stankovic.  QA76.9.D5S73 1985.

Functional Programming Languages and Computer Architecture. Nancy,
France, Sept. 1985. Lecture Notes In Computer Science. 201. ed. by
Jouannaud.QA76.7F86 1985.

CAD of Concurrent Computers. Research Studies Press. by Patrick Foulk.
QA76.9.S88F68 1985.

International Workshop On Timed Petri Nets. Torino, Italy. July 1985.
QA267.I695 1985.

IEEE Computer Society. Conference on Computer Vision and Pattern
Recognition. June, 1985. TA1632.I36 1985.

12th Annual International Symposium on Computer Architecture.
June 1985. Boston, Mass. Proceedings.

CPEUG 16th Meeting. Computer Performance Evaluation Users Group.
QA76.9.E94C65 1980.

Applied Finite Mathematics And Calculus. by Kertz.   QA37.2.K47 1985.

Interface: Calculus And the Computer. 2nd ed. by David Smith.
QA303.S596 1984 c.2

Discrete Structures: An Introduction to Mathematics For Computer
Science. by Fletcher Norris. QA76.9M35N67 1985.

Secure Communications And Asymmetric Cryptosystems. AAAS Selected
Symposium 69.  ed. by Gustavus Simmons. Z103.S42 1982.

Reliability Theory And Models: Stochastic Failure Models, Optimal
Maintenacne Policies, Life Testing, and Structures. Notes and
Reports in Computer Science and Applied Mathematics. ed. by 
Abdel-Hameed, Cinlar, and Quinn. TA169.R46 1984.

Prelude To Programming: Problem Solving and Algorithms. by
William Mitchell. QA76.6.M5294 1984.

Studies In Numerical Analysis. by Gene Golub. QA279.S83 1984.

Einfuhrung In Das Programmieren In LISP. by Christian-Michael
Hamann. QA76.73.L23H35 1982.

Information System Specification And Design Road Map. by Dennis
Connor. T58.6.C668 1985.

The Universal Machine. by Pamela McCorduck. QA76.M367 1985.

The Complete Guide To Software Testing. QED Information Sciences.
by William Hetzel.  QA76.76.T48H48 1984.

File And Data-Base Management Programs for the IBM PC. by Myron
Hecht. QA76.8I2594H39 1985.

Assembly Language and Systems Programming for the IBM PB and 
Compatibles.  by Karen Lemone. QA76.8I2594L425 1985.

Dictionary of Microelectronics and Microcomputer Technology.
Deutsch/Englisch, English/German. Reference TK7804.A88 1984.

The Software Encyclopedia. 1985/86. Reference QA76.5S664 1985
volumes 1 and 2.

Harry Llull
-------

∂27-Dec-85  0951	LIBRARY@SU-SCORE.ARPA 	Math/CS Library New Books--Computer Science    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 27 Dec 85  09:50:26 PST
Date: Fri 27 Dec 85 09:46:30-PST
From: C.S./Math Library <LIBRARY@SU-SCORE.ARPA>
Subject: Math/CS Library New Books--Computer Science
To: su-bboards@SU-SCORE.ARPA
cc: faculty@SU-SCORE.ARPA, cs%Playfair@SU-SCORE.ARPA
Message-ID: <12170491494.11.LIBRARY@SU-SCORE.ARPA>

Heuristic Reasoning About Uncertainty: An Artificial Intelligence
Approach.  Research Notes In Artificial Intelligence 2. by Paul R.
Cohen.  Q375.C64 1985.

IEEE Workshop on Languages For Automation. Cognitive Aspects In 
Information Processing.  Spain. June 28-29, 1985.  TJ212.2.I326 1985.

H. Llull
-------

∂27-Dec-85  1713	RICE@SUMEX-AIM.ARPA 	Tina
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 27 Dec 85  17:13:28 PST
Date: Fri 27 Dec 85 17:13:10-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Tina
To: aap@SUMEX-AIM.ARPA
Message-ID: <12170572808.14.RICE@SUMEX-AIM.ARPA>


A new version of the Tina system has been released.  This has a number of
major changes.  The new features of the system are documented in the new
release of the Tina System manual.

