perm filename E89.IN[LET,JMC] blob sn#877857 filedate 1989-10-06 generic text, type C, neo UTF8
COMMENT ⊗   VALID 00516 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00065 00002	∂01-Jul-89  2149	VAL 	frames    
C00066 00003	∂03-Jul-89  0125	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:bundy%aipna.edinburgh.ac.uk@nsfnet-relay.ac.uk 	RA Position at Edinburgh  
C00071 00004	∂03-Jul-89  1145	hoffman@csli.Stanford.EDU 	about to call you 
C00072 00005	∂03-Jul-89  1146	hoffman@csli.Stanford.EDU 	partial list of questions   
C00077 00006	∂03-Jul-89  1300	hoffman@csli.Stanford.EDU 	thanks very much for the interview!   
C00079 00007	∂03-Jul-89  2049	siegman@sierra.STANFORD.EDU   
C00084 00008	∂04-Jul-89  1022	Mailer 	Re: flag    
C00087 00009	∂06-Jul-89  0744	lm@cs.brown.edu 	Re:  papers  
C00089 00010	∂06-Jul-89  1726	PKR 	equality circumscription 
C00090 00011	∂06-Jul-89  1939	RFC 	Prancing Pony Bill  
C00092 00012	∂06-Jul-89  2038	BEDIT@Score.Stanford.EDU 	Summary of May computer charges.  
C00095 00013	∂07-Jul-89  1042	hayes@kanga.parc.xerox.com 	[Shrager.pa@Xerox.COM: T.L. Popejoy: The Culture and Polotics of AI]    
C00100 00014	∂07-Jul-89  1434	JONES@Score.Stanford.EDU 	teaching CS306 in the fall   
C00101 00015	∂08-Jul-89  0206	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	480 million digits of pi
C00109 00016	∂08-Jul-89  2254	CLT 	CLT itinerary etc   
C00112 00017	∂08-Jul-89  2316	CLT 	to do
C00113 00018	∂09-Jul-89  1439	Mailer 	Once upon a flag 
C00116 00019	∂09-Jul-89  2211	Mailer 	re: Once upon a flag  
C00117 00020	∂10-Jul-89  0412	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	math-funnel   
C00121 00021	∂10-Jul-89  0930	Mailer 	re: Abortion
C00123 00022	∂10-Jul-89  1109	vanMeule@ALLEGHENY.SCRC.Symbolics.COM 	re: 480 million digits of pi   
C00126 00023	∂10-Jul-89  1159	MPS  
C00127 00024	∂11-Jul-89  0036	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	480 million digits of pi
C00133 00025	∂11-Jul-89  0848	VAL 	re: parsing the past
C00135 00026	∂11-Jul-89  0947	siegman@sierra.STANFORD.EDU 	Math Book  
C00140 00027	∂11-Jul-89  1453	betsy@russell.Stanford.EDU 	Arkady Blinov    
C00142 00028	∂11-Jul-89  1515	MPS  
C00143 00029	∂11-Jul-89  1606	RWF 	re: Constitutional Ideals
C00144 00030	∂11-Jul-89  2000	JMC  
C00145 00031	∂12-Jul-89  0801	ACW@ELEPHANT-BUTTE.SCRC.Symbolics.COM 	math-funnel
C00150 00032	∂12-Jul-89  0852	@Score.Stanford.EDU:perrault@ai.sri.com 	Meeting  
C00152 00033	∂12-Jul-89  0940	perrault@ai.sri.com 	re: Meeting        
C00154 00034	∂12-Jul-89  1056	Mailer 	re: Offshore Oil Drilling Ban   
C00157 00035	∂12-Jul-89  1142	njacobs@vax.darpa.mil 	task:  DD   
C00160 00036	∂12-Jul-89  1143	njacobs@vax.darpa.mil 	task:  DF   
C00163 00037	∂12-Jul-89  1341	@Gang-of-Four.Stanford.EDU:arg@lucid.com 	new-qlisp    
C00165 00038	∂12-Jul-89  1506	DON@RML2.SRI.COM 	Re: Ban on Offshore Oil Drilling
C00169 00039	∂12-Jul-89  1731	scherlis@vax.darpa.mil 	TECHNICAL REPORTS    
C00180 00040	∂12-Jul-89  2229	Mailer 	re: Offshore Oil Drilling Ban   
C00181 00041	∂13-Jul-89  0019	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	math-funnel   
C00189 00042	∂13-Jul-89  0731	Mailer 	re: Offshore Oil Drilling Ban   
C00191 00043	∂13-Jul-89  0754	Mailer 	re: Ban on Offshore Oil Drilling
C00196 00044	∂13-Jul-89  0822	DON@RML2.SRI.COM 	re: Ban on Offshore Oil Drilling
C00200 00045	∂13-Jul-89  1009	chandler@Polya.Stanford.EDU 	Your vote on Karp    
C00202 00046	∂13-Jul-89  1024	chandler@Polya.Stanford.EDU 	re: Your vote on Karp     
C00203 00047	∂13-Jul-89  1151	ME 	failed mail returned 
C00205 00048	∂13-Jul-89  1321	DON@RML2.SRI.COM 	Nuclear Power    
C00209 00049	∂13-Jul-89  1332	Mailer 	re: Ban on Offshore Oil Drilling
C00215 00050	∂13-Jul-89  1334	siegman@sierra.STANFORD.EDU 	Re: Ban on Offshore Oil Drilling    
C00218 00051	∂13-Jul-89  1347	MPS  
C00219 00052	∂13-Jul-89  1443	PKR 	meeting on equality circumscription
C00221 00053	∂13-Jul-89  1530	Mailer 	re: Nuclear energy    
C00225 00054	∂13-Jul-89  1539	Mailer 	re: Nuclear Power
C00227 00055	∂13-Jul-89  1542	STEINBERGER@KL.SRI.COM 	re: Nuclear energy        
C00229 00056	∂13-Jul-89  1553	PKR 	re: meeting on equality circumscription 
C00230 00057	∂13-Jul-89  1900	JMC  
C00231 00058	∂14-Jul-89  0030	karish@forel.Stanford.EDU 	Re: Nuclear energy
C00233 00059	∂14-Jul-89  0315	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	535 million digits of pi?    
C00238 00060	∂14-Jul-89  0700	JMC  
C00239 00061	∂14-Jul-89  0700	JMC  
C00240 00062	∂14-Jul-89  0945	MPS  
C00241 00063	∂14-Jul-89  1430	cross@vax.darpa.mil 	DARPA Program Summaries 
C00254 00064	∂15-Jul-89  1311	ME 	re: discrepancy      
C00255 00065	∂15-Jul-89  1439	wab@sumex-aim.stanford.edu 	re: Nuke Japan Now!   
C00257 00066	∂15-Jul-89  2156	cphoenix@csli.Stanford.EDU 	re: greenhouse effect 
C00259 00067	∂16-Jul-89  0004	rdz@Gang-of-Four.Stanford.EDU 	rn instructions    
C00261 00068	∂16-Jul-89  1228	RDZ@Score.Stanford.EDU   
C00262 00069	∂17-Jul-89  0700	JMC  
C00263 00070	∂17-Jul-89  0900	JMC  
C00264 00071	∂17-Jul-89  0932	kolk@smiley.Stanford.EDU 	The articles form soc.culture.china    
C00266 00072	∂17-Jul-89  0937	kolk@labrea.stanford.edu 	The file is still around on labrea afterall 
C00267 00073	∂17-Jul-89  1336	wab@sumex-aim.stanford.edu 	re: Racism  
C00271 00074	∂17-Jul-89  1343	MPS  
C00273 00075	∂17-Jul-89  1406	eps@Princeton.EDU 	visit 
C00275 00076	∂17-Jul-89  1414	dai@csli.Stanford.EDU 	Re: [none]  
C00276 00077	∂17-Jul-89  1420	dai@csli.Stanford.EDU 	Re: [none]  
C00277 00078	∂18-Jul-89  0820	perrault@ai.sri.com 	meeting  
C00280 00079	∂18-Jul-89  1012	MPS  
C00281 00080	∂18-Jul-89  1102	davidson@psych.Stanford.EDU 	Re: "What can the white man say to the black woman?"    
C00284 00081	∂18-Jul-89  1203	wab@sumex-aim.stanford.edu 	Re: black doctors     
C00288 00082	∂18-Jul-89  1205	@Score.Stanford.EDU:clancey.pa@Xerox.COM 	Einstein remark   
C00290 00083	∂18-Jul-89  1216	harnad@clarity.Princeton.EDU 	Preprint  
C00366 00084	∂18-Jul-89  1219	Mailer 	re: missing the point 
C00367 00085	∂18-Jul-89  1300	JMC  
C00368 00086	∂18-Jul-89  1329	wab@sumex-aim.stanford.edu 	re: missing the point 
C00372 00087	∂18-Jul-89  1328	Mailer 	Re: missing the point 
C00381 00088	∂18-Jul-89  1458	A.ABIE@Macbeth.Stanford.EDU 	Re: black doctors    
C00383 00089	∂18-Jul-89  1856	Mailer 	re: greenhouse effect 
C00384 00090	∂19-Jul-89  0126	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	cf of Brillhart surds   
C00387 00091	∂19-Jul-89  0217	Mailer 	re: Protecting JMC [was Re: missing the point] 
C00389 00092	∂19-Jul-89  1050	cliff%unix.computer-science.manchester.ac.uk@NSFnet-Relay.AC.UK 	TESTING   
C00390 00093	∂19-Jul-89  1139	pi-data@vax.darpa.mil    
C00403 00094	∂19-Jul-89  1203	pi-data@vax.darpa.mil    
C00416 00095	∂19-Jul-89  1225	pi-data@vax.darpa.mil    
C00429 00096	∂19-Jul-89  1425	MRC@CAC.Washington.EDU 	re: White Man   
C00431 00097	∂19-Jul-89  1439	MRC@CAC.Washington.EDU 	re: White Man   
C00433 00098	∂19-Jul-89  1524	Mailer 	re: question for those who favor de-forestation
C00435 00099	∂19-Jul-89  1526	Mailer 	re: greenhouse effect 
C00436 00100	∂19-Jul-89  1536	RWF 	re: Nuclear waste disposal    
C00437 00101	∂19-Jul-89  1617	RWF 	re: Nuclear waste disposal    
C00438 00102	∂19-Jul-89  1608	RWF 	re: greenhouse effect    
C00439 00103	∂19-Jul-89  1920	cloutier@sierra.STANFORD.EDU 	Appointment    
C00441 00104	∂20-Jul-89  0601	cliff%unix.computer-science.manchester.ac.uk@NSFnet-Relay.AC.UK 	TESTING        
C00443 00105	∂20-Jul-89  1140	RWF 	re: Nuclear waste disposal    
C00444 00106	∂20-Jul-89  1447	pcohen@ai.sri.com 	speech acts
C00447 00107	∂20-Jul-89  1705	H.HARRY@Macbeth.Stanford.EDU 	Prof. McCarthy's third-world problem [was Re: When should I call JMC racist?]   
C00450 00108	∂21-Jul-89  0154	Mailer 	Reply and award for Prof. McCarthy [was re: Prof. McCarthy's third-world problem] 
C00455 00109	∂21-Jul-89  0944	pcohen@ai.sri.com 	re: speech acts 
C00457 00110	∂21-Jul-89  1429	RPG 	Scherlis  
C00458 00111	∂21-Jul-89  1438	RPG 	re: Scherlis   
C00459 00112	∂21-Jul-89  1522	alex@jessica.Stanford.EDU 	Orals scheduling decision: Friday August 18th 10am-12noon 
C00462 00113	∂21-Jul-89  1950	MRC@CAC.Washington.EDU 	re: Shockely's errors..   
C00465 00114	∂22-Jul-89  0700	JMC  
C00466 00115	∂22-Jul-89  0903	RPG 	re: Lisp macros, theory thereof    
C00467 00116	∂22-Jul-89  1519	Mailer 	re: IQ (USA vs Japan) 
C00469 00117	∂22-Jul-89  1553	Mailer 	re: IQ (USA vs Japan) 
C00471 00118	∂22-Jul-89  1644	gds@spam.istc.sri.com 	Re: IQ (USA vs Japan) 
C00473 00119	∂22-Jul-89  1903	Mailer 	re: IQ (USA vs Japan) 
C00476 00120	∂22-Jul-89  2352	terman@Portia.stanford.edu 	Re: IQ 
C00479 00121	∂23-Jul-89  1407	Mailer 	re: Abortion (what else?)  
C00480 00122	∂24-Jul-89  0033	BEDIT@Score.Stanford.EDU 	Summary of June computer charges. 
C00483 00123	∂24-Jul-89  0419	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	tough nombres 
C00488 00124	∂24-Jul-89  0517	wri!ilan@uunet.UU.NET 	Re: Riemann Hypothesis
C00491 00125	∂24-Jul-89  0856	ortiz@spam.istc.sri.com 	Recomendation  
C00493 00126	∂24-Jul-89  1223	Mailer 	re: IQ (USA vs Japan) 
C00496 00127	∂24-Jul-89  1702	hoffman.pa@Xerox.COM 	interview    
C00532 00128	∂24-Jul-89  2228	CLT 	file 
C00533 00129	∂25-Jul-89  0101	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Stutter-free strings    
C00541 00130	∂25-Jul-89  0730	JMC  
C00542 00131	∂25-Jul-89  0755	qphysics-owner@neat.cs.toronto.edu 	request  
C00544 00132	∂25-Jul-89  0912	MPS  
C00545 00133	∂25-Jul-89  1035	hoffman.pa@Xerox.COM 	also    
C00546 00134	∂25-Jul-89  1038	VAL 	re: meeting on Elephant  
C00547 00135	∂25-Jul-89  1158	MPS  
C00548 00136	∂25-Jul-89  1400	JMC  
C00549 00137	∂25-Jul-89  1442	MPS 	Directions to Bell's
C00551 00138	∂26-Jul-89  0055	VAL 	Counterfactuals
C00553 00139	∂26-Jul-89  0800	JMC  
C00554 00140	∂26-Jul-89  0829	MPS  
C00556 00141	∂26-Jul-89  1014	CLT 	gymboree  
C00557 00142	∂26-Jul-89  1203	VAL 	re: Counterfactuals 
C00560 00143	∂26-Jul-89  1207	BEASLEY@IBM.COM 	Alexander Gorbis  
C00562 00144	∂26-Jul-89  1306	ACW@ELEPHANT-BUTTE.SCRC.Symbolics.COM 	Stutter-free strings 
C00572 00145	∂26-Jul-89  1647	eaf@sumex-aim.stanford.edu 	Golub's nomination to NAE  
C00574 00146	∂26-Jul-89  1654	sf@csli.Stanford.EDU 	Matyasevitch visit
C00576 00147	∂26-Jul-89  2053	BEASLEY@IBM.COM 	Alex Gorbis  
C00577 00148	∂27-Jul-89  1125	wolfe@Polya.Stanford.EDU 	questions...  
C00579 00149	∂27-Jul-89  1159	JMC  
C00580 00150	∂27-Jul-89  1205	ullman@nimbin.Stanford.EDU 	Research Prof.   
C00583 00151	∂27-Jul-89  1555	sf@csli.Stanford.EDU 	re: Matyasevitch visit      
C00584 00152	∂28-Jul-89  0834	CLT 	$    
C00585 00153	∂28-Jul-89  0900	JMC  
C00586 00154	∂28-Jul-89  0913	VAL 	reply to message    
C00587 00155	∂28-Jul-89  1400	hoffman.pa@Xerox.COM 	what we will do   
C00590 00156	∂28-Jul-89  1402	GRAHAM%UMKCVAX2.BITNET@Forsythe.Stanford.EDU 	Concepts as objects ... 
C00593 00157	∂28-Jul-89  1509	Mailer@Score.Stanford.EDU 	Message of 25-Jul-89 14:58:01    
C00595 00158	∂28-Jul-89  1557	VAL 	Design stance  
C00596 00159	∂28-Jul-89  1612	VAL 	re: Design stance   
C00598 00160	∂29-Jul-89  0147	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	pi, yi, yi.   
C00604 00161	∂29-Jul-89  1338	weening@gang-of-four.Stanford.EDU 	Forsythe down Saturday   
C00607 00162	∂29-Jul-89  1500	wri!ilan@uunet.UU.NET 	Re: Sum of MoebiusMu[n] / Sqrt[n]    
C00609 00163	∂30-Jul-89  0325	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	russian intransigence [was Pi]    
C00614 00164	∂30-Jul-89  0649	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	hot series for ln2 
C00619 00165	∂31-Jul-89  0529	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	ln(2), pi
C00631 00166	∂31-Jul-89  0906	cross@vax.darpa.mil 	DARPA Project Summary   
C00633 00167	∂31-Jul-89  1000	JMC  
C00634 00168	∂31-Jul-89  1011	cloutier@sierra.STANFORD.EDU 	Rescheduling of meeting  
C00636 00169	∂31-Jul-89  1053	CLT 	DARPA Project Summary    
C00637 00170	∂31-Jul-89  1137	CLT 	DARPA Project Summary    
C00638 00171	∂31-Jul-89  1408	hoffman.pa@Xerox.COM 	sudden deadline   
C00639 00172	∂31-Jul-89 1418 GOODMAN%uamis@rvax.ccit.arizona.edu rumor has it
C00643 00173	∂31-Jul-89  1612	CLT 	summer pay
C00644 00174	∂31-Jul-89  1709	cloutier@sierra.STANFORD.EDU 	re: Rescheduling of meeting   
C00646 00175	∂01-Aug-89  0336	J.JBRENNER@Macbeth.Stanford.EDU 	re: Higgins and US policy against terrorism    
C00651 00176	∂01-Aug-89  0432	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	computer algebra challenge integrals   
C00656 00177	∂01-Aug-89  0436	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	stigidity
C00659 00178	∂01-Aug-89  0622	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	[ESC@STONY-BROOK.SCRC.Symbolics.COM: russian intransigence [was Pi]] 
C00667 00179	∂01-Aug-89  1000	JMC  
C00668 00180	∂01-Aug-89  1644	pjd@riacs.edu 	Your submission to CACM. 
C00671 00181	∂02-Aug-89  0640	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	After he's through canning Rose,  
C00675 00182	∂02-Aug-89  0651	P.DRAVID@GSB-WHY.Stanford.EDU 	re: Higgins   
C00676 00183	∂02-Aug-89  0800	JMC  
C00677 00184	∂02-Aug-89  0814	chandler@Polya.Stanford.EDU 	Pat Simmons......    
C00678 00185	∂02-Aug-89  0900	JMC  
C00679 00186	∂02-Aug-89  0928	pjd@riacs.edu 	re: Your submission to CACM.  
C00681 00187	∂02-Aug-89  0938	chandler@Polya.Stanford.EDU 	re: Pat Simmons......     
C00682 00188	∂02-Aug-89  1718	RQ.WSB@Forsythe.Stanford.EDU 	hoover press paper  
C00710 00189	∂03-Aug-89  0501	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	surprisingly icky generating function  
C00714 00190	∂03-Aug-89  0651	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	After he's through canning Rose,  
C00721 00191	∂03-Aug-89  0915	scherlis@vax.darpa.mil 	Financial Information Needed ASAP   
C00735 00192	∂03-Aug-89  0921	CLT 	financial projection for AI   
C00736 00193	∂03-Aug-89  1001	VAL 	re: bad  not   
C00737 00194	∂03-Aug-89  1041	MPS  
C00738 00195	∂03-Aug-89  1203	VAL 	Index: First draft  
C00745 00196	∂03-Aug-89  1246	scherlis@vax.darpa.mil 	acks  
C00747 00197	∂03-Aug-89  1426	CLT 	Are you on Scherlis's list SW-PI?  
C00748 00198	∂03-Aug-89  1515	pi-data@vax.darpa.mil    
C00751 00199	∂03-Aug-89  1525	pi-data@vax.darpa.mil    
C00754 00200	∂03-Aug-89  1603	CLT 	How about pi-data?  
C00755 00201	∂03-Aug-89  1715	VAL  
C00756 00202	∂03-Aug-89  1828	RFC 	Prancing Pony Bill  
C00759 00203	∂04-Aug-89  0835	deyoung@vax.darpa.mil 	RP3 AVAILABILITY VIA NET   
C00766 00204	∂04-Aug-89  1118	scherlis@vax.darpa.mil 	Important Announcement    
C00768 00205	∂04-Aug-89  2039	ME 	failed mail returned      
C00770 00206	∂04-Aug-89  2105	CLT 	tommorrow night
C00771 00207	∂05-Aug-89  0537	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	oopsilon 
C00775 00208	∂05-Aug-89  1304	danvy@gang-of-four.Stanford.EDU 	your science-fiction story 
C00783 00209	∂05-Aug-89  1310	danvy@gang-of-four.Stanford.EDU 	By the way  
C00785 00210	∂05-Aug-89  1916	jim@thrush.STANFORD.EDU 	Re: Blacks and Jews 
C00787 00211	∂06-Aug-89  0716	weening@gang-of-four.Stanford.EDU 	Offices   
C00789 00212	∂06-Aug-89  1553	rpg@lucid.com 	Meet 
C00791 00213	∂06-Aug-89  1605	ME 	DD display lossage   
C00792 00214	∂07-Aug-89  0922	rpg@lucid.com 	Meet      
C00793 00215	∂07-Aug-89  1000	JMC  
C00794 00216	∂07-Aug-89  1033	nilsson@Tenaya.Stanford.EDU 	re: cs306  
C00796 00217	∂07-Aug-89  1049	RPG 	Grumble   
C00797 00218	∂07-Aug-89  1538	JONES@vaxb.acs.unt.edu 	researchers in logical reasoning    
C00800 00219	∂07-Aug-89  1623	PAF 	Email, Email, wherefor art thou, Email? 
C00801 00220	∂07-Aug-89  1640	gregory@Polya.Stanford.EDU 	Re: suggestions solicited  
C00803 00221	∂07-Aug-89  2020	PAF 	email
C00804 00222	∂08-Aug-89  0948	WILBER@Score.Stanford.EDU 	Re: suggestions solicited   
C00814 00223	∂08-Aug-89  1003	P.REDLICH@GSB-WHY.Stanford.EDU 	suggestion   
C00817 00224	∂08-Aug-89  1010	P.REDLICH@GSB-WHY.Stanford.EDU 	How do you . . .  
C00818 00225	∂08-Aug-89  1012	wls@nova.npac.syr.edu 	Re:  RP3 AVAILABILITY VIA NET   
C00820 00226	∂08-Aug-89  1031	@NIC.DDN.MIL:sol@NIC.DDN.MIL 	re: your su.computer message  
C00822 00227	∂08-Aug-89  1116	nilsson@Tenaya.Stanford.EDU 	re: electronic mail  
C00825 00228	∂08-Aug-89  1217	GOTELLI@Score.Stanford.EDU 	Decommissioning of SCORE ***PLEASE READ***
C00827 00229	∂08-Aug-89  1334	RPG 	I forgot to mention this 
C00828 00230	∂08-Aug-89  1614	dupre@csli.Stanford.EDU 	kinds
C00830 00231	∂08-Aug-89  1816	M.MARTY@Macbeth.Stanford.EDU 	Re: abstract   
C00833 00232	∂08-Aug-89  1826	levinth@sierra.STANFORD.EDU 	problem with permission to hire someone  
C00835 00233	∂08-Aug-89  2000	JMC  
C00836 00234	∂08-Aug-89  2234	gandalf@csli.Stanford.EDU 	Re: Networks considered harmful - for electronic mail
C00841 00235	∂08-Aug-89  2323	LES 	re: Need opinions on company names.
C00842 00236	∂09-Aug-89  0705	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	an intriguing screw case for TAYLOR    
C00851 00237	∂09-Aug-89  0854	RPG  
C00852 00238	∂09-Aug-89  0910	RPG 	re: reply to message
C00853 00239	∂09-Aug-89  0928	RPG 	re: reply to message
C00854 00240	∂09-Aug-89  1146	CLT 	McCarthy's msg about a problem with permission to hire someone   
C00857 00241	∂09-Aug-89  1347	VAL 	Gelfond   
C00858 00242	∂09-Aug-89  1440	MPS  
C00859 00243	∂09-Aug-89  1518	JONES@vaxb.acs.unt.edu 	RE: re: researchers in logical reasoning 
C00862 00244	∂09-Aug-89  1643	TAJNAI@Score.Stanford.EDU 	research interests
C00865 00245	∂10-Aug-89  0200	@RELAY.CS.NET:hibino@ntt-elis.ntt.jp 	Ask your intention of using ELIS
C00870 00246	∂10-Aug-89  0727	JONES@vaxb.acs.unt.edu 	RE: re: re: researchers in logical reasoning  
C00872 00247	∂10-Aug-89  0956	C.COLE@Macbeth.Stanford.EDU 	RE: O.T.A. message to Su-etc   
C00880 00248	∂10-Aug-89  1005	tutiya@russell.Stanford.EDU 	response to your netmail article    
C00885 00249	∂10-Aug-89  1030	MPS  
C00886 00250	∂10-Aug-89  1234	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	an intriguing screw case for TAYLOR    
C00896 00251	∂10-Aug-89  1308	levinth@sierra.STANFORD.EDU 	McCarthy CPL Research Project  
C00898 00252	∂11-Aug-89  0616	boesch@vax.darpa.mil 	ISTO DIRECTOR APPOINTMENT   
C00900 00253	∂11-Aug-89  0821	BSCOTT@Score.Stanford.EDU 	Re: McCarthy CPL Research Project
C00902 00254	∂11-Aug-89  1020	MPS 	phone call
C00903 00255	∂11-Aug-89  1240	tom@Polya.Stanford.EDU 	[jcm@Polya.Stanford.EDU: was this for me?]    
C00905 00256	∂11-Aug-89  1558	MPS  
C00906 00257	∂11-Aug-89  1801	harnad@clarity.Princeton.EDU 	Editor of Current Contents    
C00909 00258	∂11-Aug-89  1804	jun@xroads.cc.u-tokyo.ac.jp 	Re: reply from jmc   
C00913 00259	∂13-Aug-89  2205	@RELAY.CS.NET:okuno@ntt-20.ntt.jp 	Inquiry about the ELIS again  
C00915 00260	∂14-Aug-89  0007	@RELAY.CS.NET:okuno@ntt-20.ntt.jp 	re: Inquiry about the ELIS again   
C00917 00261	∂14-Aug-89  0654	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Radical, Dude 
C00924 00262	∂14-Aug-89  0800	JMC  
C00925 00263	∂14-Aug-89  0900	JMC  
C00926 00264	∂14-Aug-89  1114	Mailer 	re: mathematical maturity in our youth    
C00927 00265	∂14-Aug-89  1129	cloutier@sierra.STANFORD.EDU 	Appointment    
C00929 00266	∂14-Aug-89  1203	rathmann@eclipse.stanford.edu 	Inconsistency and Multiple Theories (progress report) 
C00933 00267	∂14-Aug-89  1228	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Thesis progress 
C00935 00268	∂14-Aug-89  1356	MPS  
C00936 00269	∂14-Aug-89  1401	pjd@riacs.edu 	re: Text of signed Viewpoint  
C00937 00270	∂14-Aug-89  1743	cloutier@sierra.STANFORD.EDU 	re: Appointment
C00939 00271	∂14-Aug-89  1800	JMC  
C00940 00272	∂15-Aug-89  0300	@RELAY.CS.NET:okuno@ntt-20.ntt.jp 	re: Inquiry about the ELIS again   
C00944 00273	∂15-Aug-89  0718	njacobs@vax.darpa.mil 	Technical Reports
C00956 00274	∂15-Aug-89  0846	CLT 	research interests  
C00960 00275	∂15-Aug-89  1341	pjd@riacs.edu 	re: Text of signed Viewpoint  
C00969 00276	∂15-Aug-89  1813	CLT 	labrea account 
C00970 00277	∂16-Aug-89  0832	MPS 	dentist   
C00971 00278	∂16-Aug-89  1010	CLT 	labrea account 
C00972 00279	∂16-Aug-89  1350	siegman@sierra.STANFORD.EDU 	Electronic Mail on Networks    
C00974 00280	∂17-Aug-89  0157	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	angels on pinpoints
C00978 00281	∂17-Aug-89  0909	CLT 	gabriels salary
C00980 00282	∂17-Aug-89  0958	C.COLE@MACBETH.STANFORD.EDU 	re: O.T.A. message to Su-etc   
C00988 00283	∂17-Aug-89  1324	MPS  
C00989 00284	∂17-Aug-89  1428	cvenkat@iag.hp.com 	TeleFax - Suggestions    
C00994 00285	∂17-Aug-89  1459	VAL 	your letter    
C00995 00286	∂17-Aug-89  1632	telecom@eecs.nwu.edu 	Receipt: NETWORKS CONSIDERED HARMFUL - FOR ELECTRONIC MAIL
C00998 00287	∂17-Aug-89  1755	rpg@lucid.com 	publications   
C01009 00288	∂17-Aug-89  2301	JMC  
C01010 00289	∂18-Aug-89  0653	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	angels on pinpoints
C01019 00290	∂18-Aug-89  0900	JMC  
C01020 00291	∂18-Aug-89  1046	dgc@math.ucla.edu 	Re -- Networks Considered Harmful - For Electronic Mail 
C01025 00292	∂18-Aug-89  1531	root@Gang-of-Four.Stanford.EDU     
C01026 00293	∂18-Aug-89  1539	ullman@nimbin.Stanford.EDU 	lunch? 
C01027 00294	∂18-Aug-89  1621	CLT 	research interests  
C01028 00295	∂18-Aug-89  1657	ullman@nimbin.Stanford.EDU 	re: lunch?  
C01029 00296	∂18-Aug-89  1726	ole@csli.Stanford.EDU 	Re: Networks considered hamful - for email
C01031 00297	∂18-Aug-89  1745	ole@csli.Stanford.EDU 	re: Networks considered hamful - for email     
C01033 00298	∂18-Aug-89  2116	rick@uunet.uu.net 	Re: Networks Considered Harmful - For Electronic Mail   
C01035 00299	∂18-Aug-89  2136	rick@uunet.uu.net 	re: Networks Considered Harmful - For Electronic Mail   
C01039 00300	∂18-Aug-89  2207	thomas@mvac23.uucp 	Comments on telecom posting   
C01045 00301	∂19-Aug-89  0744	ficc!peter@uunet.UU.NET 	Re: Networks Considered Harmful - For Electronic Mail  
C01047 00302	∂19-Aug-89  0900	JMC  
C01048 00303	∂19-Aug-89  0904	rick@uunet.uu.net 	re: Networks Considered Harmful - For Electronic Mail   
C01050 00304	∂19-Aug-89  1100	JMC  
C01051 00305	∂19-Aug-89  1113	lotus!bobf@uunet.UU.NET 	Re: Networks Considered Harmful - For Electronic Mail  
C01056 00306	∂19-Aug-89  1925	doug@research.att.com    
C01061 00307	∂20-Aug-89  0901	JMC  
C01062 00308	∂20-Aug-89  1042	eaf@sumex-aim.stanford.edu 	[The Mailer Daemon <Mailer@score.stanford.edu> : Message of   
C01071 00309	∂20-Aug-89  1300	bzs@CS.BU.EDU 	Networks Considered Harmful - For Electronic Mail 
C01075 00310	∂20-Aug-89  1549	thomas@mvac23.uucp 	re: Comments on telecom posting    
C01078 00311	∂20-Aug-89  1831	ficc!peter@uunet.UU.NET 	re: Networks Considered Harmful - For Electronic Mail  
C01079 00312	∂20-Aug-89  2143	lars@salt.acc.com 	Re: Networks Considered Harmful - For Electronic Mail   
C01081 00313	∂21-Aug-89  0514	eli@chipcom.com 	comments on McCarthy's fax/email essay
C01084 00314	∂21-Aug-89  0616	greg%cs.heriot-watt.ac.uk@NSFnet-Relay.AC.UK 	Functional language workshop 
C01086 00315	∂21-Aug-89  0857	arch_ems@gsbacd.uchicago.edu 	RE: Networks Considered Harmful... 
C01095 00316	∂21-Aug-89  1431	dgc@math.ucla.edu 	Re: Re -- Networks Considered Harmful - For Electronic Mail  
C01099 00317	∂22-Aug-89  0725	JIMM%ACMVM.BITNET@Forsythe.Stanford.EDU 	Re: Signed Viewpoint, final texless version 
C01101 00318	∂22-Aug-89  1131	alex@jessica.Stanford.EDU 	thesis, requests for change...   
C01103 00319	∂22-Aug-89  1528	arc!chet@apple.com 	Networks Considered Harmful - For Electronic Mail 
C01112 00320	∂22-Aug-89  1540	pjd@riacs.edu 	Re:  Signed Viewpoint, final texless version 
C01114 00321	∂22-Aug-89  1741	qphysics-owner@neat.cs.toronto.edu 	test.    
C01116 00322	∂23-Aug-89  1400	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
C01118 00323	∂24-Aug-89  2105	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:STICKEL@warbucks.ai.sri.com 	CADE-10 Call for Papers  
C01124 00324	∂24-Aug-89  2154	cheriton@Pescadero.Stanford.EDU 	Gorbis 
C01126 00325	∂25-Aug-89  0819	harnad@clarity.Princeton.EDU 	On What Is and Isn't Symbolic 
C01147 00326	∂26-Aug-89  1633	Mailer 	failed mail returned  
C01148 00327	∂26-Aug-89  1832	@ri.cmu.edu:Gripe@FAC.CS.CMU.EDU 	Mail address warning 
C01150 00328	∂26-Aug-89  2011	Allen.Newell@NATASHA.MACH.CS.CMU.EDU 	Re: copies of two slides   
C01157 00329	∂27-Aug-89  1909	underdog@Portia.stanford.edu 	re: Columbia (vis a vis the drug cartel) (actually lunch)   
C01158 00330	∂27-Aug-89  2218	underdog@Portia.stanford.edu 	re: Columbia (vis a vis the drug cartel) (actually lunch)   
C01159 00331	∂28-Aug-89  0800	JMC  
C01161 00332	∂28-Aug-89  0907	Mailer 	re: Quote for August 28, 1989   
C01164 00333	∂28-Aug-89  0937	MPS 	Knowledge Systems   
C01165 00334	∂28-Aug-89  0959	ullman@nimbin.Stanford.EDU 	lunch  
C01166 00335	∂28-Aug-89  1345	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
C01168 00336	∂28-Aug-89  1447	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	Re:  party and dinner   
C01170 00337	∂28-Aug-89  1504	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
C01173 00338	∂28-Aug-89  1528	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
C01176 00339	ullman@nimbin
C01181 00340	∂28-Aug-89  1810	ullman@nimbin.Stanford.EDU 	Re:  stable models vs. well founded semantics  
C01189 00341	∂28-Aug-89  2000	BEDIT@Score.Stanford.EDU 	Summary of July computer charges. 
C01192 00342	∂28-Aug-89  2131	ullman@nimbin.Stanford.EDU 	re:  stable models vs. well founded semantics  
C01195 00343	∂29-Aug-89  0547	boyer@CLI.COM 	[postmaster@CLI.COM: Returned mail: User unknown] 
C01198 00344	∂29-Aug-89  0824	ARK 	smithsonian    
C01199 00345	∂29-Aug-89  0911	MPS 	Thesis    
C01200 00346	∂29-Aug-89  0954	MPS 	meeting   
C01201 00347	∂29-Aug-89  1020	ullman@nimbin.Stanford.EDU 	re:  stable models vs. well founded semantics  
C01209 00348	∂29-Aug-89  1209	MPS  
C01210 00349	∂29-Aug-89  1251	hoffman.pa@Xerox.COM 	debate  
C01212 00350	∂29-Aug-89  1302	bhr@sumex-aim.stanford.edu 	CS300 Speakers   
C01216 00351	∂29-Aug-89  1424	lars@salt.acc.com 	re: Networks Considered Harmful - For Electronic Mail   
C01219 00352	∂29-Aug-89  1436	korf@CS.UCLA.EDU 	Re: Advertisement     
C01222 00353	∂29-Aug-89  2357	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	The Saganic Verses 
C01226 00354	∂30-Aug-89  0134	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu,@albanycs.albany.edu:kapur@sutra.albany.edu 	post-doctoral fellowship
C01231 00355	∂30-Aug-89  0452	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	daily trivium 
C01233 00356	∂30-Aug-89  0555	cross@vax.darpa.mil 	Visit on 19 September 1989   
C01235 00357	∂30-Aug-89  1004	VAL 	re:  stable models vs. well founded semantics
C01237 00358	∂30-Aug-89  1103	ullman@nimbin.Stanford.EDU 	re:  stable models vs. well founded semantics  
C01240 00359	∂30-Aug-89  1107	ullman@nimbin.Stanford.EDU 	reply to VAL
C01241 00360	∂30-Aug-89  1108	ullman@nimbin.Stanford.EDU 	forget it   
C01242 00361	∂30-Aug-89  1737	betsy@russell.Stanford.EDU 	Blinov's TINLunch
C01244 00362	∂31-Aug-89  0332	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Hickerson's Life breakthrough
C01250 00363	∂31-Aug-89  1501	NIZ@Forsythe.Stanford.EDU 	AP newsfeed  
C01253 00364	∂31-Aug-89  2301	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Baker's dozen correction
C01256 00365	∂01-Sep-89  0032	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Hickerson strikes again 
C01263 00366	∂01-Sep-89  1308	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	private  
C01266 00367	∂01-Sep-89  1352	VAL  
C01267 00368	∂01-Sep-89  1354	alex@jessica.Stanford.EDU 	["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to say 
C01309 00369	∂01-Sep-89  1404	MPS 	JMC @ NCSU
C01312 00370	∂01-Sep-89  1427	alex@jessica.Stanford.EDU 	re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
C01314 00371	∂01-Sep-89  1446	alex@jessica.Stanford.EDU 	re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
C01316 00372	∂01-Sep-89  1516	CLT  
C01317 00373	∂01-Sep-89  1518	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	re: private   
C01319 00374	∂01-Sep-89  1524	yeager@sumex-aim.stanford.edu 	Re: Final version of Networks considered harmful.
C01321 00375	∂01-Sep-89  1557	alex@jessica.Stanford.EDU 	re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
C01323 00376	∂01-Sep-89  1600	MPS 	TGIF 
C01324 00377	∂01-Sep-89  1717	alex@jessica.Stanford.EDU 	re: Goodfellow    
C01325 00378	∂01-Sep-89  1732	alex@jessica.Stanford.EDU 	re: Goodfellow    
C01327 00379	∂01-Sep-89  1804	@bridge2.ESD.3Com.COM:gpz@api.ESD.3Com.COM 	Re: Final version of Networks considered harmful.  
C01330 00380	∂01-Sep-89  2000	JMC  
C01331 00381	∂01-Sep-89  2051	BEDIT@Score.Stanford.EDU 	Summary of August computer charges.    
C01334 00382	∂02-Sep-89  0445	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	this is a test     
C01336 00383	∂02-Sep-89  1302	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk 	theorem proving list
C01340 00384	∂02-Sep-89  1727	CLT  
C01341 00385	∂02-Sep-89  2158	root@NSFnet-Relay.AC.UK 	Mail Delivery Failure to uk.ac.hw.cs - Timeout    
C01343 00386	∂03-Sep-89  0353	RFC 	Prancing Pony Bill  
C01345 00387	∂03-Sep-89  1759	splat!root@ssyx.ucsc.edu 	regarding "Networks Considered Harmful"
C01354 00388	∂04-Sep-89  1459	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	simplification challenge
C01360 00389	∂04-Sep-89  1543	balzer@vaxa.isi.edu 	Publication of ISAT Prototyping White Paper 
C01413 00390	∂05-Sep-89  0802	chandler@Polya.Stanford.EDU 	Pat Simmons.... 
C01416 00391	∂05-Sep-89  1223	peters@russell.Stanford.EDU 	[MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: User unknown] 
C01420 00392	∂05-Sep-89  1257	chandler@Polya.Stanford.EDU 	Pat Simmons
C01422 00393	∂05-Sep-89  1313	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	Re:  Have you read 
C01424 00394	∂06-Sep-89  1207	MPS 	My car    
C01425 00395	∂06-Sep-89  1814	RWF 	sail 
C01426 00396	∂06-Sep-89  1840	GOODMAN%uamis@rvax.ccit.arizona.edu 	Electronic News article
C01428 00397	∂06-Sep-89  1844	ARK  
C01430 00398	∂06-Sep-89  2127	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	simplification challenge
C01437 00399	∂07-Sep-89  1029	rpg@lucid.com 	Lunch
C01438 00400	∂07-Sep-89  1039	VAL  
C01439 00401	∂07-Sep-89  1055	MPS 	Russian Visitor
C01441 00402	∂07-Sep-89  1111	VAL 	re: reply to message
C01442 00403	∂07-Sep-89  1107	weening@Gang-of-Four.Stanford.EDU 	Qlisp
C01444 00404	∂07-Sep-89  1120	shoham@Polya.Stanford.EDU     
C01445 00405	∂07-Sep-89  1120	gilberts@Polya.Stanford.EDU 	re: Russian Visitor  
C01447 00406	∂07-Sep-89  1128	VAL 	Index
C01455 00407	∂07-Sep-89  1746	VAL 	Nonmonotonic workshop    
C01460 00408	∂08-Sep-89  0451	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	eater3   
C01466 00409	∂08-Sep-89  0512	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	account       
C01469 00410	∂08-Sep-89  1238	CLT 	hardware discussion      
C01470 00411	∂08-Sep-89  1309	MLB@WHITE.SWW.Symbolics.COM 	Lisp history questions    
C01474 00412	∂08-Sep-89  1707	dresser@csli.Stanford.EDU 	re: Math question      
C01476 00413	∂08-Sep-89  2017	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 
C01478 00414	∂09-Sep-89  0814	CLT  
C01479 00415	∂09-Sep-89  0953	dresser@csli.Stanford.EDU 	re: Math question      
C01480 00416	∂09-Sep-89  0955	mrc@SUMEX-AIM.Stanford.EDU 	re: Another Quayle shot    
C01482 00417	∂11-Sep-89  0912	MPS  
C01483 00418	∂11-Sep-89  1000	JMC  
C01484 00419	∂11-Sep-89  1105	perrault@ai.sri.com 	Richard Waldinger  
C01487 00420	∂11-Sep-89  1211	Mailer 	re: Another Quayle shot    
C01489 00421	∂12-Sep-89  0430	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	flakiness
C01492 00422	∂12-Sep-89  0537	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	another attempt    
C01498 00423	∂12-Sep-89  1100	bh%anarres.Berkeley.EDU@Berkeley.EDU 	Re:  Might you be in a position to help ME in getting old SAIL dump
C01500 00424	∂12-Sep-89  1315	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk 	Research Assistantships Available  
C01506 00425	∂12-Sep-89  1332	CLT 	hardware  
C01510 00426	∂12-Sep-89  1346	weening@Gang-of-Four.Stanford.EDU 	Re: hardware   
C01516 00427	∂12-Sep-89  1443	tom@Polya.Stanford.EDU 	hardware   
C01519 00428	∂12-Sep-89  1533	CLT 	cinda
C01520 00429	∂12-Sep-89  1536	tom@Polya.Stanford.EDU 	cinda 
C01521 00430	∂12-Sep-89  1647	RWF 	re: proposal for hardware
C01522 00431	∂12-Sep-89  1839	TYSON@AI.SRI.COM 	Networks considered harmful - Email  
C01524 00432	∂13-Sep-89  0048	TYSON@Warbucks.AI.SRI.COM 	re: Networks considered harmful - Email    
C01527 00433	∂13-Sep-89  0129	ME 	Boyer 
C01528 00434	∂13-Sep-89  0458	cross@vax.darpa.mil 	Reschedule DARPA Visit  
C01530 00435	∂13-Sep-89  0659	CLT 	boyer
C01531 00436	∂13-Sep-89  0700	JMC  
C01532 00437	∂13-Sep-89  0945	CLT 	revised hardware proposal for DARPA
C01537 00438	∂13-Sep-89  1150	TYSON@Warbucks.AI.SRI.COM 	re: Networks considered harmful - Email    
C01539 00439	∂13-Sep-89  1220	MPS 	Book Review    
C01540 00440	∂13-Sep-89  1651	VAL 	Elephant programs   
C01542 00441	∂13-Sep-89  1702	VAL 	re: what was   
C01543 00442	∂13-Sep-89  1919	VAL 	Second hard example 
C01544 00443	∂16-Sep-89  2246	rdz@Gang-of-Four.Stanford.EDU 	Someone's plan file, for your amusement
C01546 00444	∂16-Sep-89  2257	restivo@Neon.Stanford.EDU     
C01547 00445	∂17-Sep-89  0231	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Life query . . .   
C01557 00446	∂17-Sep-89  0231	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Buckingham's p10 (unique period)  
C01561 00447	∂17-Sep-89  0327	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Period p guns imply period n p oscillators.  Magic.   
C01568 00448	∂17-Sep-89  0716	pamela@cs.columbia.edu 	WGBH/BBC project
C01571 00449	∂17-Sep-89  0912	@Score.Stanford.EDU,@AI.ai.mit.edu,@REAGAN.ai.mit.edu:dam@ai.mit.edu 	Justifying Resolution    
C01577 00450	∂17-Sep-89  1042	qphysics-owner@neat.cs.toronto.edu 	Call for papers: AIJ Special Issue on Qualitative Physics  
C01581 00451	∂18-Sep-89  0033	weening@Gang-of-Four.Stanford.EDU 	Re: sd to go4 when home  
C01584 00452	∂18-Sep-89  0932	weening@Gang-of-Four.Stanford.EDU 	Gang of Four   
C01585 00453	∂18-Sep-89  1253	restivo@Neon.Stanford.EDU 	Re: reply to message   
C01587 00454	∂18-Sep-89  1329	MPS 	Re:  JMC again (again)   
C01589 00455	∂18-Sep-89  1341	Mailer 	re: New Workers  
C01593 00456	∂18-Sep-89  1412	hayes@kanga.parc.xerox.com 	Vlad Dabija 
C01596 00457	∂18-Sep-89  1441	weening@Gang-of-Four.Stanford.EDU 	Sun salesman visit  
C01598 00458	∂18-Sep-89  1442	nilsson@Tenaya.Stanford.EDU 	re: Vlad Dabija 
C01600 00459	∂18-Sep-89  1640	weening@Gang-of-Four.Stanford.EDU 	Re: Sun salesman visit   
C01602 00460	∂18-Sep-89  2100	JMC  
C01603 00461	∂18-Sep-89  2303	harnad@clarity.Princeton.EDU 	Visual Search & Complexity: BBS Call for Commentators  
C01607 00462	∂19-Sep-89  0854	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	p135, 255, etc.    
C01619 00463	∂19-Sep-89  1028	DIGEX@ai.ai.mit.edu 	Electronic Mail and networks 
C01622 00464	∂19-Sep-89  1229	pamela@cs.columbia.edu 	re: WGBH/BBC project 
C01624 00465	∂19-Sep-89  1316	bergman@Polya.Stanford.EDU 	new NSF grant    
C01626 00466	∂19-Sep-89  1427	bergman@Polya.Stanford.EDU 	re: new NSF grant     
C01628 00467	∂19-Sep-89  1727	CLT 	visit
C01629 00468	∂20-Sep-89  0342	harnad@clarity.Princeton.EDU 	BBS & APS 
C01635 00469	∂20-Sep-89  0404	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	rudimentry pgm + [HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu: More c/3's]  
C01648 00470	∂20-Sep-89  0457	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Oops!  And shred that cultural litmus! 
C01651 00471	∂20-Sep-89  0745	crucible@fernwood.MPK.CA.US 	THE INTERNET CRUCIBLE - Volume 1, Issue 2
C01702 00472	∂20-Sep-89  1052	clm@prefect.es.llnl.gov 	greetins  
C01704 00473	∂20-Sep-89  1324	VAL 	Coming to Austin    
C01710 00474	∂21-Sep-89  0053	weening@Gang-of-Four.Stanford.EDU 	ClariNet  
C01714 00475	∂21-Sep-89  0700	JMC  
C01715 00476	∂21-Sep-89  0700	JMC  
C01716 00477	∂21-Sep-89  0730	harnad@clarity.Princeton.EDU 	APS & BBS: Responses
C01724 00478	∂21-Sep-89  1223	njacobs@vax.darpa.mil 	Technical Reports
C01735 00479	∂21-Sep-89  1350	VAL 	re: Technical Reports    
C01736 00480	∂21-Sep-89  1547	weening@Gang-of-Four.Stanford.EDU 	Comparison of Sun and DEC machines 
C01741 00481	∂21-Sep-89  1609	kolk@smiley.Stanford.EDU 	alternative to ap news wire  
C01800 00482	∂21-Sep-89  1651	pamela@cs.columbia.edu 	re: WGBH/BBC project 
C01801 00483	∂21-Sep-89  1747	tom@Polya.Stanford.EDU 	Comparison of Sun and DEC machines  
C01803 00484	∂21-Sep-89  1759	VAL 	reply to message    
C01804 00485	∂21-Sep-89  1953	VAL 	The scope paper
C01805 00486	∂21-Sep-89  2120	perlis@yoohoo.cs.umd.edu 	your visit    
C01808 00487	∂22-Sep-89  0945	rabin@harvard.harvard.edu     
C01809 00488	∂22-Sep-89  1311	VAL 	Seminar   
C01810 00489	∂22-Sep-89  1353	VAL 	re: Seminar    
C01811 00490	∂22-Sep-89  1546	CLT 	hardware  
C01821 00491	∂22-Sep-89  1656	sf@csli.Stanford.EDU 	Matijasevitch visit    
C01823 00492	∂22-Sep-89  1700	barwise@russell.Stanford.EDU 	Re: Matijasevitch visit  
C01825 00493	∂22-Sep-89  2004	CLT 	copier    
C01826 00494	∂23-Sep-89  0835	harnad@elbereth.rutgers.edu 	APS/BBS Poll: Late Returns
C01851 00495	∂23-Sep-89  1037	perlis@yoohoo.cs.umd.edu 	your visit    
C01854 00496	∂23-Sep-89  1557	srh@flash.bellcore.com 	"Skywriting"    
C01862 00497	∂23-Sep-89  1632	perlis@yoohoo.cs.umd.edu 	Oct 5th  
C01864 00498	∂23-Sep-89  1801	harnad@clarity.Princeton.EDU 	Symbols and Connections  
C01877 00499	∂24-Sep-89  1345	scherlis@vax.darpa.mil 	Announcement    
C01879 00500	∂25-Sep-89  0800	menke@harvard.harvard.edu 	talk at harvard, october 2  
C01891 00501	∂26-Sep-89  1629	CLT 	sail users
C01892 00502	∂26-Sep-89  1841	poser@crystals.STANFORD.EDU 	curious advertisement
C01894 00503	∂27-Sep-89  0923	@po2.andrew.cmu.edu:jhm+@andrew.cmu.edu 	Re: Networks Considered Harmful - For Electronic Mail 
C01914 00504	∂27-Sep-89  0926	boyland@sequoia.Berkeley.EDU 	looking for references about prototyping
C01916 00505	∂27-Sep-89  1126	boyland@sequoia.Berkeley.EDU 	looking for references about prototyping
C01918 00506	∂27-Sep-89  1511	cblpf!mark@att.att.com 	Networks Considered Harmful - For Electronic Mail  
C01923 00507	∂27-Sep-89  1800	harnad@clarity.Princeton.EDU 	Searle's Problem vs. Fodor's Problem    
C01928 00508	∂27-Sep-89  1939	hayes@kanga.parc.xerox.com 	Vlad Dabija 
C01930 00509	∂28-Sep-89  0902	rh@mbunix.mitre.org 	Some References on Prototyping    
C01933 00510	∂28-Sep-89  1011	smu!tanik@uunet.UU.NET 	Re:  looking for references about prototyping 
C01935 00511	∂29-Sep-89  0851	@REAGAN.AI.MIT.EDU:dam@AI.MIT.EDU 	The inverse method  
C01941 00512	∂29-Sep-89  1056	scherlis@vax.darpa.mil 	Awards
C01943 00513	∂29-Sep-89  1252	balzer@vaxa.isi.edu 	Prototyping Panel at OOPSLA  
C01945 00514	∂29-Sep-89  2253	ungar@self 	Re:  looking for references about prototyping   
C01947 00515	∂30-Sep-89  1041	harnad@clarity.Princeton.EDU 	Re: Skywriting 
C01968 00516	∂30-Sep-89  2115	harnad@clarity.Princeton.EDU 	Searle and the Symbol Grunding Problem  
C02010 ENDMK
C⊗;
∂01-Jul-89  2149	VAL 	frames    