The new release of the manual can be found in the following :-

	3602:>System-Software>Tina>Tina-System.Document

A small file which contains poi9nters to the deltas is in :-

	3602:>System-Software>Tina>Changes.Document

I'm thinking about a proper printable form of the manual.  Apologies to those
who like to read off line.

Rice.
-------

∂27-Dec-85  1905	FEHRI@SU-CSLI.ARPA  
Received: from SU-CSLI.ARPA by SU-AI.ARPA with TCP; 27 Dec 85  19:05:27 PST
Date: Fri 27 Dec 85 18:59:47-PST
From: Fassi fehri <FEHRI@SU-CSLI.ARPA>
To: folks@SU-CSLI.ARPA
cc: fehri@SU-CSLI.ARPA


I am leaving tomorrow evening, back to Morocco. I would like to thank you
all for your friendship, and especially those of you who made me feel 
that no accident has happened. Bye and happy new year,

Abdu
-------

∂27-Dec-85  2008	ullman@su-aimvax.arpa 	meetings canceled
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 27 Dec 85  20:08:01 PST
Received: by su-aimvax.arpa with Sendmail; Fri, 27 Dec 85 20:04:25 pst
Date: Fri, 27 Dec 85 20:04:25 pst
From: Jeff Ullman <ullman@diablo>
Subject: meetings canceled
To: nail@diablo

I forgot to say that last Wednesday's meeting was canceled
on account of Christmas.  I hope nobody showed up.
This Wednesday, being New Years, probably shouldn't have a
meeting either.
				---Jeff

∂29-Dec-85  1047	PAPA@SU-SCORE.ARPA 	Concurrency Control Book 
Received: from SU-AIMVAX.ARPA by SU-AI.ARPA with TCP; 29 Dec 85  10:47:42 PST
Received: from SU-SCORE.ARPA by su-aimvax.arpa with Sendmail; Sun, 29 Dec 85 10:45:03 pst
Date: Sun 29 Dec 85 10:41:31-PST
From: C. Papadimitriou <PAPA@SU-SCORE.ARPA>
Subject: Concurrency Control Book
To: nail@SU-AIMVAX.ARPA
Message-Id: <12171025799.11.PAPA@SU-SCORE.ARPA>

Believe it or not, my book ``The Theory of Database Concurrency Control'' is
almost done.  Its shape is final enough that I would like to have some
people with interest in the area look at it.  If you are seriously interested
in looking at it in the next six weeks and give me your comments and
corrections, I would appreciate it.  Please send me a message.  I plan to
make about ten copies.  The table of contents follows (in broken TeXese):

\input macros

\obeylines
\centerline{\bf CONTENTS}

\vskip .25 in
CHAPTER ONE:  INTRODUCTION
1.1 Databases and Concurrency
1.2 The Model
APPENDIX: Notation, Graph Theory, Algorithms and Complexity.

\vskip .2in
CHAPTER TWO: CORRECTNESS

2.1 Introduction
2.2 Final-State Serializability
2.3 The Critics of Final-State Serializability
2.4 View Serializability
2.5 The Complexity of View Serializability
2.6 Conflict Serializability
2.7 Special Cases

\vskip.2in
CHAPTER THREE: MORE ON CORRECTNESS

3.1 Multiple Versions
3.2 Reliability
3.3 The Complexity of Reliability

\vskip.2in
CHAPTER FOUR: SCHEDULERS

4.1 Introduction
4.2 Locking 
4.3 Structured Locking
4.4 Timestamp Scedulers
4.5 Conflict Graph Schedulers
4.6 Multiversion Schedulers
4.7 Deadlocks
4.8 Reliable Schedulers