In your paper with Hayes you suggest that the frame problem can be solved
if "a number of fluents are declared as attached to the frame and the effect
of an action is described by telling which fluents are changed, all others
being presumed unchanged". What is your current view of this idea?

∂03-Jul-89  0125	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:bundy%aipna.edinburgh.ac.uk@nsfnet-relay.ac.uk 	RA Position at Edinburgh  
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Jul 89  01:25:14 PDT
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU.#Internet) by SCORE.STANFORD.EDU with TCP; Mon 3 Jul 89 01:25:34-PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa22928;
          3 Jul 89 4:13 EDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU  3 Jul 89 04:12:09 EDT
Received: from lcs.mit.edu (CHAOS 15044) by MC.LCS.MIT.EDU  3 Jul 89 04:11:17 EDT
Received: from NSFNET-RELAY.AC.UK by mintaka.lcs.mit.edu id aa22875;
          3 Jul 89 4:08 EDT
Received: from aipna.edinburgh.ac.uk by NSFnet-Relay.AC.UK 
           via Janet with NIFTP  id aa01048; 3 Jul 89 8:51 BST
Date: Mon, 3 Jul 89 08:56:12 BST
Message-Id: <2273.8907030756@affric.aipna.ed.ac.uk>
From: Alan Bundy <bundy%aipna.edinburgh.ac.uk@nsfnet-relay.ac.uk>
Subject: RA Position at Edinburgh
To: theorem-provers%mc.lcs.mit.edu@nsfnet-relay.ac.uk

	Please circulate or post the attached job ad. Thanks.

			Alan Bundy



		Department of Artificial Intelligence
		     University of Edinburgh

			  RESEARCH FELLOW
		      (Mathematical Reasoning)


Applications are invited for an SERC supported post, tenable from 1st
November 1989, or on a mutually agreed date. Appointment will be to
September 30th 1991, initially, but with a possibility of renewal.
The research is to develop proof plans, a technique for guiding the
search for a proof in automatic theorem proving.  The main application
is to the automatic synthesis, verification and transformation of
logic programs using constructive logic.  The project is led by
Professor Alan Bundy and Dr Alan Smaill.

Candidates should possess a PhD or have equivalent research or
industrial experience.  Knowledge of logic is essential and knowledge
of artificial intelligence, formal methods in software engineering or
logic programming would be an advantage.  Salary is on the AR1A scale
in the range 9,865-15,720 pounds p.a., according to age,
qualifications and experience.

Applicants should send a CV and the names of two referees to:

		Prof. Alan Bundy. 
		Department of Artificial Intelligence, 
		University of Edinburgh, 
		80 South Bridge, 
		Edinburgh, 
		EH1 1HN.

as soon as possible.  The closing date for applications is 17th July
1989.  Further details may be obtained from Prof. Bundy (at the above
address or email to bundy@uk.ac.edinburgh or bundy@rutgers.edu)
quoting reference number 5678/E.








∂03-Jul-89  1145	hoffman@csli.Stanford.EDU 	about to call you 
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Jul 89  11:45:17 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA03664; Mon, 3 Jul 89 11:46:04 PDT
Date: Mon 3 Jul 89 11:46:03-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: about to call you
To: jmc@sail.stanford.edu
Message-Id: <615494763.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>


Following this message will be another message with a partial list of questions

thanks
reid
-------

∂03-Jul-89  1146	hoffman@csli.Stanford.EDU 	partial list of questions   
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Jul 89  11:45:53 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA03690; Mon, 3 Jul 89 11:46:40 PDT
Date: Mon 3 Jul 89 11:46:40-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: partial list of questions
To: jmc@sail.stanford.edu
Message-Id: <615494800.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>


Questions:

Recently, parallel distributed processing--usually dubbed "connectionism" or
"neural nets"--has been receiving much funding.  Many people feel that it 
solves, or has the ability to solve, many of the short-comings of more 
traditional AI approaches.  Could you share your thoughts with us about PDP?

Due to AI's emphasis on issues of knowledge, belief, "intelligent" problem-
solving, inference mechanisms and so on, some people feel that AI is actually 
closer to Psychology and Philosophy than Computer Science.  Could you comment 
on the relationship between Artificial Intelligence and Computer Science?  Do
these previous concerns have something to do with the inter-disciplinary nature
of AI?  

Sometime, it appears as if there is a division between "engineering" AI (e.g.
expert systems and other knowledge systems) and "science" AI (non-monotonic
logic, problem-solving, learning, etc.)  Sometimes, proponents of the 
"engineering" side claim that the science is done and only technological 
problems remain.  Sometimes, proponents of the "science" side claim that the
engineering approach will not accomplish nearly what it wishes to.  Could you
comments on this division?