\vskip.2in
CHAPTER FIVE:  THE PERFORMANCE OF SCHEDULERS

5.1 Efficiency and Concurrency
5.2 Information and Concurrency
5.3 Multiversion Schedulers

\vskip.2in
CHAPTER SIX: THE THEORY OF LOCKING

6.1 Uninterpreted Locking
6.2 Entity Locks
6.3 Locking Protocols
6.4?? Deadlocks

\vskip .2in
CHAPTER SEVEN: DISTRIBUTED CONCURRENCY CONTROL

7.1 The Model
7.2 Distributed Schedulers
7.3 The Complexity of Distributed Concurrency Control
7.4 Distributed Locking

\vfill\end
-------

∂29-Dec-85  1452	NILSSON@SU-SCORE.ARPA 	End-of-year Note 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 29 Dec 85  14:51:55 PST
Date: Sun 29 Dec 85 14:46:21-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: End-of-year Note
To: csd@SU-SCORE.ARPA
Message-ID: <12171070369.9.NILSSON@SU-SCORE.ARPA>

Dear Stanford computer scientists and friends of computer science at Stanford:

This is where I came in--one year ago--at the end of autumn quarter.  I
only thought I knew a lot about the Computer Science Department then!
This end-of-the-year note is primarily to thank you all for helping to
make this an exciting year for me personally and a year of progress for
the department and for all of computer science everywhere at Stanford.
We have made some important achievements: moving the CSD to the School
of Engineering (it didn't hurt much, did it?); welcoming John Hennessy
back to Stanford to lead a new CSL that enjoys the enthusiastic support
of both EE and the CSD; designing a premier CS undergraduate curriculum
that many of us hope will be put in place next year; modernizing the
rest of the CS curriculum; upgrading CS student computing facilities;
beginning to restudy thoroughly our PhD requirements; and beginning an
updated, long-range plan for the department.  We are particularly proud
of the many achievements of our outstanding students including their
winning the ACM programming competition.

I still have many things on my list that I wish we had made more
progress on.  We still have only one endowed chair in the CSD.  That has
to change!  We still have too little space, and what we have is spread
around too much.  Things will improve when we move into new quarters in
the Near West Campus (probably circa 1990), but in the meantime any
relief from overcrowding will probably spread us around even more.  We
have had a search going for a robotics professor for almost a year, and
still our search committee has not found a suitable candidate.  We are
facing the stresses of a proposed undergraduate major, and we have not
yet begun searching for the new faculty that will be required to staff
that major.  Most of the CSD faculty and staff are stressed by needing
to do more than there is time to do it in.  We can probably all point to
things that we wish had been accomplished in 1985 which just didn't get
done.

Well, there is always 1986!  One of the major tasks before us
immediately at the beginning of 1986 is for us to decide do we or don't
we do the undergraduate major starting in September.  We can talk about
the pros and cons of all of that some more at our faculty retreat on
January 4, and we'll have Jim Gibbons there to exlain how he is prepared
to commit resources to the program.  In order to get all of the
necessary university approvals for the major, I think we will have to
decide formally one way or the other at our regular faculty meeting on
January 7.  We will be making that decision in the context of
authorizations from Jim Gibbons to begin several new searches if we
decide in favor of the major.

Assuming we decide to do the major, we need to organize ourselves to
conduct some searches immediately (and to incorporate these new people
into the department in September--if we can get super-high quality
people by then).  Although long-range plans are fine things, I suppose,
our only real guarantee of being a first-rate department in the future
is to have first-rate people.  So we will have to be exceptionally
careful to ensure that we get them.  Effort on that task is worth tons
of planning!

There are many other opportunities on the 1986 horizon including a
possible new GMD-sponsored institute in Palo Alto that could bring in
substantial additional research support for the CSD (GMD is a German
quasi-government research organization); and a possible new Center for
Parallel Computing.  We expect to make progress on our planning for
computing facilities (hardware and software).  We want to increase the
strength of our beneficial ties with local computer science research
establishments, and we look forward to productive cooperation with other
departments in our new home in the School of Engineering.

I'm sure that all of the faculty join me in expressing our deep
appreciation for the dedicated efforts of the staff.  Even though we are
all working hard, a deserved compliment is rewarding both to its giver
and receiver and helps make all of our lives here more enjoyable. I'm
particularly pleased that the staff has elected two "staff bureaucrats"
who will help us in the administration listen helpfully to staff
concerns and will also keep their constituents informed about
administrative planning.

Best wishes for a happy and fulfilling 1986!     -Nils
-------

∂30-Dec-85  1405	DAVIES@SUMEX-AIM.ARPA 	More IJCAI-85 articles of possible interest    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 30 Dec 85  14:05:29 PST
Date: Mon, 30 Dec 1985  14:04 PST
Message-ID: <DAVIES.12171324920.BABYL@SUMEX-AIM.ARPA>
From: DAVIES@SUMEX-AIM.ARPA
To:   AAP@SUMEX-AIM.ARPA
Subject: More IJCAI-85 articles of possible interest
cc:   Davies@SUMEX-AIM.ARPA

Y. Nishimoto and Y. Shirai: A parallel matching algorithm for stereo
vision. (p. 977) [Parallelism by subdividing an image]

Edmund H. Durfee, Victor R. Lesser, Daniel D. Corkill: Increasing
coherence in a distributed problem-solving network. (p. 1025)
[Improved local reasoning by sharing meta-level knowledge between
nodes]

Christopher Stuart: An implementation of a multi-agent plan
synchronizer. (p. 1031) [Using synchronizing primitives to resolve
conflicts between concurrent real-world actions]

Tom Henderson, Chuck Hansen, and Bir Bhanu: A framework for
distributed sensing and control. (p. 1106) [An architecture for
integrating information from multiple, possibly heterogeneous
sensors]

∂30-Dec-85  1441	NILSSON@SU-SCORE.ARPA 	Probable New Searches 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Dec 85  14:41:40 PST
Date: Mon 30 Dec 85 14:35:46-PST
From: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Subject: Probable New Searches
To: ac@SU-SCORE.ARPA
Message-ID: <12171330587.14.NILSSON@SU-SCORE.ARPA>

John Hennessy, Bob White and I had a productive meeting with Jim Gibbons
just before the Christmas break.  Jim is preparing to commit his "six
billets" --or at least five of them--now.  John Hennessy and I are
writing memos requesting search initiations, and these have to be
officially approved.  I expect approval soon after the first of the
year.

Here's the breakdown:

1.  A programming languages person (emphasizing, perhaps,
               applicative and/or parallel languages)

2.  A CAD (for VLSI) person 

3.  An AI person who wants to pursue basic AI research motivated
    by engineering applications

4.  A computer architecture or VLSI systems person

5.  A theory-of-computer science person

All of these will also have to have evident interest in teaching,
including undergraduate teaching.

(These are also in addition to the systems and robotics searches
currently underway.  Jim had also previously approved our reinstituting
a search--jointly with math--for an "applied math/scientific computing"
person.)

I think this is a good start, and now Jim can go to Rosse--saying he's
spending his six billets--requesting more to meet the rest of our growth
plan.  These are probably even more searches than we can effectively
pursue for now, so I think it remains only to get reasonably firm
commitments to start additional searches (according to our planned
schedule) as we fill these slots.  Of course these searches (and
additional promises) are contingent on the CSD faculty approving the
undergraduate program.  I think progress is sufficient to bring the
matter to a vote at the first faculty meeting in January--given the
discussion that is bound to take place at the retreat on Jan. 4.

-Nils
-------

∂30-Dec-85  1843	@SU-SCORE.ARPA:coraki!pratt@su-navajo.arpa 	Re: Probable New Searches 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 30 Dec 85  18:43:25 PST
Received: from su-navajo.arpa by SU-SCORE.ARPA with TCP; Mon 30 Dec 85 18:39:49-PST
Received: by su-navajo.arpa with Sendmail; Mon, 30 Dec 85 18:39:55 pst
Received: by coraki.uucp (1.1/SMI-1.2)
	id AA13423; Mon, 30 Dec 85 18:35:16 pst
Date: Mon, 30 Dec 85 18:35:16 pst
From: coraki!pratt@su-navajo.arpa (Vaughan Pratt)
Message-Id: <8512310235.AA13423@coraki.uucp>
To: Nils Nilsson <NILSSON@SU-SCORE.ARPA>
Cc: ac@su-score.ARPA, pratt@su-navajo.arpa
Subject: Re: Probable New Searches
In-Reply-To: message of Mon 30 Dec 85 14:35:46-PST.
             <12171330587.14.NILSSON@SU-SCORE.ARPA>