There are some old debates with AI.  Dreyfus, Searle, Weizenbaum, and Winograd
have all taken exception with AI one way or another.  Do you have anything to
add to these debates?  [Should I make this more specific?]

What ground has AI covered in the last 10-20 years and what ground do you 
think that AI might cover in the next 10-20 years?  

Could you comment a little on the physical symbol systems hypothesis, which
might be interpreted as saying that "symbol manipulation (e.g by a machine
or by a human being) is sufficient, and perhaps necessary, for intelligence."?

Could you also comment a little on the (related) knowledge representation 
hypothesis, which might be interpreted as saying that any mechanical 
intelligent device will have some sort of structured knowledge both which we
as outside observers can interpret as propositional knowledge of the 
machine's perception of it's environment and which plays a causal role in the
machines behavior?

---this is an incomplete list, but I needed to stop typing so I could call
you now.

thanks
--reid
-------

∂03-Jul-89  1300	hoffman@csli.Stanford.EDU 	thanks very much for the interview!   
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Jul 89  13:00:53 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA07159; Mon, 3 Jul 89 13:01:41 PDT
Date: Mon 3 Jul 89 13:01:40-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: thanks very much for the interview!
To: jmc@sail.stanford.edu
Message-Id: <615499300.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>


Sorry for getting some of the questions wrong (e.g. physical symbol systems
hypothesis.)  I have some vague questions about intentionality and AI,
and some others floating around, but I wanted to write them out to see if
they amount to anything.  I will send you the transcription by next Sunday
and then we will go from there.  (The transcription will be edited.)

thanks again and hope that you enjoy Japan.

reid
-------

∂03-Jul-89  2049	siegman@sierra.STANFORD.EDU   
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 3 Jul 89  20:49:44 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Mon, 3 Jul 89 20:46:33 PDT
Date: Mon, 3 Jul 89 20:46:33 PDT
From: siegman@sierra.STANFORD.EDU (Anthony E. Siegman)
To: John McCarthy <JMC@SAIL.Stanford.EDU>
In-Reply-To: Your message of 03 Jul 89 1700 PDT 
Cc: siegman@sierra.STANFORD.EDU
Message-Id: <CMM.0.88.615527192.siegman@>

Thanks _very_ much; it was indeed in Newman's World of Mathematics
that I read it.

The question arose because I'm trying to bring about some cultural
understanding between administrators and faculty in dealing with lab
safety requirements.  I've suggested that administrators should read
and ponder Feynman's essay on his service on the Rogers Commision
investigating the Challenger disaster; and that Feynman's role
involved at least two aspects:

1) Focusing on the "physical realities" of the situation ("Mother
Nature can't be fooled"), in contrast to (and with a certain scorn
for) the "political realities" (e.g., keeping NASA politically viable)
with which Rogers was obviously also concerned;

2) Digging below the surface of the formal reports and statistics
(with an attitude of considerable skepticism toward them) with the
objective of finding out personally what's really going on at the
working level.

I've further suggested that many science and engineering faculty will
not only be aware of but will personally identify with Feynman's role
in this case, and will take it as a model for their response to safety
requirements.  Resulting recommendations:

1) Assemble statistics on lab accidents if possible (a la FAA
investigations on aircraft accidents), as justification for safety
measures; but be prepared for your audience to insist on digging below
the surface of these statistics into individual cases, and to react
with scorn if the statistics turn out to be phony or inflated.

2)  If there are certain safety policies that have to be implemented
for legal or bureaucratic or "CYA" reasons, but that have no rational
justification in terms of real safety improvement, be bluntly honest
about both of these points -- at least in dealing internally with
faculty.  Faculty will be more cooperative if the staff frankly
concedes that some measure is pointless but still has to be done than
if someone tries to make them believe, or pretends to believe, it is a
genuinely useful measure.

I'll have to review Morse, but I thought he might provide more
evidence for the "dig below the surface" line of approach.

∂04-Jul-89  1022	Mailer 	Re: flag    
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 4 Jul 89  10:22:15 PDT
Date: Tue, 4 Jul 89 10:22:34 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: Re: flag   
To: JMC@SAIL.Stanford.EDU
cc: su-etc@SAIL.Stanford.EDU
In-Reply-To: <j8tYh@SAIL.Stanford.EDU>
Message-ID: <12507342176.10.STEINBERGER@KL.SRI.COM>

FROM JMC:

> 3. It might mitigate and slow the intellectuals' war against
> traditional American culture.  That's doubtful, however.  It
> seems to me that this war has a tribal character.  I notice
> this on su-etc, where so many of the countercultural expressions
> take the form of jeering at the enemy rather than arguing about
> issues.  I don't mainly refer to the flag issue, which, I suppose,
> mainly lends itself to ad hominem arguments.

Could you possibly define what you mean by "traditional American culture"?
And then could you explain what what "an intellectuals' war against" it
implies?  And I would hope that this "tribal" business might be dropped
in favor of some more specific criticisms or concerns.  After all, you
are decrying ad hominem attacks and the use of the word "tribal" can
hardly be interpreted as flattering to anyone.

I assume that the term "war" is used in an almost theological sense here:
Anyone who dares oppose a principle considered a part of "traditional
American" values must be declared a member of an enemy camp.  There's no
room for political debates or discussions when "American values" are
at stake.  Or am I misinterpreting the implications of "war"?

Happy 4th.

-ric steinberger

←≠
-------

∂06-Jul-89  0744	lm@cs.brown.edu 	Re:  papers  
Received: from cs.brown.edu by SAIL.Stanford.EDU with TCP; 6 Jul 89  07:44:29 PDT
Received: from zaphod.cs.brown.edu by cs.brown.edu (4.1/SMI-4.0)
	id AA21275; Thu, 6 Jul 89 10:43:59 EDT
Date: Thu, 6 Jul 89 10:44:14 edt
From: lm@cs.brown.edu
Message-Id: <8907061444.AA04431@zaphod.cs.brown.edu>
To: JMC@SAIL.Stanford.EDU
Subject: Re:  papers

I have some copies of my thesis lying around (an unusual occurrence),
and I will send you one right now.  I will try to find some copies
of other papers, but this will be hard, because I am currently moving,
and most of my stuff is in boxes.  So, it may take a while for the
other papers.

I haven't had time yet to check out Epistemic MOtivated Action Theory
on the Mr. S and Mr. P problem, but will let you know when I do.

Leora Morgenstern

∂06-Jul-89  1726	PKR 	equality circumscription 
To:   JMC
CC:   PKR   
Dear Professor McCarthy,

Would it be possible to meet some time?  I have some new results
on equality circumscription that I would like to show you.  I
would be very interested in learning what you think of them, and
also in any advice you might have on what might be interesting
research topics to pursue from here.
			Thanks,
			    Peter

∂06-Jul-89  1939	RFC 	Prancing Pony Bill  
Prancing Pony bill of     JMC   John McCarthy           6 July 1989

Previous Balance            22.55
Payment(s)                  22.55  (check 6/22/89)
                           -------

Current Charges              4.00  (bicycle lockers)
                           -------
TOTAL AMOUNT DUE             4.00


PAYMENT DELIVERY LOCATION: CSD Receptionist.

Make checks payable to:  STANFORD UNIVERSITY.
Please deliver payments to the Computer Science Dept receptionist, Jacks Hall.
To ensure proper crediting, please include your PONY ACCOUNT NAME on your check.

Note: The recording of a payment takes up to three weeks after the payment is
made, but never beyond the next billing date.  Please allow for this delay.

Bills are payable upon presentation.  Interest of  1.0% per month will be
charged on balances remaining unpaid 25 days after bill date above.

An account with a credit balance earns interest of  .33% per month,
based on the average daily balance.

∂06-Jul-89  2038	BEDIT@Score.Stanford.EDU 	Summary of May computer charges.  
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 6 Jul 89  20:37:58 PDT
Date: Thu 6 Jul 89 20:36:11-PDT
From: Billing Editor <BEDIT@Score.Stanford.EDU>
Subject: Summary of May computer charges.
To: MCCARTHY@Score.Stanford.EDU
Message-ID: <12507978171.9.BEDIT@Score.Stanford.EDU>

Dear Mr. McCarthy,

Following is a summary of your computer charges for May.

Account     System   Billed    Pct      Cpu    Job   Disk  Print   Adj   Total

JMC         SAIL     2-DMA804  100   556.63  22.69 ***.**  43.87  5.00 2847.21
MCCARTHY    SCORE    2-DMA804  100      .00    .00  30.47    .00  5.00   35.47
jmc         LABREA   2-DMA804  100      .00    .00 105.90    .00  5.00  110.90

Total:                               556.63  22.69 ***.**  43.87 15.00 2993.58


University budget accounts billed above include the following. 

Account     Princip Inv      Title                      Comment             

2-DMA804    McCarthy         N00039-84-C-0211           Task 16, AI          


The preceding statement is a condensed version of the detailed summary sheet 
sent monthly to your department. 

Please verify each month that the proper university budget accounts are paying 
for your computer usage.  Please also check the list of account numbers below 
the numeric totals.  If the organizations/people associated with that account 
number should NOT be paying for your computer time, send mail to BEDIT@SCORE. 

Please direct questions/comments to BEDIT@SCORE. 
-------

∂07-Jul-89  1042	hayes@kanga.parc.xerox.com 	[Shrager.pa@Xerox.COM: T.L. Popejoy: The Culture and Polotics of AI]    
Received: from arisia (arisia.Xerox.COM) by SAIL.Stanford.EDU with TCP; 7 Jul 89  10:42:11 PDT
Received: from kanga.parc.Xerox.COM by arisia with SMTP
	(5.61+/IDA-1.2.8/gandalf) id AA03496; Fri, 7 Jul 89 09:45:08 -0700
Received: by kanga.parc.xerox.com
	(5.61+/IDA-1.2.8/gandalf) id AA21338; Fri, 7 Jul 89 09:47:31 PDT
Message-Id: <8907071647.AA21338@kanga.parc.xerox.com>
Date: Fri, 7 Jul 89 09:47:31 PDT
From: Pat Hayes <hayes@arisia>
To: JMC@SAIL.stanford.edu
Subject: [Shrager.pa@Xerox.COM: T.L. Popejoy: The Culture and Polotics of AI]


John,  I thought this might interest you.

Pat

Date: Thu, 6 Jul 89 21:20 PDT
From: Shrager.pa@Xerox.COM
Reply-To: Shrager.pa@Xerox.COM, shrager.PA@Xerox.COM
Subject: T.L. Popejoy: The Culture and Polotics of AI
To: ssl↑.PA@Xerox.COM, irl↑.PA@Xerox.COM
Line-Fold: no

Folks,

	T.L. Popejoy of Amherst College is conducting thesis research on
the "Culture and Politics of AI" under a Ford Foundation grant.  He has
contacted me in hopes of interviewing some of those at PARC who have
been involved in the socio-political aspects of AI such as CPSR, IRL
(i.e., our new view of the role of computers in learning and work), and
PARC as a progenitor of some of the important ideas in the field
including CSCW, HyperMedia, qualitative reasoning, planning, etc.

	He writes: "I am especially interested in learning about the
relationship between AI and the social/behavioral sciences which
surround AI-related projects.  What I am trying to piece together is how
the various institutional contexts of AI research (military, commercial,
academic) relate to one-another, and what the possible implications of
this relationship might be.  My work, therefore, consists of studying
both the engineering projects themselves, as well as the multitude of
discourses that surround the projects and give them meaning."

	T.L. seems to be an intelligent and articulate person, and I
think that he is well worth talking to if only for the potential of a
stimulating discussion of some topics that lurk behind our work and too
rarely surface.

	T.L. will be around the Bay Area for about a month, so there is
no rush to do this.  If you'd like to spend an hour or so with him, drop
me a note indicating approximately your preferred time (e.g., next week;
the end of July;...). I'll pass this information along to T.L. and he
will contact you directly to make appointments.

Thanks,
	Jeff

PS: If you know folks who aren't on SSL↑ or IRL↑ but whom T.L. might be
interested in talking with, e.g., at Stanford or SRI, you might pass
this note along.  I'm sure that more contacts can't hurt.

∂07-Jul-89  1434	JONES@Score.Stanford.EDU 	teaching CS306 in the fall   
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Jul 89  14:34:18 PDT
Date: Fri 7 Jul 89 14:34:27-PDT
From: "H. Roy Jones" <JONES@Score.Stanford.EDU>
Subject: teaching CS306 in the fall
To: jmc@Sail.Stanford.EDU
cc: : ;
Message-ID: <12508174463.18.JONES@Score.Stanford.EDU>


John,

We have you down for teaching CS306 this fall in Terman 156 on Tuesdays
and Thursdays from 2:45-4:00.  Is this correct?  In any case, what are
your teaching plans for 89-90?

Thanks,

Roy

-------

∂08-Jul-89  0206	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	480 million digits of pi
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 8 Jul 89  02:05:54 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 399360; Sat 8-Jul-89 04:28:25 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 100070; Sat 8-Jul-89 01:22:27 PDT
Date: Sat, 8 Jul 89 01:22 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
 Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: 480 million digits of pi
To: vanMeule@ALLEGHENY.SCRC.Symbolics.COM
cc: MLB@WHITE.SWW.Symbolics.COM, cwr@WHITE.SWW.Symbolics.COM, Math-Fun@WHITE.SWW.Symbolics.COM
In-Reply-To: <19890612173831.3.VANMEULE@PERTA.SCRC.Symbolics.COM>
Message-ID: <19890708082226.0.RWG@TSUNAMI.SPA.Symbolics.COM>
Character-Type-Mappings: (1 0 (NIL 0) (NIL :ITALIC NIL) "CPTFONTI")
Fonts: CPTFONT, CPTFONTI
bcc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM

[I'm back]
    Date: Mon, 12 Jun 89 13:38 EDT
    From: Andre van Meulebrouck <vanMeule@ALLEGHENY.SCRC.Symbolics.COM>

	Date: Mon, 12 Jun 89 09:48 PDT
	From: Marc Le Brun <MLB@WHITE.SWW.Symbolics.COM>

	    Date: Fri, 9 Jun 89 19:37 PDT
	    From: Craig W. Reynolds <cwr@WHITE.SWW.Symbolics.COM>

	    This may be Old News in mathematical circles, but I saw a report on
Chudnovsky
	    of Columbia University had computed 480 million digits of pi.

In my talk I called them the Columbian pi cartel.  I hope they took no offense,
as they were generous to me in their talk.  At least I kept their secret, even
from Salamin.  They used essentially my method, but with their series instead
of Ramanujan's.  When I castigated them about decimal (instead of c.f.), they
mumbled that although 10 is stupid, every hypergeometric number should have
at least one radix in which the digits are nonrandom.  So they may at least
attempt a moby radix conversion, when they decide to which.

	Who can expect to transcribe a Cyrillic name accurately (one book I have gave a production
	that results in 420 spellings for "Chebychev")?  

    That's sad to see that happen.  On the news, and even on Nightline, I've often
    heard the names of Soviet premieres and other dignitaries bungled, and a lot of
    other Soviet words, when it should be easy to ask any of the many Russians in
    this country for a pronounciation.  
                           ↑

The viewing public, in its insatiable craving for entertainment disguised as
information, has relentlessly selected a herd of airheads not only to
deliver the news, but also to write it.  (The problem is amplified by
multimedia conglomerates, who internally reprocess stories for their various
excretionary modalities.)  It has even reached the "educational" programming:
I recently saw a piece on Mesa Verde National Park wherein the Pueblo builders
were repeatedly called "Anastazi" instead of Anasazi (= aborigine in Navajo).

    Part of the problem is standardization--there are various transliteration
    systems floating around.

    I could probably locate my copy of the transliteration system most generally
    used by the military (and perhaps likely by the NSA and government too) if
    anyone was interested.  If you go by strict transliteration, you do get some
    horrid looking words that are very unEnglish looking and unEnglish to try to
    pronounce (as spelled), so in some cases, I'm sympathetic to fudging pure
    transliterations (except in transliterations of Soviet scientific literature in
    libraries).

    For "Chednofski", it should probably be "Chednovsky" (I'm sure about the "v" but
    there's some variation in -sky endings, though for Soviet Russian I've usually
    seen it as -sky rather than -ski, the latter I've seen more for Polish.).

    [...]

I think -(ov)ski is patronymic in Polish, the way -ovich is in Russian, while
Russian -sky is "provenonymic" (as in Minsky or Pinsky, or like the "von"
in von Dittersdorf(f)).  By Murphy's law, there is a city ending in "ov",
Chudnow, to make David and Gregory's name look Polish.  (Anybody know the
provenance of Penkovsky, as in Oleg?)

The preponderance of Jews among American owners of "-sky" surnames may be
more a consequence of "preferential" emigration than of correlation with
ethnicity.  The helmsman of the passenger liner in that Black Sea (or was
it Lake Baikal?) ship collision (~1986) was named Chudnovsky, and you can
bet he wasn't Jewish.  (Either that or he became man-at-the-helm
retroactively.)


Honey, I SHRUNK the kids??  Disney must be spinning in his grave.

            1CNN Headline News0 this morning that two brothers named Chednofski (?) 

∂08-Jul-89  2254	CLT 	CLT itinerary etc   

July 11-16
    Hotel <<La Fontana>>
    Luno lago 
    28049  Stresa Italy
    (0323)32707

    16th ICALP
    Conference Office  (8:30am - 6pm) Tue thru Sat July 15
    Palazzo dei Congressi
    Piazzale Europa, 3
    28049  Stresa Italy
    0323-31092

July 16
    Novotel Sophia-Antipolis
    Tel: 93 65 40 00

July 17-19
    Marc Hely 
    La Colle sur Loup
    tel: 93 22 64 10

    Visiting Gilles Kahn
    INRIA at Sophia-Antipolis
    2004 Route des Lucioles
    06565 Valbonne Cedex
    Tel: 93 65 77 77


Realestate law
  Duncan King  494-6000  8am  
fax: 494-3012
   2471 east bayshore #501
     cross 101 on embarcadero, turn right onto east bayshore betw mings and scotts
     park between office and mings, come in side facing mings


Geotechnologist
Joe Michelucci
Michelucci and assoc   468-3640  
(report is in folder)

Neil Springer (Springer Construction) 345-2471
29.3k for clearing termite report  (see folder)

Vini Joy
Western Construction and Engineering   366-1766
11.2k for drainage
18k for driveway
(verbal - hasn't sent formal bid yet - typist sick)

Michael Pedro
Ground Stability Construction Inc  408-453-7545 
bids for underpinning, drainage and driveway
(see folder)

Mark Almeida
Alcon  Contractors 652-0311
General contractors lic no 375006
(will call -- call him if you don't here by July 14)

∂08-Jul-89  2316	CLT 	to do
Get realestate lawyer to look at proposed agreement with
Keeley and hopefully get it signed.

Lean on Alcon man for bid if you don't hear.

∂09-Jul-89  1439	Mailer 	Once upon a flag 
To:   JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
From: Les Earnest <LES@SAIL.Stanford.EDU>

On July 3, John McCarthy posted a statement in support of the flag
deification amendment.  In response, I said:
. I find John's position curiously inconsistent on freedom of speech.
. He stoutly opposed censorship of rec.humor.funny at Stanford, yet
. he now supports censorship of those who would communicate their
. disrespect for the government by "desecrating" the flag.  I invite
. John to explain why censorship of bulletin boards is evil but
. censorship of public demonstrations is OK.
. 
. I wonder how John feels about political cartoons that deride the
. President?  Should they too be repressed under penalty of law, or
. does he feel the need to protect only abstract symbols of central
. authority?
. 
. How about the California state flag?  And should burning the Stanford
. flag be considered a violation of the Fundamental Standard?

In response, John said:
> I'm sorry I don't have time before my trip to Japan to explain
> fully the questions raised by Les Earnest and Ric Steinberger.

Now that John is back and is once again addressing the same issue, I
invite him to respond more fully to the issues raised above.

∂09-Jul-89  2211	Mailer 	re: Once upon a flag  
To:   su-etc@SAIL.Stanford.EDU, JMC@SAIL.Stanford.EDU
From: Les Earnest <LES@SAIL.Stanford.EDU>

JMC says:
> 	3. My main reason for supporting the amendment is that
> it would provide a precedent for overruling the Supreme Court
> when it manufactures law.

This sounds like a wonderful principle to demonstrate: that that any court
decisions made by those uppity Reagan appointees can be circumvented with
suitable legislation.

∂10-Jul-89  0412	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	math-funnel   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 10 Jul 89  04:11:54 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 399453; Mon 10-Jul-89 07:10:51 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 100408; Mon 10-Jul-89 03:56:42 PDT
Date: Mon, 10 Jul 89 03:56 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: math-funnel
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890710105639.4.RWG@TSUNAMI.SPA.Symbolics.COM>

Does anyone know the formula for the "gravitational"
infundibula that supposedly demonstrate orbits at science
musea?  Can they really work, even for frictionless point
marbles?  The centripetal force can't blow up at 0,
precluding a point gravity source, but then how many
black holes have *you* grazed, lately?  (And you
probably need parental consent before X-raying kids
when their marbles cross the event horizon.)

I'd like to say "relativity aside", but we seem to need
curved space here.  I.e., I don't think the surface can
deliver an r↑-2 radial force in room coordinates, because
the surface centripetally accelerates marbles as an increasing
function of their tangential velocity.  Thus, if you do manage
periodic orbits, they won't be Keplerian when projected onto
the floor.

I scribbled some equations on a notepad (which I misplaced before I
could check properly), claiming that the surface should have cross
section k = y s, where k is some negative constant depending on real
and simulated gravity, y is the height of the surface, and s is arc
length along a radius.  This leads to an elliptic integral when you
try for y(x), which would be fitting, if true.

Salamin would polish this off between gulps of sambar,
except that he's munching trail gorp in the Rockies someplace.

Analysis, anybody?

∂10-Jul-89  0930	Mailer 	re: Abortion
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Jul 89  09:30:52 PDT
Date: Mon 10 Jul 89 08:49:08-PDT
From: ABE DEANDA <A.ABIE@Macbeth.Stanford.EDU>
Subject: re: Abortion
To: JMC@SAIL.Stanford.EDU
cc: cjh@PORTIA.STANFORD.EDU, su-etc@SAIL.Stanford.EDU
In-Reply-To: <bpzaS@SAIL.Stanford.EDU>
Message-ID: <12508898033.14.A.ABIE@Macbeth.Stanford.EDU>

JMC writes:

>The average woman getting an abortion isn't white, although most likely
>the modal woman is white.  The average is somewhat off-white, unless
>there are some whiter than white in order to balance the blacks.

Whoops, I realize the semantical error in my original posting.  However,
my defense is that I was repeating a statement from the CDC's 1988 report
on abortion published in the MMWR (Morbidity and Mortality Weekly Report),
which was summarized in a December issue of the Journal of the American
Medical Association.

Abe
-------

∂10-Jul-89  1109	vanMeule@ALLEGHENY.SCRC.Symbolics.COM 	re: 480 million digits of pi   
Received: from ALLEGHENY.SCRC.Symbolics.COM ([128.81.41.45]) by SAIL.Stanford.EDU with TCP; 10 Jul 89  11:09:00 PDT
Received: from PERTA.SCRC.Symbolics.COM by ALLEGHENY.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 145723; Mon 10-Jul-89 13:27:31 EDT
Date: Mon, 10 Jul 89 13:27 EDT
From: Andre van Meulebrouck <vanMeule@ALLEGHENY.SCRC.Symbolics.COM>
Subject: re: 480 million digits of pi  
To: JMC@SAIL.Stanford.EDU, rwg@RUSSIAN.SPA.Symbolics.COM, vanMeule@ALLEGHENY.SCRC.Symbolics.COM
cc: MLB@WHITE.SWW.Symbolics.COM, cwr@WHITE.SWW.Symbolics.COM, vanmeule@ALLEGHENY.SCRC.Symbolics.COM
In-Reply-To: <bpZbz@SAIL.Stanford.EDU>
Message-ID: <19890710172728.5.VANMEULE@PERTA.SCRC.Symbolics.COM>

    Date: 08 Jul 89  1216 PDT
    From: John McCarthy <JMC@SAIL.Stanford.EDU>

    [In reply to message from rwg@RUSSIAN.SPA.Symbolics.COM sent Sat, 8 Jul 89 01:22 PDT.]

    The spelling is Chudnovsky, because that's the way the Chudnovsky's,
    U.S. citizens, spell it.

Yup, and I believe Chudnovsky is the correct transliteration as well (seeing how
it's spelled in Cyrillic could verify that).  I mistyped when I typed
"Chednovsky" (corrected in mail you probably didn't see).

[...]

∂10-Jul-89  1159	MPS  
Bob Keeley called - wants a copy of the purchase agreement
will bring a copy of the modification to your office this
week.  He will call first

Pat

∂11-Jul-89  0036	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	480 million digits of pi
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 11 Jul 89  00:36:20 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 399801; Tue 11-Jul-89 03:37:15 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 100598; Tue 11-Jul-89 00:27:15 PDT
Date: Tue, 11 Jul 89 00:27 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: 480 million digits of pi
To: Math-Fun@WHITE.SWW.Symbolics.COM
cc: "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890708082226.0.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890711072712.6.RWG@TSUNAMI.SPA.Symbolics.COM>
Character-Type-Mappings: (1 0 (NIL 0) (NIL :ITALIC NIL) "CPTFONTI")
Fonts: CPTFONT, CPTFONTI

    Date: Sat, 8 Jul 89 01:22 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

    [I'm back]
	Date: Mon, 12 Jun 89 13:38 EDT
	From: Andre van Meulebrouck <vanMeule@ALLEGHENY.SCRC.Symbolics.COM>

	    Date: Mon, 12 Jun 89 09:48 PDT
	    From: Marc Le Brun <MLB@WHITE.SWW.Symbolics.COM>

		Date: Fri, 9 Jun 89 19:37 PDT
		From: Craig W. Reynolds <cwr@WHITE.SWW.Symbolics.COM>

		This may be Old News in mathematical circles, but I saw a report on
		ε1CNN Headline Newsε0 this morning that two brothers named Chednofski (?)
    Chudnovsky
		of Columbia University had computed 480 million digits of pi.

    In my talk I called them the Columbian pi cartel.  I hope they took no offense,
    as they were generous to me in their talk.  At least I kept their secret, even
    from Salamin.  They used essentially my method, but with their series instead
    of Ramanujan's.

One interesting difference stems from the two huge storage penalties that attend
resumability.  One is unavoidable--the maintenance of the exact rational partial
sum, which contains several times as many digits as leading zeros in the first
neglected term.  The other is a useless, monster gcd (several million factorial)
that builds up in the matrix elements, except that I found the cost of removing
it exceeded the cost of tolerating it.

But the relative enormity of their undertaking forced the Chudnovskys to pay the
gcd price.
		     When I castigated them about decimal (instead of c.f.),

An irony is that you can get the c.f. as a by-product of the gcd, except that
you want the c.f. of pi, not the partial series sum, which is off by a rational
times sqrt(10005).  (Or sqrt(2), in the Ramanujan case.)
									     they
    mumbled that although 10 is stupid, every hypergeometric number should have
    at least one radix in which the digits are nonrandom.  So they may at least
    attempt a moby radix conversion, when they decide to which.

	    Who can expect to transcribe a Cyrillic name accurately (one book I have gave a production
	    that results in 420 spellings for "Chebychev")?  
 . . .

    I think -(ov)ski is patronymic in Polish, the way -ovich is in Russian, while
    Russian -sky is "provenonymic" (as in Minsky or Pinsky, or like the "von"
    in von Dittersdorf(f)).  By Murphy's law, there is a city ending in "ov",
    Chudnow, to make David and Gregory's name look Polish.  (Anybody know the
    provenance of Penkovsky, as in Oleg?)

I asked David, who says that Penkovsky means something like "dealer in hemp".
(Was he hanged or shot?)

And Tchaikovsky (obvious, when you think about it) presumably descended from
a tea merchant.  (Reminding me of my favorite Boston phone subscriber,
Esper K. Teebagy.)

∂11-Jul-89  0848	VAL 	re: parsing the past
[In reply to message rcvd 11-Jul-89 06:00-PT.]

I've never heard of continuous pattern matching. Perhaps what we match
isn't really continuous pasts, but their "finite summaries" of some kind.
For instance, if an object is moving along the trajectory x=f(t), then,
instead of the graph of f on  a subinterval [t0,t1], we use the tuple
<t0,f(t0),t1,f(t1)> as the "summary" that can be compared with a given
pattern.