I don't see any slots for logic.  It is an absolute certainty that this
is an area in which Stanford must grow in order for us not merely to
keep pace with the world but to push it along.  It is critical that we
recognize logic as a subject that has not only played an important role
to date in computer science but will further increase in importance as
the ways in which we manipulate information structures become
progressively more abstract.

-v

∂31-Dec-85  1642	RICE@SUMEX-AIM.ARPA 	Hard Copy Tina docs.    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:42:00 PST
Date: Tue 31 Dec 85 14:36:39-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Hard Copy Tina docs.
To: aap@SUMEX-AIM.ARPA, bhayes-roth@SUMEX-AIM.ARPA, hewett@SUMEX-AIM.ARPA
Message-ID: <12171592892.22.RICE@SUMEX-AIM.ARPA>


The L100 and Tina manuals are now available in hard copy form with all
of the special characters readable.  If you want a copy please could you
send me a note so that I'll know how many copies to run off.

As before the new versions of these manuals in their on line form are in

    3602:>System-Software>L100>L100-Language-Manual.Document

and

    3602:>System-Software>Tina>Tina-System.Document

respectively.  Your feedback on these would be much appreciated.  I would
like them to be useful.

Rice.
-------

∂31-Dec-85  1642	RICE@SUMEX-AIM.ARPA 	Printing out Tina models.    
Received: from SUMEX-AIM.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:42:21 PST
Date: Tue 31 Dec 85 15:43:02-PST
From: Jim Rice <RICE@SUMEX-AIM.ARPA>
Subject: Printing out Tina models.
To: aap@SUMEX-AIM.ARPA
Message-ID: <12171604975.22.RICE@SUMEX-AIM.ARPA>


This message is for those who are going to write Tina models.

If you would prefer to use the special character representation of the
operator characters in Tina and would like to print out your model then
you should load "3602:>Tools>Make-Into-Scribe-File" and then use the
procedure (Make-Into-Scribe-File infile outfile).  This will make an output
file in Scribe format which will display the special characters in their
true form.

This program will cope with all Lisp machine special characters, not just
those used by Tina.

Rice.
-------

∂31-Dec-85  1652	RICHARDSON@SU-SCORE.ARPA 	General Faculty Meeting 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:52:39 PST
Date: Tue 31 Dec 85 08:09:54-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: General Faculty Meeting
To: faculty@SU-SCORE.ARPA
cc: gotelli@SU-SCORE.ARPA, woodward@SU-SCORE.ARPA
Message-ID: <12171522485.13.RICHARDSON@SU-SCORE.ARPA>

There will be a general faculty meeting on Tuesday, Jan. 7, 1985
at 2:30 in MJH 146. The known agenda items include: 1/ conferral
of degrees and 2/ a decision about the UG major. If you have any
additional agenda items, please forward them to me.

Thanks,
Anne
-------

∂31-Dec-85  1653	BSCOTT@SU-SCORE.ARPA 	[Phyllis M. Hughes <AS.PHU@Forsythe>: Army Research Office Announcement Coming: Pls Advise Clients]    
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:53:00 PST
Date: Tue 31 Dec 85 08:25:00-PST
From: Betty Scott <BSCOTT@SU-SCORE.ARPA>
Subject: [Phyllis M. Hughes <AS.PHU@Forsythe>: Army Research Office Announcement Coming: Pls Advise Clients]
To: AC@SU-SCORE.ARPA
Message-ID: <12171525233.14.BSCOTT@SU-SCORE.ARPA>

Are any of you interested in this announcement from the USARO?  Let me know
if you are, and I will get more information. - Betty
                ---------------

Return-Path: <AS.PHU@Lindy>
Received: from Lindy by SU-SCORE.ARPA with TCP; Mon 30 Dec 85 10:31:44-PST
Date: Mon, 30 Dec 85 10:34:46 PST
From: Phyllis M. Hughes <AS.PHU@Forsythe>
To: BSCOTT@SCORE
Subject:  Army Research Office Announcement Coming: Pls Advise Clients

Bridget Morgan, Gary McDonough, Betty Scott, Dena Polyhronakis:
Please share with appropriate PI's in your department.
P.Hughes 12/30/85

To:  FA.BCM, BSCOTT@SCORE, FE.DXP

FORWARDED MESSAGE 12/20/85 07:40 FROM AS.CFB "Fred Bentley": Army Research
Office Announcement Coming: Pls Advise Clients

Val and Phyllis:

     I believe that you may also have an interest in this announcement.

         Fred

To:  AS.VGM, AS.PHU, HK.EGC

FORWARDED MESSAGE 12/19/85 10:08 FROM HK.EGC "Earl Cilley": Army Research
Office Announcement Coming: Pls Advise Clients

ARO ANNOUNCES URI RESEARCH CENTERS

On December 12 the Army Research Office (ARO) published in the
Commerce Business Daily notice of its plans for the Army version
of the University Research Initiative.  The Army is inviting
proposals in the ten research areas listed below.  Awards are
planned, subject to available appropriations, in amounts up to
$3M per year for five years.  Funds will be provided for
fundamental research, graduate fellowships and research
instrumentation.  On December 27 ARO will publish a Broad Agency
Announcement.  Proposals will be due on February 28.  The ten
areas of interest are: biosystems and biotechnology, geosciences,
electro-optics, signal processing and image understanding, high
frequency microelectronics, fast reaction kinetics of energetic
materials, advanced propulsion systems, intelligent control
systems, ultradynamic performance materials, manufacturing
science and reliability and maintainability enhancement, and
advanced construction technology.


To:  AS.GUS, MCCABE@SIERRA, AS.PLB, AS.RMK, AS.SDG, AS.CFB
cc:  HK.PLD, HK.KDC


-------

∂31-Dec-85  1657	@SU-SCORE.ARPA:ullman@su-aimvax.arpa 	Re:  Gordon Bell 
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:57:46 PST
Received: from su-aimvax.arpa by SU-SCORE.ARPA with TCP; Tue 31 Dec 85 10:09:27-PST
Received: by su-aimvax.arpa with Sendmail; Tue, 31 Dec 85 10:12:56 pst
Date: Tue, 31 Dec 85 10:12:56 pst
From: Jeff Ullman <ullman@diablo>
Subject: Re:  Gordon Bell
To: RICHARDSON@SU-Score, ac@SU-Score

Id like to talk with him for a short time.

∂31-Dec-85  1657	RICHARDSON@SU-SCORE.ARPA 	Gordon Bell   
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:57:32 PST
Date: Tue 31 Dec 85 09:41:55-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Gordon Bell
To: ac@SU-SCORE.ARPA
Message-ID: <12171539237.13.RICHARDSON@SU-SCORE.ARPA>

Gordon Bell will be visiting the Stanford CSD on Jan. 8 ( in addition to
being the colloquium speaker for CS 500 on Jan. 7 ). If anyone would like
to see him, please let me know.

-Anne
-------

∂31-Dec-85  1655	RICHARDSON@SU-SCORE.ARPA 	Happy New Year
Received: from SU-SCORE.ARPA by SU-AI.ARPA with TCP; 31 Dec 85  16:55:18 PST
Date: Tue 31 Dec 85 08:33:36-PST
From: Anne Richardson <RICHARDSON@SU-SCORE.ARPA>
Subject: Happy New Year
To: csd@SU-SCORE.ARPA
Message-ID: <12171526798.13.RICHARDSON@SU-SCORE.ARPA>

Betty Scott asked me to let you know that the CSD offices will close at
3:00 p.m. today.
-------