∂11-Jul-89  0947	siegman@sierra.STANFORD.EDU 	Math Book  
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 11 Jul 89  09:47:07 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Mon, 11 Jul 88 09:47:40 PDT
Date: Mon, 11 Jul 88 09:47:40 PDT
From: siegman@sierra.STANFORD.EDU (Anthony E. Siegman)
To: jmc@sail
Subject: Math Book 
Message-Id: <CMM.0.88.584642859.siegman@>

In a recent msg on another topic, you mentioned Newman's 4-volume
series on the history of mathematics, which I've owned for many years.

I happened to see by accident on someone else's bookshelf since then a
large single volume by Edna E. Kramer, "The Nature and Growth of
Modern Mathematics", which appeared to be also interesting and perhaps
more up to date.

You're probably already aware of it, but if not might find it
interesting.

∂11-Jul-89  1453	betsy@russell.Stanford.EDU 	Arkady Blinov    
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Jul 89  14:52:43 PDT
Received: by russell.Stanford.EDU (4.0/inc-1.0)
	id AA11183; Tue, 11 Jul 89 14:58:21 PDT
Date: Tue 11 Jul 89 14:58:21-PDT
From: Betsy Macken <BETSY@CSLI.Stanford.EDU>
Subject: Arkady Blinov
To: jmc@sail.stanford.edu
Message-Id: <616197501.0.BETSY@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

Stanley told me that you had recommended that Arkady Blinov give
a talk at CSLI sometime in October.  Do you know yet which two
weeks in October he will be available?  The schedule for CSLI talks
is filling up, and I'd like to pin down a date for Blinov if possible.
Thanks,
Betsy
-------

∂11-Jul-89  1515	MPS  
Call Jack Cate - 321-1225

∂11-Jul-89  1606	RWF 	re: Constitutional Ideals
[In reply to message rcvd 08-Jul-89 16:55-PT.]

I have very little knowledge about El Salvador, and
probably should have been more careful about
accepting it as an example.

∂11-Jul-89  2000	JMC  
Camargo 493-1075

∂12-Jul-89  0801	ACW@ELEPHANT-BUTTE.SCRC.Symbolics.COM 	math-funnel
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 12 Jul 89  08:01:41 PDT
Received: from YUKON.SCRC.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 400160; Wed 12-Jul-89 10:40:22 EDT
Received: from TALLIS.SCRC.Symbolics.COM by YUKON.SCRC.Symbolics.COM via INTERNET with SMTP id 472128; 12 Jul 89 10:46:43 EDT
Date: Wed, 12 Jul 89 10:40 EDT
From: Allan C. Wechsler <ACW@YUKON.SCRC.Symbolics.COM>
Subject: math-funnel
To: rwg@RUSSIAN.SPA.Symbolics.COM, math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890710105639.4.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890712144044.3.ACW@TALLIS.SCRC.Symbolics.COM>

    Date: Mon, 10 Jul 89 03:56 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

    Does anyone know the formula for the "gravitational"
    infundibula that supposedly demonstrate orbits at science
    musea?  Can they really work, even for frictionless point
    marbles?  The centripetal force can't blow up at 0,
    precluding a point gravity source, but then how many
    black holes have *you* grazed, lately?  (And you
    probably need parental consent before X-raying kids
    when their marbles cross the event horizon.)

    I'd like to say "relativity aside", but we seem to need
    curved space here.  I.e., I don't think the surface can
    deliver an r↑-2 radial force in room coordinates, because
    the surface centripetally accelerates marbles as an increasing
    function of their tangential velocity.  Thus, if you do manage
    periodic orbits, they won't be Keplerian when projected onto
    the floor.

I think this is false.  If we are allowed to ignore the rotational
energy of the marble, the situation is much less grave than you imagine.
Just make the surface follow the potential function.  I think the laws
of motion are completely determined by the potential function.

    I scribbled some equations on a notepad (which I misplaced before I
    could check properly), claiming that the surface should have cross
    section k = y s, where k is some negative constant depending on real
    and simulated gravity, y is the height of the surface, and s is arc
    length along a radius.  This leads to an elliptic integral when you
    try for y(x), which would be fitting, if true.

    Salamin would polish this off between gulps of sambar,
    except that he's munching trail gorp in the Rockies someplace.

    Analysis, anybody?

∂12-Jul-89  0852	@Score.Stanford.EDU:perrault@ai.sri.com 	Meeting  
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Jul 89  08:51:58 PDT
Received: from Warbucks.AI.SRI.COM ([128.18.3.1].#Internet) by SCORE.STANFORD.EDU with TCP; Wed 12 Jul 89 08:52:22-PDT
Received: from agate.ai.sri.com by Warbucks.AI.SRI.COM with INTERNET ;
          Wed, 12 Jul 89 08:52:12 PDT
Received: by agate.ai.sri.com (4.1/4.16)
	id AA07849 for jmc@score.stanford.edu; Wed, 12 Jul 89 08:53:10 PDT
Date: Wed 12 Jul 89 08:53:06-PDT
From:     PERRAULT@Warbucks.AI.SRI.COM (Ray Perrault)
Subject: Meeting
To:       jmc@score.stanford.edu
cc:       perrault@Warbucks.AI.SRI.COM
Message-Id: <616261986.0.PERRAULT@ai.sri.com>
Mail-System-Version: <SUN-MM(229)+TOPSLIB(128)@ai.sri.com>

John,

I'm going to have to ask you whether we could move our meeting
of tomorrow at 4.  How about next week at 4 on any of Monday, Wed,
or Friday?

Ray
-------

∂12-Jul-89  0940	perrault@ai.sri.com 	re: Meeting        
Received: from Warbucks.AI.SRI.COM ([128.18.3.1]) by SAIL.Stanford.EDU with TCP; 12 Jul 89  09:40:30 PDT
Received: from agate.ai.sri.com by Warbucks.AI.SRI.COM with INTERNET ;
          Wed, 12 Jul 89 09:33:56 PDT
Received: by agate.ai.sri.com (4.1/4.16)
	id AA07962 for JMC@SAIL.Stanford.EDU; Wed, 12 Jul 89 09:34:48 PDT
Date: Wed 12 Jul 89 09:34:46-PDT
From:     PERRAULT@Warbucks.AI.SRI.COM (Ray Perrault)
Subject: re: Meeting    
To:       JMC@SAIL.Stanford.EDU
cc:       PERRAULT@Warbucks.AI.SRI.COM
Message-Id: <616264486.0.PERRAULT@ai.sri.com>
In-Reply-To: <nrXDw@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(229)+TOPSLIB(128)@ai.sri.com>

see you then.  --ray
-------

∂12-Jul-89  1056	Mailer 	re: Offshore Oil Drilling Ban   
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 12 Jul 89  10:55:52 PDT
Date: Wed, 12 Jul 89 10:56:17 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Offshore Oil Drilling Ban   
To: JMC@SAIL.Stanford.EDU
cc: DON@RML2.SRI.COM, comments@KL.SRI.COM, su-etc@SAIL.Stanford.EDU,
    don@RML2.SRI.COM
In-Reply-To: <RrXJ3@SAIL.Stanford.EDU>
Message-ID: <12509445468.23.STEINBERGER@KL.SRI.COM>

> From: John McCarthy <JMC@SAIL.Stanford.EDU>
> Subject: re: Offshore Oil Drilling Ban   

> Let the bastards freeze in the dark, but what about me?

That's a cute, but irrelevant remark.  The amount of oil off the coast
of California is uncertain, but most of the estimates I've seen weigh in
at something like a few weeks to a few months of *current* US consumption.
And so the question ought to become, "Is it worth the environmental risks
for such an amount of oil?"  While considerable profits can be made from
such a quantity of oil, no long term solution to America's fossil fuel
dependency is addressed.

BTW - You won't be freezing in the dark.  Most California houses are
heated with natural gas, much of it piped in from other states.

-ric steinberger

-------

∂12-Jul-89  1142	njacobs@vax.darpa.mil 	task:  DD   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 12 Jul 89  11:42:40 PDT
Received: from sun41.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA12447; Wed, 12 Jul 89 14:36:24 EDT
Posted-Date: Wed 12 Jul 89 14:40:02-EDT
Received: by sun41.darpa.mil (4.1/5.51)
	id AA01722; Wed, 12 Jul 89 14:40:04 EDT
Date: Wed 12 Jul 89 14:40:02-EDT
From: Nicole L. Fields <NJACOBS@vax.darpa.mil>
Subject: task:  DD
To: clt@sail.stanford.edu, jmc@sail.stanford.edu
Cc: NJACOBS@vax.darpa.mil
Message-Id: <616272002.0.NJACOBS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>

This message is a repeat of the message that I sent to you about two
weeks ago.  Many of the messages sent at that time seem to have
disappeared into the ether.

The "Subject:" line of this message contains the task code for your
project.  The "To:" line contains the addresses we currently have on
file for official email correspondence with you. 

Please confirm that the email address we are using is relaible.  We
prefer that you use a virtual mailbox that will direct mail so we can
get responses even when PIs are travelling.

Please acknowledge receipt of this message right now.  Thanks!
			      Nicole Jacobs




-------

∂12-Jul-89  1143	njacobs@vax.darpa.mil 	task:  DF   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 12 Jul 89  11:42:55 PDT
Received: from sun41.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA12457; Wed, 12 Jul 89 14:36:38 EDT
Posted-Date: Wed 12 Jul 89 14:38:54-EDT
Received: by sun41.darpa.mil (4.1/5.51)
	id AA01716; Wed, 12 Jul 89 14:38:56 EDT
Date: Wed 12 Jul 89 14:38:54-EDT
From: Nicole L. Fields <NJACOBS@vax.darpa.mil>
Subject: task:  DF
To: rpg@sail.stanford.edu, jmc@sail.stanford.edu
Cc: NJACOBS@vax.darpa.mil
Message-Id: <616271934.0.NJACOBS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>

This message is a repeat of the message that I sent to you about two
weeks ago.  Many of the messages sent at that time seem to have
disappeared into the ether.

The "Subject:" line of this message contains the task code for your
project.  The "To:" line contains the addresses we currently have on
file for official email correspondence with you. 

Please confirm that the email address we are using is relaible.  We
prefer that you use a virtual mailbox that will direct mail so we can
get responses even when PIs are travelling.

Please acknowledge receipt of this message right now.  Thanks!
			      Nicole Jacobs




-------

∂12-Jul-89  1341	@Gang-of-Four.Stanford.EDU:arg@lucid.com 	new-qlisp    
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Jul 89  13:41:49 PDT
Received: from LUCID.COM by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA08889; Wed, 12 Jul 89 13:39:29 PDT
Received: from bhopal ([192.43.178.13]) by heavens-gate id AA19957g; Wed, 12 Jul 89 13:38:51 PDT
Received: by bhopal id AA13895g; Wed, 12 Jul 89 13:40:36 PDT
Date: Wed, 12 Jul 89 13:40:36 PDT
From: Ron Goldman <arg@lucid.com>
Message-Id: <8907122040.AA13895@bhopal>
To: qlisp@go4.stanford.edu
Subject: new-qlisp

really-new-qlisp is now just regular new-qlisp.

∂12-Jul-89  1506	DON@RML2.SRI.COM 	Re: Ban on Offshore Oil Drilling
Received: from RML2.SRI.COM by SAIL.Stanford.EDU with TCP; 12 Jul 89  15:06:32 PDT
Date: Wed 12 Jul 89 15:06:43-PST
From: Don Cooley <DON@RML2.SRI.COM>
Subject: Re: Ban on Offshore Oil Drilling
To: comments@KL.SRI.COM, su-etc@shelby.stanford.edu
Cc: jmc@SAIL.STANFORD.EDU, don@RML2.SRI.COM
Message-ID: <616284403.550000.DON@RML2.SRI.COM>
Mail-System-Version: <VAX-MM(243)+TOPSLIB(135)@RML2.SRI.COM>


[In reply to message from JMC@SAIL.STANFORD.EDU sent Wed 12 Jul 89 09:33]

>>Let the bastards freeze in the dark, but what about me?

   One very good reason for a ban on offshore oil drilling:

   For decades the prevailing assumption has been that a fossil-fuel-based 
   economy would be constrained by oil, gas, and coal depletion.
   Logical enough. But global warming has turned that paradigm on its head:
   It now appears that the atmosphere's ability to assimilate fossil-fuel
   wastes [CO2, etc.] will be the limiting factor. The question is no
   longer how much oil, gas, and coal we have, but how much we can afford
   to burn before the greenhouse effect results in global warming of
   such a large magnitude it endangers human society as we know it today.
   (Not to mention nonhuman life.)

   I would rather look to conservation, better energy use efficiency,
   and renewables (solar, wind, and biomass) than to offshore oil
   drilling. According to Amory Lovins, "since 1973 the U.S. has obtained
   seven times as much energy from efficiency savings as from ALL increases
   in energy supply." 

   Does this mean I endorse nuclear fission?  No.  Again, I quote Lovins:
  "The U.S. has spent a trillion dollars on nuclear, which now provides 
   half as much energy as wood. Let's back winners, not losers."

   _Don Cooley

   Some of the above is from an article by James R. Udall in the
   July/August 1989 issue of Sierra magazine called "Climate
   Shock."
-------

∂12-Jul-89  1731	scherlis@vax.darpa.mil 	TECHNICAL REPORTS    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 12 Jul 89  17:31:03 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA13961; Wed, 12 Jul 89 20:10:28 EDT
Posted-Date: Wed 12 Jul 89 20:00:10-EDT
Received: by sun45.darpa.mil (4.1/5.51)
	id AA06068; Wed, 12 Jul 89 20:00:14 EDT
Date: Wed 12 Jul 89 20:00:10-EDT
From: William L. Scherlis <SCHERLIS@vax.darpa.mil>
Subject: TECHNICAL REPORTS
To: SW-PI@vax.darpa.mil
Cc: njacobs@vax.darpa.mil
Message-Id: <616291210.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>


To all DARPA/ISTO Software and Algorithms PIs:

	ANNUAL TECHNICAL REPORTS

This is the time of year when we require information from contractors
concerning accomplishments for the current fiscal year (FY) and
objectives for the next FY.  This information is used to assist us in
planning incremental funding for those efforts that are expected to
continue into the next FY (which begins 1 Oct 89).  This information
will also be used to brief the incoming Director of ISTO.  (Note that
this position remains unfilled.  We are actively seeking candidates.
If you have nominations please contact me or, if you wish, Craig
Fields directly.)

WHAT IS REQUIRED.  This year, we expect that only two email reports
will be solicited:
	(1) A single annual technical summary.
	(2) Financial summaries.

NOTE: These email reports are in addition to the reports required as
contractual deliverables, and this informal request does not waive any
contractual requirement.

FINANCIAL SUMMARIES.  The format for the financial summaries will be
uniform throughout the entire ISTO community, in order that we can use
automatic means to insert the data you provide into our financial
database.  Details of the format and deadlines will be provided in a
separate message.  PLEASE SEND FINANCIAL SUMMARIES SEPARATELY.

TECHNICAL SUMMARIES.  The technical summaries provide us with an
up-to-date view of the state of activity in our community.  The
challenge is to be concise yet substantive.  Responses are needed by
the morning of 20 July.

PLEASE ADHERE to the format below.  Specifically, (1) do not include
markup commands from a text processor (except in formulas that are
especially complex), and (2) do not use leading indentation or any
other extra embedded horizontal whitespace.  I suggest grabbing this
text with a text editor and filling in the blanks.

Here is what is required:
================================================================

(1) BASIC DATA.

(1.a) DARPA/ISTO project code, ARPA Order number, agent, contract number.

Example: AA, 1111, SPAWAR, N0037-C-0004
The first of these is the "task code" you were just sent.  (Call
Nicole Jacobs if you don't receive this by noon, 13 July.)

(1.b) Institution.

The institution contracting with the government.

(1.c) Project title. 

One line project title.

(1.d) Project mailbox.  

An internet address for official project email.  This should be active
even when PIs are travelling.  At worst, we will accept a list of
addresses. 

(1.e) PIs. 

For each: name, phone numbers, email address.

(2) PROJECT SUMMARY.

A short paragraph (approximately four sentences) outlining the
specific need for your effort, the opportunity it represents, and the
results to be delivered.  Your summary should indicate the expected
impact of the project, i.e., how future technologies will be different
because of this investment by DARPA.

(3) APPROACH.

A short paragraph describing the overall approach taken to research.
That is, what actual steps will be taken to achieve research goals?

(4) ACCOMPLISHMENTS. 

Two to four concrete accomplishments for fy89.  Emphasize technical
results with externally recognizable impact.  If you achieved a major
milestone of broad community impact, indicate so and describe it.
(There should be such an event every year or two.)

(5) OBJECTIVES.

Two to four for fy90, each as a short (two to three sentence)
paragraph.  Emphasize both technical results and impact.  Be specific.

(6) TECHNOLOGY TRANSFER progress and plans.  

(6.a) Customers. 

Customers for your results, whether the results be theorems or
software components or software interface definitions.  (Include
potential users outside of your research group with whom you have had
discussions as well as actual current users.  Include both research
groups and development groups.)  This and the other responses in this
section should be only two or three sentences each.

(6.b) Interfaces and consensus.

Opportunities pursued for accelerating community consensus, where
needed and where possible, for systems interface definitions that your
work creates, contributes to, and/or depends on.  (Any producer of a
large system prototype should address this issue.  At the research
prototype stage, efforts involving collaboration among research groups
to agree on component interfaces should be described.)

(6.c) Sources of technology.

Major producers of technology that you rely upon, including technical
results, interface definitions, and systems components.  (Do not
mention common commercial components.  You might mention Mach or the
Boyer-Moore theorem prover, but you wouldn't mention DEC or Sun
workstations.)

(7) OTHER INFORMATION.

(7.a) Major personnel changes.   

New hires, departures, etc.

(7.b) Major recent publications, honors, etc.

Send us copies of important publications once in a while.

================================================================

WHAT TO DO.

One report should be sent for each independently funded effort (i.e.,
each project code).  If you receive multiple copies of this message, it
is probably because you are involved with multiple funded efforts.  If
there is any question about what constitutes an "independently funded
effort," please call Nicole or me.

Send reports by email directly to Nicole Jacobs at DARPA:
	njacobs@vax.darpa.mil		(202)694-5800

If you do not receive an acknowledgement within a week, call Nicole to
ensure she has received your message.

Follow the format guidance for your responses.

MAILING ADDRESSES.  Help us keep our mailing list current.  We often
send official correspondence by email.  The best way to do this is for
you to have a virtual project mailbox at your site that will forward
to PIs and other responsible people.  Again, if there is a question,
call Nicole and check to see what address we have on file.

Deadline: Please respond by the morning of 20 July.  Early responses
are requested.  If you MUST respond late, call us first.  Thanks.

Finally: Many of your contracts require regular reports to be sent to
the contracting agent and to certain parts of DARPA.  Please also send
one copy of the reports directly to me at DARPA.

Thanks,
				Bill 

-------

∂12-Jul-89  2229	Mailer 	re: Offshore Oil Drilling Ban   
To:   STEINBERGER@KL.SRI.COM, JMC@SAIL.Stanford.EDU
CC:   su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

[In reply to message from STEINBERGER@KL.SRI.COM sent Wed, 12 Jul 89 10:56:17 PDT.]

OK, I grant that we know the oil off the coast of CA is a few weeks
of US consumption.  Now what fraction of current US pollution is
the expected pollution resulting from drilling off CA?
 

∂13-Jul-89  0019	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	math-funnel   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  00:18:58 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 400484; Thu 13-Jul-89 03:16:55 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 101010; Wed 12-Jul-89 23:43:47 PDT
Date: Wed, 12 Jul 89 23:43 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: math-funnel
To: ACW@YUKON.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890712144044.3.ACW@TALLIS.SCRC.Symbolics.COM>
Message-ID: <19890713064339.3.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Wed, 12 Jul 89 10:40 EDT
    From: Allan C. Wechsler <ACW@YUKON.SCRC.Symbolics.COM>

	Date: Mon, 10 Jul 89 03:56 PDT
	From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

	Does anyone know the formula for the "gravitational"
	infundibula that supposedly demonstrate orbits at science
	musea?  Can they really work, even for frictionless point
	marbles?  The centripetal force can't blow up at 0,
	precluding a point gravity source, but then how many
	black holes have *you* grazed, lately?  (And you
	probably need parental consent before X-raying kids
	when their marbles cross the event horizon.)

	I'd like to say "relativity aside", but we seem to need
	curved space here.  I.e., I don't think the surface can
	deliver an r↑-2 radial force in room coordinates, because
	the surface centripetally accelerates marbles as an increasing
	function of their tangential velocity.  Thus, if you do manage
	periodic orbits, they won't be Keplerian when projected onto
	the floor.

    I think this is false.  If we are allowed to ignore the rotational
    energy of the marble,

(point marbles presumably have none)

			  the situation is much less grave than you imagine.
    Just make the surface follow the potential function.  I think the laws
    of motion are completely determined by the potential function.

Analysis by Salamin:  Seductive, but dubitable.  Your surface will produce
the correct r(t), but it would be a miracle if theta(t) had the same period.
Put another way, suppose you had your z=-1/r system alongside another
which really was Keplerian (say, a perflectly flat surface with a frictionless
puck with an electrostatic charge opposite that of a central "sun", and no
magnetism (= relativity).)  Now suppose these systems behaved identically
when projected onto the floor.  Look at the kinetic energies.  They agree
at apoapsis and periapsis, but they'll disagree everywhere else, because
there will be a z velocity component for the funnel system, yet the x and y
components already constitute the correct Keplerian energy.

We think the only way to win is via a radial stretch function that
distorts the conics into the curved surface.  This should be around
page minus thirteen of Misner, Thorne & Wheeler.

(RWG:  By the way, these exhibits work poorly enough that it is possible
that they incorrectly use hyperboloids, with blame for the error misplaced
on friction and the finiteness of the marbles.  After all, what do they
mean by "parabolic" where they say "Place marble on this ramp to see a
parabolic trajectory."?  If Salamin and I are right, there could be at most
one trajectory with a parabolic projection.  Also, the direction of the
ramp is fixed.  If the exhibit really worked, you should be able to pivot
the ramp about its apex, yet always get parabolas, since they are the
trajectories of zero net energy.)

	I scribbled some equations on a notepad (which I misplaced before I
	could check properly), claiming that the surface should have cross
	section k = y s,

k = z s
			 where k is some negative constant depending on real
	and simulated gravity, y
z
				 is the height of the surface, and s is arc
	length along a radius.  This leads to an elliptic integral when you
	try for y(x),
z(r),
		      which would be fitting, if true.

Notice the similarity, though:  z= -1/s  vs  z = -1/r.

	Salamin would polish this off between gulps of sambar,
	except that he's munching trail gorp in the Rockies someplace.

He got dragged back for an emergency project.

	Analysis, anybody?

Maybe it's time to call the Exploratorium.

∂13-Jul-89  0731	Mailer 	re: Offshore Oil Drilling Ban   
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  07:30:59 PDT
Date: Thu, 13 Jul 89 07:31:26 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Offshore Oil Drilling Ban   
To: RWF@SAIL.Stanford.EDU
cc: JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
In-Reply-To: <prc$o@SAIL.Stanford.EDU>
Message-ID: <12509670320.16.STEINBERGER@KL.SRI.COM>

> From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

> OK, I grant that we know the oil off the coast of CA is a few weeks
> of US consumption.  Now what fraction of current US pollution is
> the expected pollution resulting from drilling off CA?
 
Of course, this is not question with a known answer, but rather an exercise
in risk prediction and pollution definition.  One would like to know
what are the risks associated with offshore drilling, and what is the
compensation in oil for taking the risks.  

Here's an interesting hypothetical question for Alaskans:  If it were a
virtual certainty that a major oil spill would occur every 10 - 15 years
in Alaska, would you, as a state resident, favor proceeding with the
pipeline project [might have been asked in early 1970s]?

-------

∂13-Jul-89  0754	Mailer 	re: Ban on Offshore Oil Drilling
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  07:54:53 PDT
Date: Thu, 13 Jul 89 07:55:26 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Ban on Offshore Oil Drilling
To: JMC@SAIL.Stanford.EDU
cc: DON@RML2.SRI.COM, comments@KL.SRI.COM, su-etc@SAIL.Stanford.EDU
In-Reply-To: <ps0d4@SAIL.Stanford.EDU>
Message-ID: <12509674689.16.STEINBERGER@KL.SRI.COM>

> From: John McCarthy <JMC@SAIL.Stanford.EDU>


> Energy obtained from saving cannot readily be estimated
> and is readily exaggerated.  

While it's possible to exagerate almost anything, it's incorrect to state the
the energy obtained from conservation cannot be estimated.  In fact there are
many cases where it can be precisely estimated.  For example, if I replace a
75-Watt incandescent bulb with a newer fluorescent one that uses only 18 Watts
(and lasts 10 times as long), but provides an equivalent amount of luminosity,
it's a relatively easy exercise to compare expected bulb lifetimes, energy
consumptions and costs/savings.


> In fact the ``renewable sources'' have generated little energy
> in spite of vast research investment since 1973.  There are no
> even plausible ideas.

It is unbelievable that you can critize energy conservation activists for
exageration and then come up with the above statement.  I suppose designing
houses with southern exposures, improved insulation, solar water heaters,
double or triple pane windows are bad ideas that cost the US government
monies that ought to have been invested in nuclear fission.  Now here's
an industry that required Federal legislation limiting insurance liability
(Price-Anderson Act) just to prode private companies to build reactors.
And if you want to speak of 'vast research investment', how many dollars
has the federal government devoted to commercial fission/fusion research
since 1945?  Compare that with the federal 'dole' to solar/wind/architectural
researches.


> France gets 75 percent of its electricity from nuclear energy,
> and Japan has just passed the U.S. in percentage.  France has
> the best estimates of the cost.

And I suppose these countries can safely store radioactive wastes for tens
of thousands of years.  In other words, the true costs of nuclear power
generation will remain unknown well into the future.

> It is disappointing that the Sierra Club is still publishing
> those old lies.  Blind opposition to progress, I guess.

What is really disappointing is to see highly-talented people venomizing
people or organizations with perspectives different from their own.
-------

∂13-Jul-89  0822	DON@RML2.SRI.COM 	re: Ban on Offshore Oil Drilling
Received: from RML2.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  08:22:48 PDT
Date: Thu 13 Jul 89 08:22:59-PST
From: Don Cooley <DON@RML2.SRI.COM>
Subject: re: Ban on Offshore Oil Drilling
To: comments@KL.SRI.COM, su-etc@shelby.stanford.edu, ric@RML2.SRI.COM, 
    don@RML2.SRI.COM
Cc: jmc@SAIL.STANFORD.EDU
Message-ID: <616346579.510000.DON@RML2.SRI.COM>
Mail-System-Version: <VAX-MM(243)+TOPSLIB(135)@RML2.SRI.COM>


Date: 13 Jul 89  0055 PDT
From: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: Ban on Offshore Oil Drilling
To:   DON@RML2.SRI.COM, comments@KL.SRI.COM,
      su-etc@SAIL.Stanford.EDU  

[In reply to message from JMC@SAIL.STANFORD>EDU sent Thur 13 Jul 89]

 >>[deleted stuff]
 >>It is disappointing that the Sierra Club is still publishing
 >>those old lies.  Blind opposition to progress, I guess.

  What can I say, you think most (all?) of what the Sierra Club
  is saying is a pack of lies.  Fine. I don't agree, however.

  You also think that anyone that doesn't agree with your pro-nuclear
  fission position is opposed to progress.  I can only say,
  some of us don't believe that increasing our dependence on
  nuclear fission is progress. 

  In fact, I deeply resent the attitude of the pro-nuke group that
  anyone who opposes nuclear fission are anti-technology, or that
  they have an irrational fear of radiation (probably because
  they are too stupid to understand nuclear physics). There
  are many pro-technology, physics literate people who oppose
  nuclear power (as it currently exists) for a variety of
  rational, scientific, and economic reasons. But, why rehash all this? 
  (See Ric's previous comment, or a variety of comments by Ric, myself,and
   others over the past 6 months or so to su-etc and comments@kl.sri.com.)

  My original posting was to encourage those opposed to offshore oil drilling
  to telegram key senators to support the ban. Those in
  favor of offshore oil drilling can telegram the same senators 
  and voice their support of drilling.  But, nobody has given me a good
  reason to be in favor of offshore oil drilling.  All I got
  from you is "let the bastards freeze in the dark."


  -Don Cooley
 
-------

∂13-Jul-89  1009	chandler@Polya.Stanford.EDU 	Your vote on Karp    
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Jul 89  10:09:02 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA22458; Thu, 13 Jul 89 10:09:13 -0700
Date: Thu, 13 Jul 1989 10:09:12 PDT
From: "Joyce R. Chandler" <chandler@Polya.Stanford.EDU>
To: feigenbaum@sumex-aim, goldberg@Polya.Stanford.EDU, guibas@dec.com, zm@sail,
        jmc@sail, winograd@csli
Cc: chandler@Polya.Stanford.EDU
Subject: Your vote on Karp
Message-Id: <CMM.0.88.616352952.chandler@Polya.Stanford.EDU>

The Academic Council has voted in the past to make an offer to Dick Karp
provided we could find the necessary billet.  We have 1/2 billet available
and Operations Research is voting today to combine 1/2 of their billets with
ours to create the necessary billet.  At the 7/11 general faculty meeting it
was moved and seconded that Karp be offered a faculty position.  The motion
passed unanimously.  Please give me your vote at your earliest opportunity.

Thanks again!!!

∂13-Jul-89  1024	chandler@Polya.Stanford.EDU 	re: Your vote on Karp     
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Jul 89  10:24:12 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA23098; Thu, 13 Jul 89 10:24:42 -0700
Date: Thu, 13 Jul 1989 10:24:41 PDT
From: "Joyce R. Chandler" <chandler@Polya.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: Your vote on Karp 
In-Reply-To: Your message of 13 Jul 89 1018 PDT 
Message-Id: <CMM.0.88.616353881.chandler@Polya.Stanford.EDU>

So noted.  Thanks John.

∂13-Jul-89  1151	ME 	failed mail returned 
The following message was undeliverable to the address(es) below
because the destination Host Name(s) are Unknown:
davef@Jessica.UUCP

[ME - I don't know where this .UUCP host name came from, but if
this is the Stanford host Jessica, you can just omit the ".UUCP"
and remail to davef.  (The mail went to su-etc OK.)]

------- Begin undelivered message: -------
 13-Jul-89  1136	Mailer 	re: police brutality and OR
To:   davef@Jessica.UUCP, su-etc@SAIL.Stanford.EDU   
From: John McCarthy <JMC@SAIL.Stanford.EDU>

[In reply to message from davef@Jessica.UUCP sent 13 Jul 89 16:18:15 GMT.]

What should be done to Earth First is the same as what was done
to the Ku Klux Klan.  A civil suit established that the Klan's
activities had caused a murder, and the judgment was large
enough to put the Klan out of business in a certain state.

------- End undelivered message -------

∂13-Jul-89  1321	DON@RML2.SRI.COM 	Nuclear Power    
Received: from RML2.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  13:21:32 PDT
Date: Thu 13 Jul 89 13:21:41-PST
From: Don Cooley <DON@RML2.SRI.COM>
Subject: Nuclear Power
To: comments@KL.SRI.COM, su-etc@shelby.stanford.edu
Cc: jmc@SAIL.STANFORD.EDU, don@RML2.SRI.COM
Message-ID: <616364501.460000.DON@RML2.SRI.COM>
Mail-System-Version: <VAX-MM(243)+TOPSLIB(135)@RML2.SRI.COM>


  "Fission energy is safe only if a number of critical devices work
   as they should, if a number of people in key positions follow
   all of their instructions, if there is no sabotage, no hijacking
   of transports, if no reactor fuel processing plant or waste
   repository anywhere in the world is situated in a region of riots
   or guerrilla activity, and no revolution or war--even a
   "conventional one"--takes place in these regions. The enormous
   quantities of extremely dangerous material must not get into
   the hands of ignorant people or desperados. No acts of God can
   be permitted."

   Quote from Nobel Prize winning physicist Hannes Alfen.
   (Gee, I guess not all experts are pro-nuke afterall.)

   -Don Cooley   "I'd rather freeze in the dark than glow in the dark."
    and "I'd rather take my chances in Ted Kennedy's car, than
    at Chernobyl or Rocky Flats, Colorado, after all, Ted's been 
    driving for 40 years and only lost one passenger."

    I just had a great idea: How about building a nuclear power
    plant right on the Stanford campus. There is lot's of room
    along Palm Drive, and maybe it would drive down home prices
    in the area that where forced up by those nasty Sierra Club
    people. I can see it now, nice 4 bedroom, 3 bath home,
    on 1/2 acre lot, within walking distance to Stanford.
    Was $1,200,000 , now only $67,234! Not only that, but electricity
    rates should fall too, after 20 years of customers paying double
    to pay for the construction of the nuclear power plant.
    Nuclear Engineering majors at Stanford would also
    no longer have to go to France to study, they could just
    walk down to the friendly campus nuclear power plant.
-----

    
-------

∂13-Jul-89  1332	Mailer 	re: Ban on Offshore Oil Drilling
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  13:32:29 PDT
Date: Thu, 13 Jul 89 13:32:55 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Ban on Offshore Oil Drilling
To: JMC@SAIL.Stanford.EDU
cc: DON@RML2.SRI.COM, comments@KL.SRI.COM, su-etc@SAIL.Stanford.EDU
In-Reply-To: <bspVo@SAIL.Stanford.EDU>
Message-ID: <12509736125.21.STEINBERGER@KL.SRI.COM>

> From: John McCarthy <JMC@SAIL.Stanford.EDU>

> 1. It is intentionally confusing to regard conservation as a
> source of energy.  It involves using less, but it doesn't supply
> energy.

If I can do a given task with X Joules that used to require Y (Y > X) Joules,
is that not thermodynamically equivalent to producing Y - X Joules of energy?

>  As for solar water heaters, I recently bought
> a house that has one.  It doesn't work, and the company that
> sold it is out of business.  Maybe it can be fixed, but since
> it has a natural gas backup, there's no urgency.
> I'll bet a large fraction of solar hot water systems no longer
> work.

I'm sorry your water heater doesn't work properly.  However that's not a
thorough inductive proof that the majority of solar water heaters are
improperly designed.  Israel has a mature solar water heater industry and many
residential buildings there have these devices installed.
	I agree that, at the moment, while fossil fuels remain relatively
inexpensive in the U.S., many solar-based strategies are often uneconomic.
However, given the past 20 years experience with rising energy prices and
diminishing and uncertain supplies, this is likely to be a static condition.
	Conserving existing supplies of energy remains the most attractive
least expensive method for saving money and energy.

> 3. Yes, France and Japan have figured out how to store the waste.
> All the proposals will work, but implementation needs to be
> delayed to allow further decay of heat generating radioactivity.
> After 500 years the radioactivity will be less than that of the
> original uranium ore, so throwing the waste into uranium mines is
> a workable, if far from optimal, solution.  The anti-nukes have
> succeeded in delaying U.S. decisions on waste disposal by their
> usual dishonest tactics of exaggerating hypothetical dangers.

Sounds like reasoning based more on faith and hope than solid experience.


> 4. It is wrong, and usually dishonest, to say or imply that
> there is substantial scientific controversy about the overall
> safety of nuclear energy.  The major scientific organizations
> that have made studies of relevant issues, e.g. the American
> Physical Society and the National Research Council, have
> all determined that the problems they studied were solvable
> and that American reactors were reasonably safe.

Other organizations, like the Union of Concerned Scientists, disagree.
"Reasonably safe" is an interesting concept, and certainly subject to
interpretation.  Would you like to live 1/2 mile from a reacror deemed
by the NRC to be "reasonably safe."

> 5. I'll be back on the Sierra Club later.  It has done immense
> harm, and it and its ideological allies have a major
> responsibility for the fact that while my generation could
> purchase homes out of current income, most of the current
> generation of young people cannot but must wait for their parents
> to die.

I suppose you could blame the Sierra Club for the failure of the US to
win more medals at the Seoul Olympics, but what's the point?  Blaming
the Sierra Club for high housing prices is a an attempt to displace
anger and frustration for real estate conditions.  If that organization
were really so influential, I'd probably be opposing them as well.
-------

∂13-Jul-89  1334	siegman@sierra.STANFORD.EDU 	Re: Ban on Offshore Oil Drilling    
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 13 Jul 89  13:34:41 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Thu, 13 Jul 89 13:35:06 PDT
Date: Thu, 13 Jul 89 13:35:06 PDT
From: siegman@sierra.STANFORD.EDU (Anthony E. Siegman)
To: JMC@SAIL.STANFORD.EDU
Subject: Re: Ban on Offshore Oil Drilling
Newsgroups: su.etc
In-Reply-To: <bspVo@SAIL.Stanford.EDU>
Organization: Stanford University
Cc: 

In article <bspVo@SAIL.Stanford.EDU> you write:

>1. It is intentionally confusing to regard conservation as a
>source of energy.  It involves using less, but it doesn't supply
>energy.

I can't agree that it's even confusing, much less "intentionally
confusing".

There's a certain minor degree of rhetorical license in referring to
conservation as a "source"; but the underlying point is important,
valid, and not misunderstood.  We could live every bit as well, maybe
a bit better, with substantially less energy consumption (and
pollution and waste), if we just put our minds to it.

As a minor but nontrivial example, I think of some huge enormous
10-mpg Detroit-iron boats I've driven in earlier years, that had less
interior space, less comfort, and less pep than the 35-mpg Honda
Accords I later owned (and that were much more awkward to garage and
park).  Pressures for conservation played a major role in making this
newer kind of car available; Detroit's mentality would never have
produced without this.

∂13-Jul-89  1347	MPS  
Hi

Don DeHaven, NSF, called regarding your proposal with
US-Japan.  He is going to be in his office for the next
15 minutes only.  He would like you to call.  Otherwise,
when can he reach you tomorrow?

202-357-7494

Peter Rathmann would like to make an appointment with you.
You had told him previously to call and make an appointment.
He can see you anytime tomorrow, except between 3:15 and 4:15
when he will be at a seminar.  His e-mail is PFR@sail.

Thaks.

Pat

∂13-Jul-89  1443	PKR 	meeting on equality circumscription
To:   JMC
CC:   PKR   
[In reply to message rcvd 13-Jul-89 13:55-PT.]

I sent you a message last week, and I was kind of vague on the
topic for the meeting.  A few months ago I gave a seminar on
circumscribing equality (the seminar which Vladimir
coordinates.)  At the time you had some interesting comments on
fixing up the axiom and so on.

That work has progressed some, and I wanted to show you the new
stuff, as well as get any opinions or suggestions you might have
for where to take it.  In short, I guess I am asking for the
standard kind of student/professor meeting.

				-Peter

∂13-Jul-89  1530	Mailer 	re: Nuclear energy    
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  15:30:42 PDT
Date: Thu, 13 Jul 89 15:31:07 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Nuclear energy    
To: JMC@SAIL.Stanford.EDU
cc: jester@JESSICA.STANFORD.EDU, su-etc@SAIL.Stanford.EDU
In-Reply-To: <12sqK1@SAIL.Stanford.EDU>
Message-ID: <12509757644.21.STEINBERGER@KL.SRI.COM>

> From: John McCarthy <JMC@SAIL.Stanford.EDU>

> Steinberger implies that the damage to humans from radiation releases
> associated with nuclear energy has not been estimated and might be
> very large.  On the contrary, it has been estimated accurately, and
> rather low upper bounds can be given. 

I am claiming that studies have been done in addition to Cohen's that
do in fact conclude that thousands of people who were living in the 
Harrisburg, PA, region during the Three Mile Island accident will eventually
develop cancer.  The point here is not that the expert studies I read are
correct while the ones JMC reads are not.  The point is that it is still
an open issue with reputable scientists taking different sides.



>	This is an entirely different statistical result
> than was obtained when smoking and cancer were related.

No really so.  For decades, many physicians and citizens were on both
sides of the fence on the issue of smoking and lung cancer.  It was only
in 1962 (or thereabouts) when the US government formally concluded that
a firm connection between smoking and lung cancer existed.  The tobacco
industries argue to this day that a "connection" is not a causality
and continue to deny that smoking can *cause* cancer.
    We know that some kinds of radiation can and do result in cancers
developing in exposed individuals.  Still being researched is the precise
amounts of various types of radiation, the type and length of exposure, the
age of the exposed persons and other statistical factors.  We do know that
nuclear reactors can accidently release radioactive materials into the air and
sourrounding waters.  We do know that highly radioactive wastes are produced
as a result of normal reactor operations.  Methods of storing such wastes for
hundreds to thousands of years have been proposed but none have been
universally accepted or praised.
    This is not a closed issue.  Intelligent technically astute citizens
may be found who strongly support differing positions on commercial
nuclear power and the waste problem.
-------

∂13-Jul-89  1539	Mailer 	re: Nuclear Power
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  15:39:42 PDT
Date: Thu, 13 Jul 89 15:40:04 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Nuclear Power
To: JMC@SAIL.Stanford.EDU
cc: DON@RML2.SRI.COM, comments@KL.SRI.COM, su-etc@SAIL.Stanford.EDU
In-Reply-To: <nsrpG@SAIL.Stanford.EDU>
Message-ID: <12509759274.21.STEINBERGER@KL.SRI.COM>

> Alfven, note spelling, is not a nuclear physicist but an
> astrophysicist. 

 Gee, astrophysicists don't know anything at all about fission, do they?  %-)

This kind of argument is an attempt to claim that only experienced 
professionals who have spent years researching or working in the nuclear
industry are knowledgeable and objective enough to ascertain its relative
safety.  Similar arguments have been made regarding who should decide
the US defense policy: only long-time DoD employees!

-ric s.

-------

∂13-Jul-89  1542	STEINBERGER@KL.SRI.COM 	re: Nuclear energy        
Received: from KL.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Jul 89  15:42:25 PDT
Date: Thu, 13 Jul 89 15:42:57 PDT
From: Richard Steinberger <STEINBERGER@KL.SRI.COM>
Subject: re: Nuclear energy    
To: JMC@SAIL.Stanford.EDU
In-Reply-To: <1Rss#X@SAIL.Stanford.EDU>
Message-ID: <12509759796.21.STEINBERGER@KL.SRI.COM>

Sigh. . . we just had our house painted yesterday and my references and books 
are all in boxes.  I will try to find something in the next few days.
I seem to remember another study out of U of Pitt. and a study by John
Goffman of LLL on the 3MI accident.

-ric s.

-------

∂13-Jul-89  1553	PKR 	re: meeting on equality circumscription 
To:   JMC
CC:   PKR   
[In reply to message rcvd 13-Jul-89 14:53-PT.]

10am tomorrow is fine, see you then.		-Peter

∂13-Jul-89  1900	JMC  
gabriel 854-3106

∂14-Jul-89  0030	karish@forel.Stanford.EDU 	Re: Nuclear energy
Received: from forel.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Jul 89  00:30:14 PDT
Received: by forel.Stanford.EDU (5.51/inc-1.0)
	id AA12721; Fri, 14 Jul 89 00:31:28 PDT
Date: Fri, 14 Jul 89 00:31:28 PDT
From: karish@forel.Stanford.EDU (Chuck Karish)
Message-Id: <8907140731.AA12721@forel.Stanford.EDU>
To: JMC@SAIL.STANFORD.EDU
Subject: Re: Nuclear energy
Newsgroups: su.etc
In-Reply-To: <1tssb0@SAIL.Stanford.EDU>
Organization: When necessary
Cc: 

In article <1tssb0@SAIL.Stanford.EDU> you wrote:
>[In reply to message from gscott@Portia.Stanford.EDU sent 13 Jul 89 20:55:07 GMT.]
>
>The usual phrase is "accident in a commercial nuclear reactor".

That wasn't the phrase used in the article that started the thread to
which you responded.  As it was phrased, the original assertion was
incorrect.

∂14-Jul-89  0315	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	535 million digits of pi?    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 14 Jul 89  03:15:48 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 400893; Fri 14-Jul-89 06:13:43 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 101961; Fri 14-Jul-89 02:02:04 PDT
Date: Fri, 14 Jul 89 02:02 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: 535 million digits of pi?
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890714090209.8.RWG@TSUNAMI.SPA.Symbolics.COM>

Today I got a weird call from the Chudnovskys.  They say Bailey
(29M, Cray 2) told them that Kanada (200M, NEC) told Bailey that
he had 525 million:  80 hrs using Salamin's, 85 hrs to check using
Kanada's improved quartic Borwein's, on a Hitachi monster with
5(?) gigabytes.  (Chudnovskys were/are multiplying tape-to-tape!)

There are several oddities here.  Kanada normally sends
congratulations when he is bested, but Chudnovskys have heard
zilch.  When Kanada had 200M, he told Yomiuri Shimbun that he
would not try again for less than 1G.  But the strangest thing
of all was David's reaction to Bailey:  something like,
"Well, remember, our computation was resumable.  Perhaps 535
is not the record after all."

(DHB:  is this what you heard?)

This could mean a number of things.  Perhaps they have been
continuing the computation since their announcement.  But what gets
resumed is not that actual stream of digits, but rather just the
series summation, which is a rational number, off by a factor of
sqrt(10005).  So, if they really have more digits than they
announced, they had to do at least another giant multiply and divide.

But they also have checked the computation by repeating it on
different machines, and the 479M they announced may have been the min
of two efforts.  (They have fancy congruence stuff to check the actual
series, so the only thing that needs repeating is the (/ num den
(sqrt 10005)).

Thus Kanada has been given a moving, and partially submerged target.
How many digits do the Chudnovskys have?  How fast is the SR-71?
Nipponese digit war goes classified??

If Kanada switches to a resumable method, then we might see a real,
ongoing race, with digits piling up like MacDonald's hamburgers.
Claims could be staked by publishing 100 digit bursts every 10M, or so.

But what the hell good are digits?  I'd almost rather have the burgers.

∂14-Jul-89  0700	JMC  
Don DeHaven 202-357-7494

∂14-Jul-89  0700	JMC  
pkr at 10am

∂14-Jul-89  0945	MPS  
Vishal Sikka - 408 247-5653 would like to set up an appointment
with you for next week.  Reference Doug Leonard and his AM program

∂14-Jul-89  1430	cross@vax.darpa.mil 	DARPA Program Summaries 
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 14 Jul 89  14:29:46 PDT
Received: from sun35.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA20159; Fri, 14 Jul 89 17:20:01 EDT
Posted-Date: Fri 14 Jul 89 17:28:41-EDT
Received: by sun35.darpa.mil (4.1/5.51)
	id AA03528; Fri, 14 Jul 89 17:28:44 EDT
Date: Fri 14 Jul 89 17:28:41-EDT
From: Steve Cross <CROSS@vax.darpa.mil>
Subject: DARPA Program Summaries
To: cross-pi-addresses@vax.darpa.mil
Message-Id: <616454921.0.CROSS@SUN35.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@SUN35.DARPA.MIL>

=====================================================================

     TO:  Principal Investigators Whom I Support

*********************************************************************
*********************************************************************

                This is an important message.  
          Please acknowledge receipt immediately to:  

                     cross@vax.darpa.mil 
                            AND
                     sully@vax.darpa.mil

*********************************************************************
*********************************************************************

Greetings! I've finally arrived at DARPA and I will be taking over the
management of the KBS, Planning, and Math Modeling work.  The
following request is a necessary "evil" this time of year.
The current budget and leadership uncertainties are additional complicating 
factors. I am a WILLING advocate for the AI community within this agency, 
but to be an EFFECTIVE advocate I need your utmost cooperation. Any additional 
information you can provide me to help me get smart on your project (tech reports 
or copies of recent papers) would also be appreciated. 

                                              Thanks in advance, 
                                              Stephen (Steve) E. Cross
                                              Program Manager, Machine Intelligence

*********************************************************************

In the next two weeks, I need two sets of data about each of your
current or anticipated contracts:
     a.  A 1-2 page program summary -- to educate the new Director of
         ISTO (not yet named) among others.
     b.  Certain administrative data -- to help me allocate FY-90 funds
         wisely.
Instructions and templates are given below.

The due date for submissions is Friday, 28 July.  If some data are not
available by that date, please send what you have and indicate when
the rest will follow.

Please send your submission(s) by e-mail to both of the following:
         wayne@vax.darpa.mil
         sully@vax.darpa.mil            

Please direct any questions to Mary Sullivan at 703/276-3532.

To help us sort submissions, please use one of the following
capitalized titles as the subject line of your message:
     SUMMARY IC4        for Knowledge-Based Systems
     SUMMARY IC4        for Math Modeling
     SUMMARY IC5        for Planning

In the near future you may receive an ISTO-wide request for financial
data in a format suitable for updating a database.  The fiscal data
that you will be collecting for me may help you answer that request
also.

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

A separate program summary with associated administrative data is
needed for each current or anticipated contract.  The instructions
below are designed to make your submissions consistent and effective.

Your program summary will be an important vehicle for explaining and
defending your program.  It should be compact and technically
interesting, drafted so as to be understood by the new Director of
ISTO or by other DARPA program managers.  It is worth writing well.

A well-written portfolio of program summaries will help us defend
current budgets in a time of contraction and possibly argue for
increased funding.

Your summary should be strong and positive, but without hyperbole.
Give a clear, top-level view.  Leave out minutiae.  Avoid jargon.
Write in the third person.

To help us quickly put many such summaries into Interleaf format,
please be sure to:
   a.  Include the capitalized titles used below, substituting your
       words for the lower case instructions.
   b.  Use plain ASCII with no markup commands from a text processor.
   c.  Do not use leading indentation or any extraneous embedded
       whitespace (but do leave blank lines between paragraphs).
   d.  Keep lines to a maximum of 70 characters.
   e.  Limit each program summary to a maximum of 7000 characters.
       (This will let us reformat your text to fit within 2 pages.
       Summaries for small efforts should be even shorter.  These
       limits do not apply to the separate administrative data.)
   
==========================================================================

                          === PROGRAM SUMMARY ===

ORGANIZATION

University or company name

PRINCIPAL INVESTIGATOR

Name, Phone, email address, US Mail Address

TITLE OF EFFORT

Use the title of the contract and the ARPA order number.

OBJECTIVE

A concise statement of what you are attempting to accomplish and why.
At most a few sentences.

APPROACH

A high-level description of your approach, both technical and
procedural.  Give enough context to make sense, but keep it brief, as
the emphasis should be on what you are doing.  Indicate what is
innovative and why it is promising.  Break into paragraphs if
appropriate.

PROGRESS

A brief discussion of how far you have come and where you are headed
in the total contract.  Include quantitative results (here and below)
if appropriate.  Perhaps a few sentences overall.

RECENT ACCOMPLISHMENTS

For existing contracts only: A crisp list of your most significant
accomplishments in the last 12-18 months; one sentence per item.
The list may include important new ideas.

FY-90 PLANS

A crisp list of the most significant new items that you expect to
accomplish in FY-90; one sentence per item.

Note:  Under both RECENT ACCOMPLISHMENTS and FY-90 PLANS, please
       phrase your sentences like the two examples below and insert a
       blank line between each sentence:

Developed a message understanding system (PUNDIT) to extract key 
data (e.g., who did what to whom) from telegraphic military messages.

Ported PUNDIT to several Navy domains (CASPREPS, RAINFORMS, OPREPS),
to maintenance reports, to natural language database queries, to
medical abstracts, and to air traffic control transmissions.

==========================================================================

                          === ADMINISTRATIVE DATA ===

 
  a.  ARPA ORDER NUMBER:  This is a 4-digit number

  b.  CONTRACT NUMBER:

  c.  AGENT:  The agency that administers your contract (e.g. ONR).

  d.  CONTRACT TITLE:  If you proposed a new title in your program
                       summary, put it here and also include the 
                       official title.

  e.  ORGANIZATION:    University or company name

  f.  PI:  

  g.  ACTUAL OR EXPECTED START DATE:

  h.  EXPECTED END DATE IF OPTIONS NOT EXERCISED:

      The "options" referred to here and below are official options
      for additional work negotiated at the time of award, but not yet
      exercised by the government. Options which have already been
      exercised are considered part of the contract.  (Many contracts
      do not have options.)

  i.  EXPECTED END DATE IF OPTIONS EXERCISED:

  j.  TOTAL PRICE IF OPTIONS NOT EXERCISED:

  k.  SPENDING AUTHORITY PROVIDED SO FAR:

  l.  DATE WHEN THESE FUNDS WILL BE FULLY EXPENDED:

  m.  ADDITIONAL FUNDS EXPECTED PER CONTRACT (by FY):

  n.  MINIMUM ACCEPTABLE ADDITIONAL FUNDS (by FY):

  o.  ADDITIONAL FUNDS EXPECTED (by FY) IF OPTIONS EXERCISED:

  p.  ANYTHING ELSE YOU NEED (date and item):

  q.  OTHER REMARKS:

==========================================================================



-------

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

-------

∂15-Jul-89  1311	ME 	re: discrepancy      
 ∂15-Jul-89  1236	JMC 	discrepancy    
I note that WHO says the system has been up for 3 days, while there
are five sails. Curious!

ME - That was from a hack I put in yesterday for Bastille Day, to
show lots of fireworks.  Check it now.  It should be back to normal.

∂15-Jul-89  1439	wab@sumex-aim.stanford.edu 	re: Nuke Japan Now!   
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 15 Jul 89  14:39:15 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA11389; Sat, 15 Jul 89 14:41:21 PDT
Date: Sat, 15 Jul 1989 14:41:21 PDT
From: "William A. Brown" <wab@sumex-aim.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Cc: wab@sumex-aim.stanford.edu, su-etc@sumex-aim.stanford.edu
Subject: re: Nuke Japan Now! 
In-Reply-To: Your message of 14 Jul 89 2116 PDT 
Message-Id: <CMM.0.88.616542081.wab@sumex-aim.stanford.edu>

For once, JMC and I agree on something. It's true that regarding trade issue
the Japanese don't fight fairly, but then again, neither does the United
States. Still, the anti-Japanese sentiment rising in this country hits a
little too close to home for me.

--Bill

∂15-Jul-89  2156	cphoenix@csli.Stanford.EDU 	re: greenhouse effect 
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 15 Jul 89  21:56:53 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA02448; Sat, 15 Jul 89 21:57:47 PDT
Date: Sat, 15 Jul 89 21:57:47 PDT
From: cphoenix@csli.Stanford.EDU (Chris Phoenix)
Message-Id: <8907160457.AA02448@csli.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: re: greenhouse effect

s'possible.  I just know what I was told.  But I also understand that coal 
takes just the right conditions to form, and I do know that vegetable matter
in swamps must decay sometime, or else they'd fill up too quickly.

∂16-Jul-89  0004	rdz@Gang-of-Four.Stanford.EDU 	rn instructions    
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 16 Jul 89  00:04:42 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA09431; Sun, 16 Jul 89 00:02:37 PDT
Date: Sun, 16 Jul 89 00:02:37 PDT
From: Ramin Zabih <rdz@Gang-of-Four.Stanford.EDU>
Message-Id: <8907160702.AA09431@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: rn instructions

When you log in to go4, type "rn soc.culture.china".  If it asks you
any questions about adding other news groups, just type carriage return
as an answer.  Then it will tell you how many unread articles there
are -- currently, about 970.  Answer "y" when it asks you if you want
to read this group.  This will put you into the main loop.

At any point, typing "h" will get you help.  Within an article, typing
space will scroll forward.  "n" will get you to the chronologically
next article, while control-n will get you to the next article with
the same title.  Also, "s foo" will save the current article in the
file ~jmc/News/foo.

If you have any trouble, feel free to call me at home.


			 Ramin

∂16-Jul-89  1228	RDZ@Score.Stanford.EDU   
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 16 Jul 89  12:28:23 PDT
Date: Sun, 16 Jul 1989  12:28 PDT
Message-ID: <RDZ.12510510781.BABYL@SCORE.STANFORD.EDU>
From: RDZ@Score.Stanford.EDU
To:   jmc@Sail.Stanford.EDU

You can always FTP the file to Sail and print it from there.  The file(s)
should probably be in /user/jmc/News.

			Ramin

∂17-Jul-89  0700	JMC  
searle

∂17-Jul-89  0900	JMC  
Alcon + bids

∂17-Jul-89  0932	kolk@smiley.Stanford.EDU 	The articles form soc.culture.china    
Received: from smiley.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Jul 89  09:31:58 PDT
Received: by smiley.Stanford.EDU (4.0/inc-1.0)
	id AA04314; Mon, 17 Jul 89 09:29:18 PDT
Date: Mon, 17 Jul 89 09:29:18 PDT
From: kolk@smiley.Stanford.EDU (Dan Kolkowitz)
Message-Id: <8907171629.AA04314@smiley.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: The articles form soc.culture.china

Should be in my directory on labrea in a file called china.tar in the week
of June 15-22.  The best shot would probably be about June 20.  I'll send
a message to farhad saying that you can take it from my directory.

Dan
(china.tar is a tar file which will unpack with the command tar xf china.tar.
It will create subdirectories holding the articles.  The subdirectories
are arranged by date received.)

∂17-Jul-89  0937	kolk@labrea.stanford.edu 	The file is still around on labrea afterall 
Received: from labrea.stanford.edu by SAIL.Stanford.EDU with TCP; 17 Jul 89  09:36:55 PDT
Received: by labrea.stanford.edu; Mon, 17 Jul 89 09:36:41 PDT
Date: Mon, 17 Jul 89 09:36:41 PDT
Message-Id: <8907171636.AA00962@labrea.stanford.edu>
From: Dan Kolkowitz <kolk@labrea.stanford.edu>
Subject: The file is still around on labrea afterall
To: jmc@sail.stanford.edu

It is /tmp/weening/china.tar.  So, there is no need to get it off backups.
I would take it off labrea as quick a possible.
Dan

∂17-Jul-89  1336	wab@sumex-aim.stanford.edu 	re: Racism  
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 17 Jul 89  13:36:04 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA11920; Mon, 17 Jul 89 13:38:09 PDT
Date: Mon, 17 Jul 1989 13:38:08 PDT
From: "William A. Brown" <wab@sumex-aim.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Cc: wab@sumex-aim.stanford.edu, su-etc@sumex-aim.stanford.edu
Subject: re: Racism 
In-Reply-To: Your message of 15 Jul 89 2142 PDT 
Message-Id: <CMM.0.88.616711088.wab@sumex-aim.stanford.edu>

Sorry, JMC.

If all of them received jobs "in accordance with their qualifications" I
would be a very happy medical student. This has never been the case. I have
plenty of doctors in my family who can attest to that. Sometimes I am amazed
at the racial ingnorance in this country. That is not a flame, nor was the
statement meant for non-minorities only. That's like saying all Blacks
receive loan acceptances in accordance with their economic status - and we
all know THAT myth was blown way out of the water this year. You know, I used
to be angry at some of the opinion in this country - now I am just amazed at
the lack of information. I suppose you also think all medical students
receive privilages in accordance with their academic performance? I have a
Black friend right now who is ardently fighting the practice in her lab (AT
STANFORD IN THE BECKMAN CENTER) of the other lab assistants (all non-Black)
holding regular unscheduled meetings without her. I had the same thing happen
to me in the lab of Dr. Ellen Berkowitz of Boston University School of
Medicine, last year. I hav friends around the nation in medical schools
putting up with the same crap. Your naiivity (sp?) is almost touching - I am
being neither sarcastic nor patronizing. Please ASK someone, preferrably
several people, before making such statements. In fact, I would suggest
asking someone FYI. It will surely be a revealing conversation.


--Bill

∂17-Jul-89  1343	MPS  
Hi

Susan called.  She can be reached at 923-5656.

Pat

∂17-Jul-89  1406	eps@Princeton.EDU 	visit 
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 17 Jul 89  14:06:17 PDT
Received: from sycamore.Princeton.EDU by Princeton.EDU (5.58+++/2.20)
	id AA07698; Mon, 17 Jul 89 17:06:46 EDT
Received: by sycamore.Princeton.EDU (4.0/1.83)
	id AA03737; Mon, 17 Jul 89 17:06:40 EDT
Date: Mon, 17 Jul 89 17:06:40 EDT
From: eps@Princeton.EDU (Elisha Sacks)
Message-Id: <8907172106.AA03737@sycamore.Princeton.EDU>
To: jmc@sail.stanford.edu
Subject: visit

Hi there.  My name is Elisha Sacks from Princeton.  We spoke
briefly in Israel about dynamics.  You said that you would like
to discuss my work in more detail, but it didn't work out there.
I will be in Palo Alto on Aug. 8-11 for the third qualitative
physics workshop.  Would you like to get together and talk on one
of those days?  I'm free on the afternoon of the 8th and have
some free hours on the other days?

thanks,
elisha.

∂17-Jul-89  1414	dai@csli.Stanford.EDU 	Re: [none]  
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Jul 89  14:14:38 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA25600; Mon, 17 Jul 89 14:15:32 PDT
Date: Mon, 17 Jul 89 14:15:32 PDT
From: dai@csli.Stanford.EDU (Jim Dai)
Message-Id: <8907172115.AA25600@csli.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 17 Jul 89  1410 PDT <purMC@SAIL.Stanford.EDU>
Subject: Re: [none]


Dear Professor McCarthy,

  I don't have a work  phone. If you need to reach me, may I get your
number and call you back?

Jim Dai

∂17-Jul-89  1420	dai@csli.Stanford.EDU 	Re: [none]  
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Jul 89  14:17:46 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA25746; Mon, 17 Jul 89 14:18:40 PDT
Date: Mon, 17 Jul 89 14:18:40 PDT
From: dai@csli.Stanford.EDU (Jim Dai)
Message-Id: <8907172118.AA25746@csli.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 17 Jul 89  1415 PDT <16urRK@SAIL.Stanford.EDU>
Subject: Re: [none]


I will try to find a phone to call you either at work or at home.

Jim Dai

∂18-Jul-89  0820	perrault@ai.sri.com 	meeting  
Received: from Warbucks.AI.SRI.COM ([128.18.3.1]) by SAIL.Stanford.EDU with TCP; 18 Jul 89  08:20:44 PDT
Received: from agate.ai.sri.com by Warbucks.AI.SRI.COM with INTERNET ;
          Tue, 18 Jul 89 08:21:06 PDT
Received: by agate.ai.sri.com (4.1/4.16)
	id AA13517 for jmc@sail.stanford.edu; Tue, 18 Jul 89 08:22:07 PDT
Date: Tue 18 Jul 89 08:22:03-PDT
From:     PERRAULT@Warbucks.AI.SRI.COM (Ray Perrault)
Subject: meeting
To:       jmc@sail.stanford.edu
cc:       perrault@Warbucks.AI.SRI.COM
Message-Id: <616778523.0.PERRAULT@ai.sri.com>
Mail-System-Version: <SUN-MM(229)+TOPSLIB(128)@ai.sri.com>

wed at 4 is fine.  My office at Cordura?

ray
-------

∂18-Jul-89  1012	MPS  

They will not take American Express for the conference in
Detroit (AAAI).  Will you mail me your Visa No., or is it
on file here and if so where?  Thanks

∂18-Jul-89  1102	davidson@psych.Stanford.EDU 	Re: "What can the white man say to the black woman?"    
Received: from psych.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Jul 89  11:02:35 PDT
Received: by psych.Stanford.EDU (3.2/4.7); Tue, 18 Jul 89 11:00:15 PDT
Date: Tue, 18 Jul 89 11:00:15 PDT
From: davidson@psych.Stanford.EDU (Martin Davidson)
To: JMC@SAIL.STANFORD.EDU, su-etc@psych.Stanford.EDU
Subject: Re: "What can the white man say to the black woman?"
Newsgroups: su.etc
In-Reply-To: <RuCJN@SAIL.Stanford.EDU>
Organization: Stanford University
Cc: 

In article <RuCJN@SAIL.Stanford.EDU> you write:
>[In reply to message from ramsey@Polya.Stanford.EDU sent 18 Jul 89 04:41:42 GMT.]
>
>What are the life expectancies at birth today of white men
>and black women?

I am astounded by this response.  First, what is your point?  Second,
what does that point have to do with Walker's message?

I react strongly because Walker blended nicely, (I thought)
historical fact, and personal emotion.  It reflects the feelings and
attitudes of many black women in this country and has implications
for the attitudes of other historically oppressed people, yet
the first response I see on this board is your vague and
seemingly deflective question!  I believe there is a message for anyone who
reads that piece about the emotional impact of racism (defined
precisely as the systematic use of power to perpetuate prejudiced
attitudes) but it requires more thoughful reflection than is apparent
in your response.
-MD

∂18-Jul-89  1203	wab@sumex-aim.stanford.edu 	Re: black doctors     
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 18 Jul 89  12:03:20 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA05890; Tue, 18 Jul 89 12:05:25 PDT
Date: Tue, 18 Jul 1989 12:05:21 PDT
From: "William A. Brown" <wab@sumex-aim.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Cc: wab@sumex-aim.stanford.edu, su-etc@sumex-aim.stanford.edu
Subject: Re: black doctors 
In-Reply-To: Your message of 18 Jul 89 1142 PDT 
Message-Id: <CMM.0.88.616791921.wab@sumex-aim.stanford.edu>

JMC,
   I don't know the particular problem with the Palo Alto clinic. I can tell
you from experience, however, that your last statement "like almost all other
institutions, they have ... their recruiting ads ... in publications that
black doctors and medical students are claimed to read." is completely false.
Most non-minority organizations do NOT recruit at all in minority
professional publications. The noted exception is the federal government, but
even it does not recruit for many positions in minority publications. Heck,
htis medical school only gives East Coast interviews in the Boston area - as
if those are the only schools that exist. This medical school certainly
doesn't recruit at Black universities. Most don't. The same is true of other
graduate and professional schools. If you were really looking for some Black
computer scientists to even out the lack of minority computer scientists
here, I suggest you contact the computer science departments of Howard
university and North Carolina A&T. I'm sure they could put you in contact
with some outstanding scientists. In fact, I knew the system manager at
Howard when he was getting undergraduate degrees in mechanical engineering,
electrical engineering, civil engineering, and fine arts. Talk about an
OVER-qualified individual...

--Bill

∂18-Jul-89  1205	@Score.Stanford.EDU:clancey.pa@Xerox.COM 	Einstein remark   
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Jul 89  12:05:22 PDT
Received: from Xerox.COM by SCORE.STANFORD.EDU with TCP; Tue 18 Jul 89 12:01:59-PDT
Received: from Cabernet.ms by ArpaGateway.ms ; 18 JUL 89 11:35:45 PDT
Date: 18 Jul 89 11:35 PDT
From: Bill Clancey <clancey.pa@Xerox.COM>
Subject: Einstein remark
To: mccarthy@score.stanford.edu
Message-ID: <890718-113545-9682@Xerox>

John,

	I'd like to cite your remark that the field needs an Einstein, etc.  I
believe I saw this in print in an AI Magazine article, but can't find it.
Could you please give me a pointer?
Thanks.

	Bill

∂18-Jul-89  1216	harnad@clarity.Princeton.EDU 	Preprint  
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 18 Jul 89  12:16:19 PDT
Received: from clarity.Princeton.EDU by Princeton.EDU (5.58+++/2.20)
	id AA13899; Tue, 18 Jul 89 14:42:15 EDT
Received: by clarity.Princeton.EDU (4.0/1.83)
	id AA29131; Tue, 18 Jul 89 14:28:03 EDT
Date: Tue, 18 Jul 89 14:28:03 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8907181828.AA29131@clarity.Princeton.EDU>
To: srj@flash.bellcore.com
Subject: Preprint

Here is a copy of a paper on the symbol grounding problem
(in troff -ms format). I'd be happy to hear any reactions.

Stevan Harnad

--------------------------------------------------------------------
.LP
.nh
              THE SYMBOL GROUNDING PROBLEM

      Presented at CNLS Conference on Emergent Computation,
   Session on "The Emergence of Symbols" Los Alamos, May 1989
                Submitted to Physica D.

                Stevan Harnad
                Department of Psychology 
                Princeton University
                Princeton NJ 08544

ABSTRACT: There has been much discussion recently about the scope and
limits of purely symbolic models of the mind and about the proper role
of connectionism in mental modeling. This paper describes the "symbol
grounding problem": In a purely symbolic system, how can the meanings
of meaningless symbol tokens, manipulated solely on the basis of their
shapes, be grounded in anything but other meaningless symbols? A
candidate solution is sketched: Symbolic representations must be
grounded bottom-up in nonsymbolic representations of two kinds: (1)
.I "iconic representations" ,
which are analogs of the sensory projections of objects and events, and
(2)
.I "categorical representations" ,
which are learned and innate feature-detectors that pick out the
invariant features of object and event categories from their sensory
projections. Elementary symbols are the names of these object and event
categories, assigned on the basis of their (nonsymbolic) categorical
representations. Higher-order (3)
.I "symbolic representations" ,
grounded in these elementary symbols, consist of symbol strings
describing category membership relations. Connectionism is one natural
candidate for the mechanism that learns the invariant features
underlying categorical representations. In this way connectionism can
be seen as a complementary component in a hybrid nonsymbolic/symbolic
model of the mind, rather than a rival to purely symbolic modeling.
Such a hybrid model would not have an autonomous symbolic "module,"
however; the symbolic functions would emerge as an intrinsically
"dedicated" symbol system as a consequence of the bottom-up grounding
of categories' names in their sensory representations -- iconic and
categorical.


.na
.ce1
.B
1. Modeling the Mind
.R

.B "1.1 From Behaviorism to Cognitivism."
For many years the only empirical approach in psychology was
behaviorism, its only explanatory tools input/input and
input/output associations (in the case of classical conditioning;
Turkkan 1989) and the reward/punishment history that "shaped" behavior
(in the case of operant conditioning; Catania & Harnad 1988). In a
reaction against the subjectivity of armchair introspectionism,
behaviorism had declared it to be just as illicit to theorize about what went
on in the
.I head
of the organism to generate its behavior as to theorize
about what went on in its
.I mind.
Only 
.I observables
were to be the subject
matter of psychology, and, apparently, these were expected to explain
themselves.

Psychology became more like an empirical science when, with the gradual
advent of cognitivism (Miller 1956, Neisser 1967, Haugeland 1978), it
became acceptable to make inferences about the
.I unobservable processes
underlying behavior.
Unfortunately, cognitivism let mentalism in again
by the back door too, and the hypothetical internal processes came
embellished with subjective interpretations. In fact, semantic interpretability
(meaningfulness),
as we shall see, was one of the defining features of the most prominent
contender for the theoretical vocabulary of cognitivism, the "language
of thought" (Fodor 1975), which became the dominant paradigm for cognitive
theory for several decades in the form of the "symbolic" model of the
mind: The mind is a symbol system and cognition is symbol
manipulation. The possibility of generating complex behavior through
symbol manipulation was empirically demonstrated by successes in the
field of artificial intelligence (AI).

.B "1.2 Symbol Systems."
What is a symbol system? From Newell (1980) Pylyshyn (1984),
Fodor (1987) and the classical work of Von Neumann, Turing, Goedel,
Church, etc.(see Kleene 1969) on the foundations of computation, we can
reconstruct the following definition:

A symbol system is:

(1) a set of arbitrary
.I "physical tokens"
(scratches on paper, holes on a tape, events in a digital computer,
etc.) that are
.br
(2) manipulated on the basis of
.I "explicit rules"
that are
.br
(3) likewise physical tokens and
.I strings
of tokens. The rule-governed symbol-token manipulation is based
.br
(4) purely on the
.I shape
of the symbol tokens (not their "meaning"), i.e., it is purely
.I syntactic,
and consists of
.br 
(5)
.I "rulefully combining"
and recombining symbol tokens. There are
.br 
(6) primitive
.I atomic
symbol tokens and
.br 
(7)
.I composite
symbol-token strings. The entire system and all its parts -- the
atomic tokens, the composite tokens, the
syntactic manipulations (both actual and possible) and the rules -- are
all
.br 
(8)
.I "semantically interpretable:"
The syntax can be 
.I systematically
assigned a
meaning (e.g., as standing for objects, as describing states of
affairs).

According to proponents of the symbolic model of mind such as Fodor
(1980) and Pylyshyn (1980, 1984), symbol-strings of this sort capture
what mental phenomena such as thoughts and beliefs are. Symbolists
emphasize that the symbolic level (for them, the mental level) is a
natural functional level of its own, with ruleful regularities that are
independent of their specific physical realizations. For symbolists,
this implementation-independence is the critical difference between
cognitive phenomena and ordinary physical phenomena and their
respective explanations. This concept of an autonomous symbolic level also
conforms to general foundational principles in the theory of
computation and applies to all the work being done in symbolic AI, the
branch of science that has so far been the most successful in
generating (hence explaining) intelligent behavior.

All eight of the properties listed above seem to be critical to this
definition of symbolic. Many phenomena have some of the properties, but
that does not entail that they are symbolic in this explicit,
technical sense. It is not enough, for example, for a phenomenon to be
.I interpretable
as rule-governed, for just about anything can be interpreted as
rule-governed. A thermostat may be interpreted as following the
rule: Turn on the furnace if the temperature goes below 70 degrees
and turn it off if it goes above 70 degrees, yet nowhere in the
thermostat is that rule explicitly represented.
Wittgenstein (1953) emphasized the difference between
.I explicit 
and
.I implicit
rules: It is not the same thing to "follow" a rule (explicitly) and
merely to behave "in accordance with" a rule (implicitly).\**
.FS

.nh
.na
.ls2
Similar
considerations apply to Chomsky's (1980) concept of "psychological
reality" (i. e., are Chomskian rules really physically represented in the
brain or do they merely "fit" our performance regularities, without
being what actually governs them?). Another version of the distinction
concerns explicitly represented rules versus hard-wired physical
constraints (Stabler 1985). In each case, an explicit, decomposable,
systematic representation
would be symbolic whereas an implicit physical constraint would not,
although
.I both
would be semantically "intepretable" as a "rule" if construed in
isolation rather than as part of a system.
.FE
The critical
difference is in the compositeness (7) and systematicity (8) criteria.
The explicitly represented symbolic rule is part of a formal system, it is
decomposable (unless primitive), its application and manipulation is
purely formal (syntactic, shape-dependent), and the entire system must be
semantically interpretable, not just the chunk in question. An isolated
("modular") chunk cannot be symbolic; being symbolic is a systematic property.

So the mere fact that a behavior is "interpretable" as
ruleful does not mean that it is really governed by a symbolic rule.\**
.FS
.ls2
.nh
.na
Analogously, the mere fact that a behavior is interpretable as purposeful
or conscious or meaningful does not mean that it is really any of those
things. (For arguments to the contrary, see Dennett 1983).
.FE
Semantic interpretability must be coupled with explicit representation
(2), syntactic manipulability (4), and systematicity (8) in order to be
symbolic. None of these criteria is arbitrary, and, as far as I can
tell, if you weaken them, you lose the grip on what looks like a
natural category and you sever the links with the formal theory of
computation, leaving a sense of "symbolic" that is merely unexplicated
metaphor (and probably differs from speaker to speaker). Hence it is
only this formal sense of "symbolic" and "symbol system" that will be
considered in this discussion of the emergence of symbol systems.

.ce1
.B "2. Dynamical Systems"

.B "2.1 Connectionism."
An early rival to the symbolic model of mind appeared (Rosenblatt 1962),
was overcome by symbolic AI (Minsky & Papert 1969) and has recently
re-appeared in a stronger form that is currently vying with AI to be
the general
theory of cognition and behavior (McClelland, Rumelhart et al. 1986).
Variously described as "dynamical systems," "neural networks,"
"parallel distributed processing" and "connectionism," this approach
has a multiple agenda, which includes providing a theory of brain
function. Much can be said for and against studying behavioral and
brain function independently, but in this paper it will be assumed
that, first and foremost, a cognitive theory must stand on its own
merits, which depend on how well it explains our observable behavioral
capacity; whether or not it does so in a sufficiently brainlike way is
another matter, and a downstream one, in the course of theory development.
Very little is known of brain structure and vegetative function so far;
and the nature of higher brain function is itself a theoretical matter. To
"constrain" a cognitive theory to account for behavior in a brainlike
way is hence premature in two respects: (1) It is far from clear yet what
"brainlike" means, and (2) we are far from having accounted for a
lifesize chunk of behavior yet, even without added constraints.
Moreover, the formal principles underlying connectionism are those of
dynamical systems governed by causal interactions and their
statistical and associative structure; a neural network is merely one
possible implementation of a dynamical system.

Connectionism will accordingly only be considered here as a cognitive
theory. As such, it has lately challenged the symbolic approach
to modeling the mind. According to connectionism, cognition is not
symbol manipulation but dynamic patterns of activity in a multilayered
network of nodes or units with weighted positive and negative
interconnections. The patterns change according to internal network
constraints governing how the activations and connection strengths are
adjusted on the basis of new inputs (e.g., the "delta rule" and
"backpropogation,"McClelland, Rumelhart et al. 1986). The result is a system
that learns, recognizes patterns, solves problems, and can even exhibit
motor skills.

.B "2.2 Scope and Limits."
It is far from clear what the scope and limits of either symbolic AI
or connectionism are. The former seems better at formal and
language-like tasks, the latter at sensory, motor and learning tasks,
but there is considerable overlap and neither has gone much beyond the
stage of "toy" tasks toward lifesize behavioral capacity. Moreover,
there has been some disagreement as to whether or not connectionism
itself is symbolic. We will adopt the position here that it is not,
because connectionist networks fail to meet several of the criteria for
being symbol systems, as Fodor & Pylyshyn (1988) have argued recently.
In particular, although, like everything else, their behavior and
internal states can be given isolated semantic interpretations, nets
fail to meet the compositeness (7) and systematicity (8) criteria
listed earlier:  The patterns of interconnections do not decompose,
combine and recombine according to a formal syntax that can be given a
systematic semantic interpretation.\**
.FS
.ls2
.nh
.na
There is some misunderstanding of this point because it is often
conflated with a mere implementational issue: Connectionist
networks can be simulated using symbol systems, and symbol systems can
be implemented using a connectionist architecture, but that is
independent of the question of what each can do
.I qua
symbol system or connectionist network, respectively. By way of
analogy, silicon can be used to build a computer, and a computer can
simulate the properties of silicon, but the functional
properties of silicon are not those of computation, and the functional
properties of computation are not those of silicon.
.FE
Instead, nets seem to do what they do
.I non symbolically.
According to
Fodor & Pylyshyn, this is a severe limitation, because many of our
behavioral capacities appear to be symbolic, and hence the most natural
hypothesis about the underlying cognitive processes that generate them
would be that they too must be symbolic. Our linguistic capacities are
the primary examples here, but many of the other skills we have --
logical reasoning, mathematics, chess-playing, perhaps even our
higher-level perceptual and motor skills -- also seem to be symbolic.
In any case, when we interpret our sentences, mathematical formulas,
and chess moves (and perhaps some of our perceptual judgments and motor
strategies) as having a systematic
.I meaning
or
.I content,
we know at first hand that that's literally true, and not just a figure
of speech. Connectionism hence seems to be at a disadvantage in
attempting to model these cognitive capacities.

Yet it is not clear whether connectionism should for this reason aspire to
be symbolic, for the symbolic approach turns out to suffer from a
severe handicap, one that may be responsible for the limited nature of
its success to date (especially in modeling human-scale capacities) as
well as the uninteresting and ad hoc nature of the symbolic "knowledge"
it attributes to the "mind" of the symbol system. The handicap has been
noticed in various forms since the advent of computing; I have
dubbed a recent manifestation of it the "symbol grounding problem"
(Harnad 1987b).

.ce1
.B "3. The Symbol Grounding Problem"

.B "3.1 The Chinese Room."
Before defining the symbol grounding problem I will give two examples
The first comes from Searle's (1980) celebrated "Chinese Room Argument," in
which the symbol grounding problem is referred to as the problem of
intrinsic meaning (or "intentionality"): Searle challenges the core
assumption of symbolic AI that a symbol system able to generate
behavior indistinguishable from that of a person must have a mind. More
specifically, according to the symbolic theory of mind, if a computer
could pass the Turing Test (Turing 1964)
in Chinese -- i.e., if it could respond to
all Chinese symbol strings it receives as input with Chinese symbol
strings that are indistinguishable from the replies that a real Chinese
speaker would make (even if we keep testing till doomsday) -- then the
computer would understand the meaning of Chinese symbols in the same
sense that I understand the meaning of English symbols.

Searle's simple demonstration that this cannot be so consists of
imagining himself doing everything the computer does -- receiving the
Chinese input symbols, manipulating them purely on the basis of their shape
(in accordance with (1) to (8) above), and finally returning the Chinese
output symbols. It is evident that Searle (who knows no Chinese) would
not be understanding Chinese under those conditions -- hence neither
could the computer. The symbols and the symbol manipulation, being all
based on shape rather than meaning, are systematically
.I interpretable
as having meaning -- that, after all, is what it is to be a symbol
system, according to our definition. But the interpretation will not be
.I intrinsic
to the symbol system itself: It will be parasitic on the fact that the
symbols have meaning for
.I us,
in exactly the same way that the meanings of the symbols in a book are
not intrinsic, but derive from the meanings in our heads. Hence, if the
meanings of symbols in a symbol system are extrinsic rather than
intrinsic like the meanings in our heads, then they are not a viable
model for the meanings in our heads: Cognition cannot be just symbol
manipulation.

.B "3.2 The Chinese/Chinese Dictionary-Go-Round."
My own example of the symbol grounding problem has two versions,
one difficult, and one, I think, impossible. The difficult version is: Suppose
you had to learn Chinese as a second language and the only source of
information you had was a Chinese/Chinese dictionary. The trip through
the dictionary would amount to a merry-go-round, passing endlessly from
one meaningless symbol or symbol-string (the definientes) to another (the
definienda), never coming to a halt on what anything meant.\**
.FS
.ls2
.nh
.na
Symbolic AI abounds with symptoms of the symbol grounding problem. One
well-known (though misdiagnosed) manifestation of it is Minsky's (1961)
"frame" problem: It is a frustrating but familiar experience in writing
"knowledge-based" programs that a system apparently behaving perfectly
intelligently for a while can be foiled by an unexpected case that
demonstrates its utter stupidity: A "scene-understanding" program will
blithely describe the goings-on in a visual scene and answer questions
demonstrating its comprehension (who did what, where, why) and then
suddenly reveal that it doesn't "know" that hanging up the phone and
leaving the room does not make the phone disappear, or something like
that. (It is important to note that these are not the kinds of lapses
and gaps in knowledge that people are prone to; rather, they are such
howlers as to cast serious doubt on whether the system has anything
like "knowledge" at all.)
The "frame" problem has been optimistically defined as the
problem of formally specifying ("framing") what varies and what stays
constant in a particular "knowledge domain," but in reality it's the
problem of second-guessing all the contingencies the programmer has not
anticipated in symbolizing the knowledge he is attempting to symbolize.
The contingencies are probably unbounded, for practical purposes,
because purely symbolic "knowledge" is ungrounded. Merely adding on more
symbolic contingencies is like taking a few more turns in the
Chinese/Chinese Dictionary-Go-Round. There is in reality no ground in
sight: just enough "intelligent" symbol-manipulation to persuade the
programmer that its meaningfulness surely couldn't just be parasitic on
the meanings he is projecting onto it from the grounded meanings in his
own head. Yet parasitism it is, as the next "frame problem" lurking
around the corner is ready to confirm. (A similar form of
over-interpretation has occurred in the ape "language" experiments
[Terrace 1979]. Perhaps both apes and computers should be trained using
Chinese code, to immunize their experimenters and programmers
against spurious over-interpretations. But
since the actual behavioral tasks in both domains are still so trivial,
there's probably no way to prevent their being decrypted. In fact, 
there seems to be an irresistible tendency to overinterpret toy task
performance itself, preemptively extrapolating and "scaling it up"
conceptually to lifesize without any justification in practice.)
.FE

The only reason cryptologists of ancient languages and secret codes
seem to be able to successfully accomplish something very like this is
that their efforts are
.I grounded
in a first language and in real world experience and knowledge.\**
.FS
.ls2
.nh
.na
Cryptologists also use statistical information about word frequencies,
inferences about what an ancient culture or an enemy government are
likely to be writing about, decryption algorithms, etc.
.FE
The second variant of the Dictionary-Go-Round, however, goes far beyond
the conceivable resources of cryptology: Suppose you had to learn
Chinese as a
.I first
language and the only source of information you had was a
Chinese/Chinese dictionary! This is more like the actual task faced by
a purely symbolic model of the mind: How can you ever get off the
symbol/symbol merry-go-round? How is symbol meaning to be grounded in
something other than just more meaningless symbols? This is the symbol
grounding problem.\**
.FS
.ls2
.nh
.na
Note that, strictly speaking, symbol grounding is a
problem only for cognitive modeling, not for AI in general. If
symbol systems alone
succeed in generating the intelligent machine
performance AI is interested in -- e.g., an automated dictionary --
then there is no reason whatsoever to demand that their symbols have
intrinsic meaning. On the other hand, the fact that our own symbols do
have intrinsic meaning whereas the computer's do not, and the fact
that we can do things that the computer so far cannot may be
indications that even in AI there may be performance gains to be made
(especially in robotics and machine vision) from endeavouring to ground
symbol systems.
.FE

.B "3.3 Connecting to the World."
The standard reply of the symbolist (e.g., Fodor 1980, 1985) is that the
meaning of the symbols comes from connecting the symbol system to the
world "in the right way." But it seems apparent that the problem of
connecting up with the world in the right way is virtually coextensive
with the problem of cognition itself. If each definiens in a
Chinese/Chinese dictionary were somehow connected to the world in the right
way, we'd hardly need the definienda! Many symbolists believe that
cognition, being symbol-manipulation, is an autonomous functional
module that need only be hooked up to peripheral devices in order to
"see" the world of objects to which its symbols refer (or, rather, to
which they can be systematically interpreted as referring). The
homuncular view underlying this position is quite apparent. Among other
things, it trivializes the difficulty of picking out the objects,
events and states of affairs in the world that symbols refer to, i.e.,
it trivializes the symbol grounding problem.

It is a candidate solution to this problem, confronted directly, that
will now be sketched. What is proposed is a dedicated hybrid
nonsymbolic/symbolic system in which the elementary symbols are
grounded in two kinds of nonsymbolic representations that pick out,
from their proximal sensory projections, the distal object categories
to which the elementary symbols refer. Most of the components of which
the model is made up (analog projections and transformations,
discretization, invariance detection, connectionism, symbol
manipulation) have also been proposed in various configurations by
others, but they will be put together in a specific bottom-up way here
that has not, to my knowledge, been previously suggested, and it is
this specific configuration on which the potential success of the
grounding scheme critically depends. Before proceeding, however, let us
take stock of the relative strengths and weaknesses of connectionism
and symbolism, the two current rival candidates for explaining
.I all
of cognition single-handedly.

.ce1
.B
4. Connectionism Vs. Symbol Systems

.B
4.1 Strengths of Connectionism:
.R

.I
(1) Nonsymbolic Function:
.R
As long as it does not aspire to be a symbol system, a connectionist
network has the advantage of not being subject to the symbol grounding
problem.

.I
(2) Generality:
.R
Connectionism applies the same small family of algorithms to many problems,
whereas symbolism, being a methodology rather than an algorithm, relies
on endless problem-specific symbolic rules.

.I
(3) "Neurosimilitude":
.R
Connectionist architecture seems more brain-like than a
Turing machine or a digital computer.

.I
(4) Pattern Learning:
.R
Connectionist networks are especially suited to the learning of
patterns from data.

.B
4.2 Weaknesses of Connectionism:

.R
.I
(1) Nonsymbolic Function:
.R
Connectionist networks, because they are not symbol systems, do not
have the systematic semantic properties that many cognitive phenomena
appear to have.

.I
(2) Generality:
.R
Not every problem amounts to pattern learning. Some cognitive tasks
may call for problem-specific rules, symbol manipulation, and
standard computation.

.I
(3) "Neurosimilitude" :
.R
Connectionism's brain-likeness may be superficial and may (like toy
models) camoflauge deeper performance limitations.

.B
4.3 Strengths of Symbol Systems:

.R
.I
.I
(1) Symbolic Function:
.R
Symbols have the computing power of Turing Machines and
the systematic properties of a formal syntax that is semantically
interpretable.

.I
(2) Generality:
.R
All computable functions (including all cognitive functions) are
equivalent to a computational state in a Turing Machine.

.I
(3) Practical Successes:
.R
Symbol systems' ability to generate intelligent behavior is
demonstrated by the successes of Artificial Intelligence.

.B
4.4 Weaknesses of Symbol Systems:

.R
.I
(1) Symbolic Function:
.R
Symbol systems are subject to the symbol grounding problem.

.I
(2) Generality:
.R
Turing power is too general. The solutions to AI's many toy problems do
not give rise to common principles of cognition but to a vast variety
of ad hoc symbolic strategies.

Having compared the strengths and weaknesses of the two approaches to
modeling cognition, let us now look more closely at the behavioral
capacities that human cognition subserves.

.ce1
.B
5. Human Behavioral Capacity
.R

Since the advent of cognitivism, psychologists have continued to gather
behavioral data, although to a large extent the relevant
evidence is already in: We already know what human beings are able to
do. They can (1)
.I discriminate,
(2)
.I manipulate,\**
.FS
.ls2
.nh
.na
Although they are no doubt as important as perceptual skills, motor
skills will not be explicitly considered here. It is assumed that
the relevant features of the sensory story (e.g., iconicity) will
generalize to the motor story. In addition, large parts of the motor
story may not be cognitive, drawing instead upon innate motor patterns
and sensorimotor feedback. Gibson's (1979) concept of "affordances" --
the invariant stimulus features that are detected by the motor
possibilities they "afford" -- is relevant here, though Gibson
underestimates the processing problems involved in finding the
invariants (Ullman 1980).
.FE
(3)
.I identify
and (4)
.I describe
the objects, events and
states of affairs in the world they live in, and that they can also (5)
.I "produce descriptions"
and (6)
.I "respond to descriptions"
of those objects,
events and states of affairs. Cognitive
theory's burden is now to explain
.I how
human beings (or any other devices) do all this.\**
.FS
.ls2
.nh
.na
If a candidate model were to exhibit all these behavioral capacities,
both
.I linguistic
(5-6)
and
.I robotic
(i.e., sensorimotor),
(1-3)
it would pass the
"Total Turing Test" (Harnad 1989).
The standard Turing Test (Turing 1964) calls for linguistic performance
capacity only: symbols in and symbols out. This makes it equivocal
about the status, scope and limits of pure symbol manipulation and
hence subject to the symbol grounding problem. A model that could pass
the Total Turing Test, however, would be grounded in the world.
.FE

.B "5.1 Discrimination and Identification."
Let us first look more closely at discrimination and identification.
To be able to
.I discriminate
is to able to judge whether two inputs are
the same or different, and, if different, 
.I how
different they are. Discrimination is a relative judgment; it is our
ability to tell things apart and judge their degree of similarity. To
be able to
.I identify
is to be able to assign a unique (usually arbitrary) response -- a
"name" -- to a class of inputs, treating them all as equivalent or
invariant in some respect. Identification is an absolute judgment; it
is our ability to tell whether or not a given input is a member of a
particular
.I category.

Consider the symbol "horse." We are able, in viewing different horses
(or the same horse in different positions, or at different times) to
tell them apart and to judge which of them are more alike, and
even how alike they are. This is discrimination.
In addition, in viewing a horse, we can reliably call it a horse,
rather than, say, a mule or a donkey (or a giraffe, or a stone). This
is identification. What sort of internal representation would be needed
in order to generate these two kinds of performance?

.B "5.2 Iconic representations."
According to the model being proposed here (Harnad 1987b), our ability
to discriminate inputs depends on our forming
.I "iconic representations"
of them. These are internal analog transforms of the
projections of distal objects on our sensory surfaces (Shepard &
Cooper 1982). In the case of
horses (and vision), they would be analogs of the many shapes that horses
cast on our retinas.\**
.FS
.ls2
.nh
.na
There are many problems having to do with figure/ground
discrimination, smoothing, size constancy, shape constancy, etc., that
make the problem of discrimination much more complicated than what is
described here, but these do not change the basic fact that iconic
representations are a natural candidate substrate for our
capacity to discriminate.
.FE
Same/different judgments would be based on the sameness or difference
of these iconic representations, and similarity judgments would be
based on their degree of congruity. No homunculus is involved here;
simply a process of superimposing icons and registering their degree
of disparity. Nor are there memory problems, since the inputs are
either simultaneously present or available in rapid enough succession
to draw upon their persisting sensory icons.

So we need horse icons to discriminate horses. But what about
identifying them? Discrimination is independent of identification. I
could be discriminating things without knowing what they were. Will the
icon allow me to identify horses? Although there are theorists who
believe it would (Paivio 1986), I have tried to show why it could not
(Harnad 1982, 1987b). Except in a world where there were bold, unmistakable
natural discontinuities between the categories we had to (or chose to)
sort and identify, so that members of one category couldn't be confused
with members of another category, icons would be useless for
identification because there are too many of them and because they
blend continuously\**
.FS
.ls2
.nh
.na
Elsewhere (Harnad 1987a,b) I have tried to show how the phenomenon of
"categorical perception" could generate internal discontinuities where
there is external continuity. There is evidence that our perceptual
system is able to segment a continuum, such as the color spectrum, into
relatively discrete, bounded regions or categories. Physical
differences of equal magnitude are more discriminable across the
boundaries between these categories than within them. This boundary
effect, both innate and learned, may play an important role in the
representation of the elementary perceptual categories out of which the
higher-order ones are built.
.FE
into one another; hence it is an independent problem to
.I identify
which of them are icons of members of the category and which are not.
Icons of sensory projections are too unselective. For identification,
icons must be selectively reduced to those
.I "invariant features"
of the sensory projection that
will reliably distinguish a member of a category from any nonmembers
with which it could be confused. Let us call this category-specific
feature detector the
.I "categorical representation" .
In some cases these representations may be innate, but
since evolution could hardly anticipate all of
the categories we may ever need or choose to identify, most of these
feature-detectors must be learned from experience. In particular,
our categorical representation of a horse is probably a learned one.
(I will defer till section 6 the problem of how the invariant features
underlying identification might be learned.)

Note that both iconic and categorical representations are nonsymbolic.
The former are analog copies of the sensory projection, preserving
its "shape" faithfully; the latter are selectively filtered to preserve only
some of the features of the shape of the sensory projection: those that
reliably distinguish members from nonmembers of a category. But both
representations are still sensory and nonsymbolic. There is no
problem about their connection to the objects they pick
out: It is a purely causal connection, based on the relation between
distal objects, proximal sensory projections and the acquired internal
changes that result from a history of behavioral interactions with
them. Nor is there any question of semantic interpretation, or whether
the semantic interpretation is justified.
Iconic representations no more "mean" the objects of which they are the
projections than the image in a camera does. Both icons and
camera-images can of course be interpreted as meaning or standing for
something, but the interpretation would clearly be derivative rather
than intrinsic.\**
.FS
.ls2
.nh
.na
On the other hand, the resemblance on which discrimination performance
is based -- the degree of isomorphism between the icon and the sensory
projection, and between the sensory projection and the distal object --
seems to be intrinsic, rather than just a matter of interpretation. The
resemblance can be objectively characterized as the degree of
invertibility of the transformation from object to icon (Harnad 1987b).
.FE

.B "5.3 Symbolic Representations."
Nor can categorical representations yet be interpreted as "meaning"
anything. It is true that they pick out the class of objects they
"name," but the names do not have the systematic properties of symbols
and symbol systems described earlier. They are just an inert taxonomy.
For systematicity it must be possible to combine and recombine them 
rulefully into
propositions that can be semantically interpreted. "Horse" is so far
just an arbitrary response that is reliably made in the presence
of a certain category of objects. There is
no justification for interpreting it holophrastically as meaning "This
is a [member of the category] horse" when uttered in the presence of a
horse, because the other expected systematic properties of "this" and
"a" and the all-important "is" of predication are not exhibited by mere
passive taxonomizing. What would be required to generate these other
systematic properties? Merely that the 
grounded names in the category taxonomy be
strung together into 
.I propositions
about further category
membership relations. For example:

(1) Suppose the name "horse" is grounded by iconic and categorical
representations, learned from experience, that reliably discriminate
and identify horses on the basis of their sensory projections.

(2) Suppose "stripes" is similarly grounded.

Now consider that the following category can be constituted out of
these elementary categories by a symbolic description of category
membership alone:

(3) "Zebra" = "horse" & "stripes"

What is the representation of a zebra? It is just the symbol string
"horse & stripes." But because "horse" and "stripes" are grounded in their
respective iconic and categorical representations, "zebra"
inherits the grounding, through its grounded
.I symbolic
representation.
In principle, someone who had never seen a zebra (but had seen
and learned to identify horses and stripes) could identify a zebra on first
acquaintance armed with this symbolic representation alone (plus the
nonsymbolic representations of horses and stripes that ground it).

Once one has the grounded set of elementary symbols provided
by a taxonomy of names (and the iconic and categorical
representations that give content to the names and allow them to pick
out the objects they identify) the rest of the symbol strings
of a natural language can be generated by symbol composition alone,\**
.FS
.nh
.na
.ls2
with the help of some standard logical connectives such as negation,
conjunction, etc.
.FE
and they will all inherit the intrinsic grounding of the elementary set.
Hence, the ability to discriminate and categorize (and its
underlying nonsymbolic representations) has led naturally to the ability
to describe and to produce and respond to descriptions through
symbolic representations.

.ce1
.B
6. A Complementary Role for Connectionism
.R

The symbol grounding scheme just described has one prominent gap: No
mechanism has been suggested to explain how the all-important
categorical representations could be formed: How does the hybrid system
find the invariant features of the sensory projection that make it
possible to categorize and identify objects correctly?\**
.FS
.ls2
.nh
.na
Although it is beyond the scope of this paper to discuss it at length,
it must be mentioned that this question has often been begged in the
past, mainly on the grounds of "vanishing intersections." It has been
claimed that one cannot find invariant features in the sensory
projection because they simply do not exist: The intersection of all
the projections of the members of a category such as "horse" is empty.
The British empiricists have been criticized for thinking otherwise;
Wittgenstein's (1953) discussion of "games" and "family resemblances"
has been taken to have discredited their view. And current research on
human categorization (Rosch & Lloyd 1978) has been interpreted as
confirming that intersections vanish and hence categories are not
represented in terms of invariant features. The problem of vanishing
intersections has even been cited by thinkers such as Fodor (1985,
1987) as a justification for extreme nativism. The present paper is
frankly empiricist. In my view, the reason intersections have not been
found is because no one has yet looked for them properly. Introspection
certainly isn't the way to look. And general pattern learning
algorithms such as connectionism are relatively new; their inductive
power remains to be tested. In addition, a careful distinction has not
been made between pure sensory categories (which, I claim, must have
invariants, otherwise we could not successfully identify them as we do)
and higher-order categories that are
.I grounded
in sensory categories; the latter's representations may be symbolic
rather than sensory, and hence not based directly on sensory
invariants. For further discussion of this problem, see Harnad 1987b).
.FE
Connectionism, with its general pattern learning capability, seems to
be the most natural candidate. Icons, paired with feedback indicating
their names, could be processed by a connectionist network that learns
to identify icons correctly by dynamically adjusting the weights on the
features and feature combinations reliably associated with the names,
reducing the icons to these invariant features of the category to which
they are assigned. In effect, the "connection" between the names and
the objects that give rise to their sensory projections and their icons
would be provided by connectionist networks.

This circumscribed complementary role for connectionism in a hybrid
system seems to remedy the weaknesses of the two current competitors
in their attempts to model the mind independently.
In a pure symbolic model the crucial connection between
the symbols and their referents is missing; an autonomous symbol
system, though amenable to a systematic semantic interpretation, is
ungrounded.
In a pure connectionist model, names are connected to
objects through invariant patterns in their sensory projections,
learned through exposure and feedback, but the crucial compositional
property is missing; a network of names, though grounded, is not
amenable to a systematic semantic interpretation.
In the hybrid system proposed here, there is no longer any autonomous
symbolic level at all; instead, there is an intrinsically dedicated
symbol system, its elementary symbols (names) connected to nonsymbolic
representations that can pick out the objects to which they refer, via
connectionistic networks that extract the invariant features of their
analog sensory projections.

.ce1
.B
7. Conclusions
.R

The expectation has often been expressed that "top-down" (symbolic)
approaches to cognition will somehow meet "bottom-up" (sensory)
approaches somewhere in between. If the grounding considerations in
this paper are valid, this expectation is hopelessly modular and there
is really only one viable route from sense to symbols: from the ground
up. A free-floating symbolic level like the software level of a
computer will never be reached by this route -- nor is it clear why we
should try to reach such a level, since getting there may well be
tantamount to unmooring our symbols from their intrinsic meanings
(and reducing ourselves to the functional equivalent of a programmable
computer).

In an intrinsically dedicated symbol system there are more constraints
on the symbol tokens than merely syntactic ones. Symbols are
manipulated not only on the basis of the arbitrary shape of their
tokens, but also on the basis of the decidedly nonarbitrary
"shape" of the iconic and categorical representations connected to
the grounded elementary symbols out of which the higher-order
symbols are composed. Of these two kinds of constraints, the
iconic/categorical ones are primary. I am not aware of any
formal analysis of such dedicated symbol systems, but this may be
because they are unique to cognitive and robotic modeling and their
properties will depend on the specific kinds of robotic (i.e., behavioral)
capacities they are designed to exhibit.

It is appropriate that the properties of dedicated symbol systems
should turn out to depend on behavioral considerations. The present
grounding scheme is still in the spirit of behaviorism in that the
only tests proposed for whether a semantic interpretation will bear the
semantic weight placed on it consist of one formal one (does it meet
the eight criteria for being a symbol system?) and one behavioral one
(can it discriminate, identify and describe all the objects and states
of affairs to which its symbols refer?). If both tests are passed, then
the semantic interpretation of its symbols is "fixed" by the behavioral
capacity of the dedicated symbol system, as exercised on the objects
and states of affairs in the world to which its symbols refer; the
symbol meanings are accordingly not just parasitic on the meanings in the
head of the interpreter, but intrinsic to the dedicated symbol system
itself. This is still no guarantee that our model has captured subjective
meaning, of course. But if the system's behavioral capacities are
lifesize, it's as close as we can ever hope to get.

.ce1
.B References

.nh
.na
Catania, A. C. & Harnad, S. (eds.) (1988)
.I
The Selection of Behavior.
The Operant Behaviorism of B. F. Skinner: Comments and Consequences.
.R
New York: Cambridge University Press.

Chomsky, N. (1980) Rules and representations.
.I "Behavioral and Brain Sciences"
3: 1-61.

Dennett, D. C. (1983)
Intentional systems in cognitive ethology.
.I
Behavioral and Brain Sciences 6:
.R
343 - 90.

Fodor, J. A. (1975)
.I "The language of thought"
New York: Thomas Y. Crowell

Fodor, J. A. (1980) Methodological solipsism considered as a
research strategy in cognitive psychology.
.I "Behavioral and Brain Sciences
3: 63 - 109.

Fodor, J. A. (1985) Pr\*'ecis of "The Modularity of Mind."
.I "Behavioral and Brain Sciences"
8: 1 - 42.

Fodor, J. A. (1987)
.I Psychosemantics
Cambridge MA: MIT/Bradford.

Fodor, J. A. & Pylyshyn, Z. W. (1988) Connectionism and cognitive
architecture: A critical appraisal.
.I Cognition
28: 3 - 71.

Gibson, J. J. (1979)
.I "An ecological approach to visual perception."
Boston: Houghton Mifflin

Harnad, S. (1982) Metaphor and mental duality.
In T. Simon & R. Scholes, R. (Eds.)
.I "Language, mind and brain."
Hillsdale, N. J.: Lawrence Erlbaum Associates

Harnad, S. (1987a) Categorical perception: A critical overview.
In S. Harnad (Ed.)
.I "Categorical perception: The groundwork of Cognition."
New York: Cambridge University Press

Harnad, S. (1987b) Category induction and representation.
In S. Harnad (Ed.)
.I "Categorical perception: The groundwork of Cognition."
New York: Cambridge University Press

Harnad, S. (1989) Minds, Machines and Searle.
.I "Journal of Theoretical and Experimental Artificial Intelligence"
1: 5-25.

Haugeland, J. (1978) The nature and plausibility of cognitivism.
.I "Behavioral and Brain Sciences"
1: 215-260.

Kleene, S. C. (1969)
.I "Formalized recursive functionals and formalized realizability."
Providence, R.I.: American Mathematical Society.

McClelland, J. L., Rumelhart, D. E., and the PDP Research Group (1986)
.I "Parallel distributed processing: Explorations in the
microstructure of cognition,"
Volume 1. Cambridge MA: MIT/Bradford.

Miller, G. A. (1956) The magical number seven, plus or minus two: Some
limits on our capacity for processing information.
.I "Psychological Review"
63: 81 - 97.

Minsky, M. (1961) Steps towards artificial intelligence.
.I "Proceedings of the Institute of Radio Engineers"
49: 8 - 30.

Minsky, M. & Papert, S. (1969)
.I "Perceptrons: An introduction to computational geometry."
Cambridge MA: MIT Press (Reissued in an Expanded Edition, 1988).

Newell, A. (1980) Physical Symbol Systems.
.I "Cognitive Science 4:"
135 - 83.

Neisser, U. (1967)
.I "Cognitive Psychology"
NY: Appleton-Century-Crofts.
.I
Cognitive Psychology
.R

Paivio, A. (1986)
.I "Mental representation: A dual coding approach."
New York: Oxford 

Rosch, E. & Lloyd, B. B. (1978)
.I "Cognition and categorization."
Hillsdale NJ: Erlbaum Associates

Rosenblatt, F. (1962)
.I
Principles of neurodynamics.
.R
NY: Spartan
Searle, J. R. (1980) Minds, brains and programs.
.I "Behavioral and Brain Sciences"
3: 417-457.

Shepard, R. N. & Cooper, L. A. (1982)
.I "Mental images and their transformations."
Cambridge: MIT Press/Bradford.

Stabler, E. P. (1985) How are grammars represented?
.I "Behavioral and Brain Sciences"
6: 391-421.

Terrace, H. (1979)
.I Nim.
NY: Random House.

Turkkan, J. (1989) Classical conditioning: The new hegemony.
.I "Behavioral and Brain Sciences 12:"
121 - 79.

Turing, A. M. (1964) Computing machinery and intelligence. In:
.I "Minds and machines,
A.R. Anderson (ed.), Engelwood Cliffs NJ: Prentice Hall.

Ullman, S. (1980) Against direct perception.
.I "Behavioral and Brain Sciences"
3: 373 - 415.

Wittgenstein, L. (1953)
.I "Philosophical investigations."
New York: Macmillan

∂18-Jul-89  1219	Mailer 	re: missing the point 
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Jul 89  12:19:10 PDT
Date: Tue 18 Jul 89 12:18:35-PDT
From: Harinder Singh <H.HARRY@Macbeth.Stanford.EDU>
Subject: re: missing the point 
To: JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
In-Reply-To: <juZ49@SAIL.Stanford.EDU>
Message-ID: <12511033315.77.H.HARRY@Macbeth.Stanford.EDU>


	Thanks for fessing up, Doc :-)

		I.

-------

∂18-Jul-89  1300	JMC  
glue together some expert and common sense papers

∂18-Jul-89  1329	wab@sumex-aim.stanford.edu 	re: missing the point 
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 18 Jul 89  13:28:56 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA08334; Tue, 18 Jul 89 13:31:02 PDT
Date: Tue, 18 Jul 1989 13:31:01 PDT
From: "William A. Brown" <wab@sumex-aim.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Cc: wab@sumex-aim.stanford.edu, su-etc@sumex-aim.stanford.edu
Subject: re: missing the point 
In-Reply-To: Your message of 18 Jul 89 1127 PDT 
Message-Id: <CMM.0.88.616797061.wab@sumex-aim.stanford.edu>

JMC,
  The article wasn't "addressed" to any particular audience, except the
literate audience.
  Furthermore, it doesn't make the Black audience want to murder the first
White it sees - it DOES make one want to insist that the history books show
true history - and that is a good thing.

  History is used to guide our present policies. As such, it should truly
reflect the past. By ignoring many of the past injustices, we doom ourselves
to repeat them. That may or may not be bad from your standpoint, but from
mine, I'll fight that to the teeth.

 Finally, poverty causes the high mortality rate among Black infants.
Furthermore, if you are going to compare death rates, you compare WITHIN a
SEX. As you know, women in general live longer than men. In addition, it is
the political policies of the mostly White legislature TODAY that keeps
infant mortality high in impoverished communities. It has long been proven
than the government's prenatal care programs for poor mothers was much more
economically feasible than caring for those same children born with a low
birth weight. The low birth weight children costs the givernment more to take
care of during those first critical weeks of life. Ignoring the moral issue
of when are we as a nation going to say that any expectant mother has a righ
to good prenatal care despite her ability to pay, I mean let's just ignore
that argument for a moment, I want to know why the US legislature selects for
the more costly alternative or raising the number of low birth-weight babies?

Besides that, MRC, their are so many racial barriers AFTER birth that limit
the QUALITY and QUANTITY of life for many Americans that just asking
aboutlife expectantcy is ridiculous. I suppose you would be as happy as a
janitor, if you could live as long?


--Bill

∂18-Jul-89  1328	Mailer 	Re: missing the point 
Received: from psych.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Jul 89  13:28:47 PDT
Received: by psych.Stanford.EDU (3.2/4.7); Tue, 18 Jul 89 13:26:33 PDT
From: davidson@psych.Stanford.EDU (Martin Davidson)
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Cc: sierra!singh@psych.Stanford.EDU, su-etc@SAIL.Stanford.EDU,
        davidson@PSYCH.Stanford.EDU
Subject: Re: missing the point 
In-Reply-To: Your message of 18 Jul 89 11:27:00 PDT.
             <juZ49@SAIL.Stanford.EDU> 
Date: Tue, 18 Jul 89 13:26:31 PDT


JMC writes:

>Mr. Singh is correct that my response to the Walker article was
>dismissive.  This was intentional.  Here are some further comments.

>1. It isn't clear whether the article is addressed to a black audience
>or to a white audience.  If the former it is destructive for two
>reasons.  First, it generates hostility.

Walker's piece did not evoke hostility in me--your message did.  My
reaction to Walker's piece was complex ranging from sadness through
anger to inspiration.

>Second, it puts the
>responsibility for reducing black infant mortality entirely on
>white men.  While white men (e.g. doctors) are doing something about
>reducing black infant mortality, much more can be done by black
>women avoiding behavior destructive of themselves and their
>children.  Alice Walker's article diverts black political
>resources from productive activities to unproductive ones.

I believe your interpretation to be valid, but I did not interpret it
the same way.  I saw her making connections between the way black women
have been treated in white Euro-American male dominated societies
historically and the present conditions in the world.  I believe such
connections are important points of departure for understanding the
plight of women and men of color in the US and abroad.  I did not see
Walker address the issue of black women's responsibility in this
dilemma.  She would not deny such responsibility, but that is not
what this piece is about.

>If the latter [directed to white audiences] it is one more harangue
>of a kind of which we
>have all seen many.  Imagine, however, that someone, say
>u.underdog, hasn't seen the like before and is bowled over
>by it into great feelings of guilt.  The behavior it would
>excuse would be some expression of anti-establishment rage,
>also not likely to improve the situation.

Please do not assume that feelings of guilt (or defensiveness about
that guilt) is evoked for any white person who reads the piece.  I
doubt you have evidence for such an assertion,  and I have numerous
counterexamples.  Your hypothetical u.underdog is a single white
individual and I would hope that you let him deal with his feelings
in any he deems appropriate.

>2. Am I included in the white men Alice Walker is attacking?
>Both my parents immigrated to the U.S. after 1900.  
>I suspect I am not supposed to escape - being of the
>wrong color.

Are you or your parents innocents of racial prejudice?  If not, then
I can see how the article could speak to you.  It is beyond my
imagination to consider the other venue (not meant to be a personal
attack--I know of no one, myself included, who is innocent of
racial/group prejudice).

...

>5. Mr. Davidson accuses me of insensitivity to this factual and
>emotional article.  He is right; I don't belong to his liberal
>religion, and the insensitivity is intentional.  The sentiment
>expressed in the article has killed millions of people.  It can
>serve as a mobilizing call to a totalitarian movement.  I suppose
>the National Council of Churches staff people who helped Puerto
>Rican nationalist terrorists make bombs and plan bombings were
>motivated by emotions like those the article is supposed to
>arouse.

>The "facts" cited are selective in support of the emotion.
>Many of them aren't facts.

JMC, What's with all of this guilt stuff?  I don't believe I ever
said that you or any other white man should feel guilty about the
circumstances Walker addresses.

First, are you saying that the facts Walker brings up are false?  If
you believe that, I suggest you check basic historical sources
(I apologize for not producing the references, but I trust you
are resourceful enough to find them if you choose).  If you
concede that even some of the facts are true (i.e., that a power structure
dominated by white males of European descent has inflicted oppression
on people of color), then the ball is in your
court.  You may feel guilty, or defensive or
forthright, or nothing at all in reponse to the validity of these
statements.  That really is your choice.  But you
will deal with those facts in some way.

My request was merely that that you do so in a more thoughful
manner than is reflected by the pointless question about mortality rates
(which have done somewhat in your reply).  I make that
request not because I want your admission of guilt (frankly, I've had
enough admissions of guilt from white folks to last me a lifetime), but
I want constructive discourse about these issues, a value I believe
we share.

-MD

P.S.  By the way, I understand your points about the potential danger
of purely emotional responses--it's a point well taken.  I must also
ask if you deny the either the importance and power of emotion solely
because it is sometimes misused?  THrowing the baby out with the
bathwater?  Would you argue that intellect is not also a tool of
oppression? (Perhaps this is another discussion!)

P.P.S.  I regret that I have not read Inder's posting at the sending
of this message.

∂18-Jul-89  1458	A.ABIE@Macbeth.Stanford.EDU 	Re: black doctors    
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Jul 89  14:57:57 PDT
Date: Tue 18 Jul 89 14:57:00-PDT
From: ABE DEANDA <A.ABIE@Macbeth.Stanford.EDU>
Subject: Re: black doctors 
To: wab@sumex-aim.stanford.edu
cc: JMC@sail.stanford.edu, su-etc@sumex-aim.stanford.edu
In-Reply-To: <CMM.0.88.616791921.wab@sumex-aim.stanford.edu>
Message-ID: <12511062154.85.A.ABIE@Macbeth.Stanford.EDU>

Bill, you write:

>Heck, this medical school only gives East Coast interviews in the
>Boston area - as if those are the only schools that exist.  This
>medical school certainly doesn't recruit at Black universities.

Hate to ruin a good point (I think your general point is well taken),
but Stanford interviews at other places besides Boston, and does
recruit from Black university(s) .. (the parentheses since the major
effort is at Howard).  I know this since I did interview trips for
the admission office last year at both New York city and Washington, D.C.

Abe
-------

∂18-Jul-89  1856	Mailer 	re: greenhouse effect 
To:   JMC@SAIL.Stanford.EDU, U.UNDERDOG@MACBETH.Stanford.EDU,
      su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

[In reply to message from JMC rcvd 15-Jul-89 13:18-PT.]

The difficulty with proposals to generate and save more wood
in the rain forests is that those forests are almost completely
nutrient-limited, their soils being very poor and the nutrients
being almost completely locked up in existing vegetation.

∂19-Jul-89  0126	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	cf of Brillhart surds   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 19 Jul 89  01:26:24 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 402201; Wed 19-Jul-89 04:24:47 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 102822; Wed 19-Jul-89 00:58:31 PDT
Date: Wed, 19 Jul 89 00:58 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: cf of Brillhart surds
To: math-fun@RUSSIAN.SPA.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890719075822.2.RWG@TSUNAMI.SPA.Symbolics.COM>

The Chudnovskys reminded me that the real root of x↑3-8x-10
has bizarrely large partial quotients among its first 162.

Through sheer wimp force search, I found 11*X↑3-35*X↑2+43*X+15,
which has 1.2 billion as its 124th.  7x↑3+30x↑2+98 is pretty good,
too.  A somewhat better algebraic integer satisfies x↑3+x↑2+23x-27.

All these are just homographic transformations of Brillhart's
polynomial, which has something to do with modular identities
(i.e. q-land).

∂19-Jul-89  0217	Mailer 	re: Protecting JMC [was Re: missing the point] 
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Jul 89  02:17:08 PDT
Date: Wed 19 Jul 89 02:16:26-PDT
From: Harinder Singh <H.HARRY@Macbeth.Stanford.EDU>
Subject: re: Protecting JMC [was Re: missing the point] 
To: JMC@SAIL.Stanford.EDU
cc: su-etc@SAIL.Stanford.EDU
In-Reply-To: <Fv05V@SAIL.Stanford.EDU>
Message-ID: <12511185840.79.H.HARRY@Macbeth.Stanford.EDU>


	Prof. McCarthy complains that I haven't answered his
questions. Those who find being ``dismissive'' an acceptable
tactic for their own use cannot expect sympathy when the roles
are reversed.

	Why is Prof. McCarthy unhappy about this?

		I.

-------

∂19-Jul-89  1050	cliff%unix.computer-science.manchester.ac.uk@NSFnet-Relay.AC.UK 	TESTING   
Received: from NSFnet-Relay.AC.UK by SAIL.Stanford.EDU with TCP; 19 Jul 89  10:50:32 PDT
Received: from unix.computer-science.manchester.ac.uk by NSFnet-Relay.AC.UK 
           via Janet with NIFTP  id aa08081; 19 Jul 89 17:50 BST
From: Cliff Jones <cliff%unix.computer-science.manchester.ac.uk@NSFnet-Relay.AC.UK>
Date: Wed, 19 Jul 89 18:08:24
To: jmc@sail.stanford.edu
Subject: TESTING

John,
	Thanks for the chat - don't be surprised if you get no reply for a
	while - I'm off to France on Friday for 2 weeks.

cliff jones

∂19-Jul-89  1139	pi-data@vax.darpa.mil    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 19 Jul 89  11:39:45 PDT
Received: from sun44.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA05267; Wed, 19 Jul 89 14:38:41 EDT
Date: Wed, 19 Jul 89 14:39:44 EDT
From: pi-data@vax.darpa.mil
Posted-Date: Wed, 19 Jul 89 14:39:44 EDT
Message-Id: <8907191839.AA05937@sun44.darpa.mil>
Received: by sun44.darpa.mil (4.1/5.51)
	id AA05937; Wed, 19 Jul 89 14:39:44 EDT
To: jmc@sail.stanford.edu
Cc: pi-data@vax.darpa.mil

MEMO FOR PI OF TASK: QLISP FOR PARALLEL PROCESSORS
          TASK CODE: BOE

In order to avoid the yearly rush to collect financial data in
support of Fall funding, we are re-instituting last year's 
reporting process with the intention of turning it into a 
routine, quarterly process.  In the future we intend to maintain
the information in a database, and extract update information
automatically from quarterly email reports (which will require 
less information than this initial report).

This report is not a requirement of your contract or grant and does
not take the place of any report required in the contract or grant.  
We are asking for this information in order to do a better job of
managing our funds in order to support you better.

Please send your report to pi-data@vax.darpa.mil, not later than 
28 July 89.  Questions on the report procedure may also be directed
to that address.  If absolutely necessary telephone inquiries may be
directed to Louise Lorenzen at 202-695-1766.  

If you do not have email available, mail to PI-Data,
DARPA/ISTO, 1400 Wilson Blvd, Arlington, VA 22209. 

We try to keep the reporting workload on the research community as
light as possible.  Thank you for helping us support you.  

                DARPA/ISTO Program Management

---------------------------------------------------------------------
Here is the information we need: (Two minus signs denote
comments. In your response, you may annotate with 
comments headed with --. However, please leave the format of the
reporting lines exactly as shown.)

CODE:  Task code from first line of this message (THIS IS IMPORTANT)
TITLE: Task name from first line of this message
CONTRACTOR:  Contractor name (e.g. company or university)
PI-NAME:  PI name
PI-EMAIL:  PI email address
PI-PHONE:  PI phone number
ADMIN-NAME:  Admin name
ADMIN-EMAIL:  Admin email address
ADMIN-PHONE:  Admin phone number
ARPA-ORDER: ARPA Order number
CONTRACT: Contract number
START: Contract start date
EARLYSTART: Date of pre-contract work authorization(omit if none)
END: Planned end date for effort on current contract
VALUE: Contract value (include funded options)
AUTH-RCD: Total spending authority received to date
-- For following line
-- If you know that the agent or others has directly contributed
-- to this contract/task:  show total amount of non-DARPA funding
-- to date.
-- If you know there was jointfunding but don't know how much:
-- include the line JOINTFUNDING with no 5930.
-- If there was no joint funding, you are unsure, or don't
-- know what I am talking about: omit the JOINTFUNDING line.
JOINTFUNDING: joint funding 
BILLED: Total amount billed to government as of 1 July 89
-- The following lines are to establish a profile of spending on the
-- project. They are intended to be used by the program manager for the
-- estimation of his budget requirements for this year and for outyears.
-- NOTE: We ask for these numbers because they are likely to have changed
-- since the project was selected for funding.  These numbers will be
-- treated as your estimate of required funding for the project as is
-- currently agreed.  They need not match the profile listed in the
-- contract nor will they be considered contractually binding.  They will
-- form a basis for the managers estimate of nextyear funding. 
-- 
-- Each line represents one calendar year. The four fields represent the
-- four calendar quarters.  (Report in thousands of dollars) There should
-- be one line for each calendar during which this contract/task is
-- active.  These values represent your actual expenses (for past
-- quarters) and your estimate of expenses (for future quarters).  Use
-- RECUR for recurring costs such as labor/travel/maintenance/overhead
-- use NONRECUR for large "one shot" expenses such as equipment
-- purchases.  Generally NONRECUR will be zero except for quarters with
-- major purchases.
RECUR: CY88 ddd ddd ddd ddd
RECUR: CY89 ddd ddd ddd ddd
RECUR: CY90 ddd ddd ddd ddd
RECUR: CY91 ddd ddd ddd ddd
RECUR: CY92 ddd ddd ddd ddd
NONRECUR: CY88 ddd ddd ddd ddd 
NONRECUR: CY89 ddd ddd ddd ddd
NONRECUR: CY90 ddd ddd ddd ddd
NONRECUR: CY91 ddd ddd ddd ddd
NONRECUR: CY92 ddd ddd ddd ddd
-- The following allows planning for options.
-- if there are unselected options that may be invoked at a later
-- time, include them here.
-- Repeat the group of OPTION-NAME and OPTION for each unselected
-- option. (Note: options currently on contract should be covered above).
-- If there are no options then omit this section.
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
-- The following is general remarks. Caveats. Comments. Suggestions ....
REMARKS: This is free-form and is terminated by end of message.
------------------------------------------------------------------------
SPECIAL NOTES: 

a. We would like the admin email address to be one which is read
   frequently, whether the PI is available or not.  This will
   provide a backup contact for actions which must be completed
   quickly.

b. All amounts are in thousands of dollars. If any item is zero,
   please provide it anyhow. Punctuation in dollar values is optional:
   dollar signs, commas and "K" will be ignored.

c. Data pertain to current contract or grant identified by the
   task code; please do not report amounts for planned/pending
   renewals. 

d. END should be at or near end of contract period.  If it is 
   after end of contract, you should request a no-cost extension
   so contract covers period of performance.  Through your Program
   Manager separately.

e. For tasking or requirements contracts, data refer to
   just your task, not the whole contract.  Please identify any 
   designator established by the Contracting Officer in CONTRACT. 
   (This will NOT be the same as the Task Code DARPA/ISTO is using 
   to identify your effort.)

f. We plan to input directly to our database from your message, 
   so please use the exact names above, and separate data from 
   description with a colon (THIS IS IMPORTANT). 

g. Please follow this example as closely as possible:

CODE: Z1
TITLE: Advanced Computing Environment
CONTRACTOR: Creative State University
PI-NAME: J.J. Smith
PI-EMAIL: j.smith@cs.creative.edu
PI-PHONE: 123-555-1234
ADMIN-NAME:: M.A. Jones
ADMIN-EMAIL: contract@cs.creative.edu
ADMIN-PHONE: 123-555-4321
ARPA-ORDER: 1432
CONTRACT: N00139-87-C-0876
START: 10 Apr 89
EARLYSTART: 1 Apr 89
END: 9 Apr 92
VALUE: 11K
AUTH-RCD: 50K
BILLED: 36K
RECUR: CY89  0 31 45 50
RECUR: CY90 50 50 50 50
RECUR: CY91 50 50 65 65
RECUR: CY92 50  5  0  0
NONRECUR: CY89 0 120 0 0 
NONRECUR: CY90 0 0 80 0
NONRECUR: CY91 0 0 0 0
NONRECUR: CY92 0 0 0 0
OPTION-NAME: More Creative Stuff
OPTION: CY91 0 0 30 80
OPTION: CY92 120 50 5 0
REMARKS: Spending behind plan due to trouble hiring programmer.
	 We are planning to get new programmer sometime next month.
--END OF MESSAGE----

∂19-Jul-89  1203	pi-data@vax.darpa.mil    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 19 Jul 89  12:03:27 PDT
Received: from sun49.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA05537; Wed, 19 Jul 89 15:02:26 EDT
Date: Wed, 19 Jul 89 15:05:32 EDT
From: pi-data@vax.darpa.mil
Posted-Date: Wed, 19 Jul 89 15:05:32 EDT
Message-Id: <8907191905.AA00709@sun49.darpa.mil>
Received: by sun49.darpa.mil (4.1/5.51)
	id AA00709; Wed, 19 Jul 89 15:05:32 EDT
To: jmc@sail.stanford.edu
Cc: pi-data@vax.darpa.mil

MEMO FOR PI OF TASK: QLISP FOR PARALLEL PROCESSORS
          TASK CODE: BOE

In order to avoid the yearly rush to collect financial data in
support of Fall funding, we are re-instituting last year's 
reporting process with the intention of turning it into a 
routine, quarterly process.  In the future we intend to maintain
the information in a database, and extract update information
automatically from quarterly email reports (which will require 
less information than this initial report).

This report is not a requirement of your contract or grant and does
not take the place of any report required in the contract or grant.  
We are asking for this information in order to do a better job of
managing our funds in order to support you better.

Please send your report to pi-data@vax.darpa.mil, not later than 
28 July 89.  Questions on the report procedure may also be directed
to that address.  If absolutely necessary telephone inquiries may be
directed to Louise Lorenzen at 202-695-1766.  

If you do not have email available, mail to PI-Data,
DARPA/ISTO, 1400 Wilson Blvd, Arlington, VA 22209. 

We try to keep the reporting workload on the research community as
light as possible.  Thank you for helping us support you.  

                DARPA/ISTO Program Management

---------------------------------------------------------------------
Here is the information we need: (Two minus signs denote
comments. In your response, you may annotate with 
comments headed with --. However, please leave the format of the
reporting lines exactly as shown.)

CODE:  Task code from first line of this message (THIS IS IMPORTANT)
TITLE: Task name from first line of this message
CONTRACTOR:  Contractor name (e.g. company or university)
PI-NAME:  PI name
PI-EMAIL:  PI email address
PI-PHONE:  PI phone number
ADMIN-NAME:  Admin name
ADMIN-EMAIL:  Admin email address
ADMIN-PHONE:  Admin phone number
ARPA-ORDER: ARPA Order number
CONTRACT: Contract number
START: Contract start date
EARLYSTART: Date of pre-contract work authorization(omit if none)
END: Planned end date for effort on current contract
VALUE: Contract value (include funded options)
AUTH-RCD: Total spending authority received to date
-- For following line
-- If you know that the agent or others has directly contributed
-- to this contract/task:  show total amount of non-DARPA funding
-- to date.
-- If you know there was jointfunding but don't know how much:
-- include the line JOINTFUNDING with no 702.
-- If there was no joint funding, you are unsure, or don't
-- know what I am talking about: omit the JOINTFUNDING line.
JOINTFUNDING: joint funding 
BILLED: Total amount billed to government as of 1 July 89
-- The following lines are to establish a profile of spending on the
-- project. They are intended to be used by the program manager for the
-- estimation of his budget requirements for this year and for outyears.
-- NOTE: We ask for these numbers because they are likely to have changed
-- since the project was selected for funding.  These numbers will be
-- treated as your estimate of required funding for the project as is
-- currently agreed.  They need not match the profile listed in the
-- contract nor will they be considered contractually binding.  They will
-- form a basis for the managers estimate of nextyear funding. 
-- 
-- Each line represents one calendar year. The four fields represent the
-- four calendar quarters.  (Report in thousands of dollars) There should
-- be one line for each calendar during which this contract/task is
-- active.  These values represent your actual expenses (for past
-- quarters) and your estimate of expenses (for future quarters).  Use
-- RECUR for recurring costs such as labor/travel/maintenance/overhead
-- use NONRECUR for large "one shot" expenses such as equipment
-- purchases.  Generally NONRECUR will be zero except for quarters with
-- major purchases.
RECUR: CY88 ddd ddd ddd ddd
RECUR: CY89 ddd ddd ddd ddd
RECUR: CY90 ddd ddd ddd ddd
RECUR: CY91 ddd ddd ddd ddd
RECUR: CY92 ddd ddd ddd ddd
NONRECUR: CY88 ddd ddd ddd ddd 
NONRECUR: CY89 ddd ddd ddd ddd
NONRECUR: CY90 ddd ddd ddd ddd
NONRECUR: CY91 ddd ddd ddd ddd
NONRECUR: CY92 ddd ddd ddd ddd
-- The following allows planning for options.
-- if there are unselected options that may be invoked at a later
-- time, include them here.
-- Repeat the group of OPTION-NAME and OPTION for each unselected
-- option. (Note: options currently on contract should be covered above).
-- If there are no options then omit this section.
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
-- The following is general remarks. Caveats. Comments. Suggestions ....
REMARKS: This is free-form and is terminated by end of message.
------------------------------------------------------------------------
SPECIAL NOTES: 

a. We would like the admin email address to be one which is read
   frequently, whether the PI is available or not.  This will
   provide a backup contact for actions which must be completed
   quickly.

b. All amounts are in thousands of dollars. If any item is zero,
   please provide it anyhow. Punctuation in dollar values is optional:
   dollar signs, commas and "K" will be ignored.

c. Data pertain to current contract or grant identified by the
   task code; please do not report amounts for planned/pending
   renewals. 

d. END should be at or near end of contract period.  If it is 
   after end of contract, you should request a no-cost extension
   so contract covers period of performance.  Through your Program
   Manager separately.

e. For tasking or requirements contracts, data refer to
   just your task, not the whole contract.  Please identify any 
   designator established by the Contracting Officer in CONTRACT. 
   (This will NOT be the same as the Task Code DARPA/ISTO is using 
   to identify your effort.)

f. We plan to input directly to our database from your message, 
   so please use the exact names above, and separate data from 
   description with a colon (THIS IS IMPORTANT). 

g. Please follow this example as closely as possible:

CODE: Z1
TITLE: Advanced Computing Environment
CONTRACTOR: Creative State University
PI-NAME: J.J. Smith
PI-EMAIL: j.smith@cs.creative.edu
PI-PHONE: 123-555-1234
ADMIN-NAME:: M.A. Jones
ADMIN-EMAIL: contract@cs.creative.edu
ADMIN-PHONE: 123-555-4321
ARPA-ORDER: 1432
CONTRACT: N00139-87-C-0876
START: 10 Apr 89
EARLYSTART: 1 Apr 89
END: 9 Apr 92
VALUE: 11K
AUTH-RCD: 50K
BILLED: 36K
RECUR: CY89  0 31 45 50
RECUR: CY90 50 50 50 50
RECUR: CY91 50 50 65 65
RECUR: CY92 50  5  0  0
NONRECUR: CY89 0 120 0 0 
NONRECUR: CY90 0 0 80 0
NONRECUR: CY91 0 0 0 0
NONRECUR: CY92 0 0 0 0
OPTION-NAME: More Creative Stuff
OPTION: CY91 0 0 30 80
OPTION: CY92 120 50 5 0
REMARKS: Spending behind plan due to trouble hiring programmer.
	 We are planning to get new programmer sometime next month.
--END OF MESSAGE----

∂19-Jul-89  1225	pi-data@vax.darpa.mil    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 19 Jul 89  12:24:55 PDT
Received: from sun49.darpa.mil by vax.darpa.mil (5.54/5.51)
	id AA05891; Wed, 19 Jul 89 15:23:48 EDT
Date: Wed, 19 Jul 89 15:26:54 EDT
From: pi-data@vax.darpa.mil
Posted-Date: Wed, 19 Jul 89 15:26:54 EDT
Message-Id: <8907191926.AA00935@sun49.darpa.mil>
Received: by sun49.darpa.mil (4.1/5.51)
	id AA00935; Wed, 19 Jul 89 15:26:54 EDT
To: jmc@sail.stanford.edu
Cc: pi-data@vax.darpa.mil

MEMO FOR PI OF TASK: FORMAL REASONING RESEARCH
          TASK CODE: SIM

In order to avoid the yearly rush to collect financial data in
support of Fall funding, we are re-instituting last year's 
reporting process with the intention of turning it into a 
routine, quarterly process.  In the future we intend to maintain
the information in a database, and extract update information
automatically from quarterly email reports (which will require 
less information than this initial report).

This report is not a requirement of your contract or grant and does
not take the place of any report required in the contract or grant.  
We are asking for this information in order to do a better job of
managing our funds in order to support you better.

Please send your report to pi-data@vax.darpa.mil, not later than 
28 July 89.  Questions on the report procedure may also be directed
to that address.  If absolutely necessary telephone inquiries may be
directed to Louise Lorenzen at 202-695-1766.  

If you do not have email available, mail to PI-Data,
DARPA/ISTO, 1400 Wilson Blvd, Arlington, VA 22209. 

We try to keep the reporting workload on the research community as
light as possible.  Thank you for helping us support you.  

                DARPA/ISTO Program Management

---------------------------------------------------------------------
Here is the information we need: (Two minus signs denote
comments. In your response, you may annotate with 
comments headed with --. However, please leave the format of the
reporting lines exactly as shown.)

CODE:  Task code from first line of this message (THIS IS IMPORTANT)
TITLE: Task name from first line of this message
CONTRACTOR:  Contractor name (e.g. company or university)
PI-NAME:  PI name
PI-EMAIL:  PI email address
PI-PHONE:  PI phone number
ADMIN-NAME:  Admin name
ADMIN-EMAIL:  Admin email address
ADMIN-PHONE:  Admin phone number
ARPA-ORDER: ARPA Order number
CONTRACT: Contract number
START: Contract start date
EARLYSTART: Date of pre-contract work authorization(omit if none)
END: Planned end date for effort on current contract
VALUE: Contract value (include funded options)
AUTH-RCD: Total spending authority received to date
-- For following line
-- If you know that the agent or others has directly contributed
-- to this contract/task:  show total amount of non-DARPA funding
-- to date.
-- If you know there was jointfunding but don't know how much:
-- include the line JOINTFUNDING with no 702.
-- If there was no joint funding, you are unsure, or don't
-- know what I am talking about: omit the JOINTFUNDING line.
JOINTFUNDING: joint funding 
BILLED: Total amount billed to government as of 1 July 89
-- The following lines are to establish a profile of spending on the
-- project. They are intended to be used by the program manager for the
-- estimation of his budget requirements for this year and for outyears.
-- NOTE: We ask for these numbers because they are likely to have changed
-- since the project was selected for funding.  These numbers will be
-- treated as your estimate of required funding for the project as is
-- currently agreed.  They need not match the profile listed in the
-- contract nor will they be considered contractually binding.  They will
-- form a basis for the managers estimate of nextyear funding. 
-- 
-- Each line represents one calendar year. The four fields represent the
-- four calendar quarters.  (Report in thousands of dollars) There should
-- be one line for each calendar during which this contract/task is
-- active.  These values represent your actual expenses (for past
-- quarters) and your estimate of expenses (for future quarters).  Use
-- RECUR for recurring costs such as labor/travel/maintenance/overhead
-- use NONRECUR for large "one shot" expenses such as equipment
-- purchases.  Generally NONRECUR will be zero except for quarters with
-- major purchases.
RECUR: CY88 ddd ddd ddd ddd
RECUR: CY89 ddd ddd ddd ddd
RECUR: CY90 ddd ddd ddd ddd
RECUR: CY91 ddd ddd ddd ddd
RECUR: CY92 ddd ddd ddd ddd
NONRECUR: CY88 ddd ddd ddd ddd 
NONRECUR: CY89 ddd ddd ddd ddd
NONRECUR: CY90 ddd ddd ddd ddd
NONRECUR: CY91 ddd ddd ddd ddd
NONRECUR: CY92 ddd ddd ddd ddd
-- The following allows planning for options.
-- if there are unselected options that may be invoked at a later
-- time, include them here.
-- Repeat the group of OPTION-NAME and OPTION for each unselected
-- option. (Note: options currently on contract should be covered above).
-- If there are no options then omit this section.
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
-- The following is general remarks. Caveats. Comments. Suggestions ....
REMARKS: This is free-form and is terminated by end of message.
------------------------------------------------------------------------
SPECIAL NOTES: 

a. We would like the admin email address to be one which is read
   frequently, whether the PI is available or not.  This will
   provide a backup contact for actions which must be completed
   quickly.

b. All amounts are in thousands of dollars. If any item is zero,
   please provide it anyhow. Punctuation in dollar values is optional:
   dollar signs, commas and "K" will be ignored.

c. Data pertain to current contract or grant identified by the
   task code; please do not report amounts for planned/pending
   renewals. 

d. END should be at or near end of contract period.  If it is 
   after end of contract, you should request a no-cost extension
   so contract covers period of performance.  Through your Program
   Manager separately.

e. For tasking or requirements contracts, data refer to
   just your task, not the whole contract.  Please identify any 
   designator established by the Contracting Officer in CONTRACT. 
   (This will NOT be the same as the Task Code DARPA/ISTO is using 
   to identify your effort.)

f. We plan to input directly to our database from your message, 
   so please use the exact names above, and separate data from 
   description with a colon (THIS IS IMPORTANT). 

g. Please follow this example as closely as possible:

CODE: Z1
TITLE: Advanced Computing Environment
CONTRACTOR: Creative State University
PI-NAME: J.J. Smith
PI-EMAIL: j.smith@cs.creative.edu
PI-PHONE: 123-555-1234
ADMIN-NAME:: M.A. Jones
ADMIN-EMAIL: contract@cs.creative.edu
ADMIN-PHONE: 123-555-4321
ARPA-ORDER: 1432
CONTRACT: N00139-87-C-0876
START: 10 Apr 89
EARLYSTART: 1 Apr 89
END: 9 Apr 92
VALUE: 11K
AUTH-RCD: 50K
BILLED: 36K
RECUR: CY89  0 31 45 50
RECUR: CY90 50 50 50 50
RECUR: CY91 50 50 65 65
RECUR: CY92 50  5  0  0
NONRECUR: CY89 0 120 0 0 
NONRECUR: CY90 0 0 80 0
NONRECUR: CY91 0 0 0 0
NONRECUR: CY92 0 0 0 0
OPTION-NAME: More Creative Stuff
OPTION: CY91 0 0 30 80
OPTION: CY92 120 50 5 0
REMARKS: Spending behind plan due to trouble hiring programmer.
	 We are planning to get new programmer sometime next month.
--END OF MESSAGE----

∂19-Jul-89  1425	MRC@CAC.Washington.EDU 	re: White Man   
Received: from hanna.cac.washington.edu by SAIL.Stanford.EDU with TCP; 19 Jul 89  14:25:52 PDT
Received: from tomobiki-cho.cac.washington.edu by hanna.cac.washington.edu
	(5.61/UW-NDC Revision: 2.1 ) id AA00511; Wed, 19 Jul 89 14:26:24 -0700
Date: Wed, 19 Jul 1989 14:22:11 PDT
From: Mark Crispin <MRC@CAC.Washington.EDU>
Sender: mrc@Tomobiki-Cho.CAC.Washington.EDU
Subject: re: White Man
To: John McCarthy <JMC@sail.stanford.edu>
In-Reply-To: <Fue3Y@SAIL.Stanford.EDU>
Message-Id: <MailManager.616886531.899.mrc@Tomobiki-Cho.CAC.Washington.EDU>

In <Fue3Y@SAIL.Stanford.EDU>, John McCarthy writes:
>Everyone is responsible only for
>his own misbehavior.  We atheists don't agree with the idea
>of "visiting the sins of the fathers upon the sons".

You have my 1000% agreement on that.

-------

∂19-Jul-89  1439	MRC@CAC.Washington.EDU 	re: White Man   
Received: from hanna.cac.washington.edu by SAIL.Stanford.EDU with TCP; 19 Jul 89  14:39:02 PDT
Received: from tomobiki-cho.cac.washington.edu by hanna.cac.washington.edu
	(5.61/UW-NDC Revision: 2.1 ) id AA00583; Wed, 19 Jul 89 14:39:35 -0700
Date: Wed, 19 Jul 1989 14:34:50 PDT
From: Mark Crispin <MRC@CAC.Washington.EDU>
Sender: mrc@Tomobiki-Cho.CAC.Washington.EDU
Subject: re: White Man
To: John McCarthy <JMC@sail.stanford.edu>
In-Reply-To: <12vryM@SAIL.Stanford.EDU>
Message-Id: <MailManager.616887290.7886.mrc@Tomobiki-Cho.CAC.Washington.EDU>

In <12vryM@SAIL.Stanford.EDU>, John McCarthy writes:
>[In reply to message sent Wed, 19 Jul 1989 14:22:11 PDT.]
>You and what nine other people?

1000% being intended as the current venacular term for "enthusiastic agreement
equivalent to the complete agreement of 10 people".

-------

∂19-Jul-89  1524	Mailer 	re: question for those who favor de-forestation
To:   JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

[In reply to message from JMC rcvd 19-Jul-89 00:10-PT.]

I disagree that the rate of carbon fixation is the most
important parameter.  It's rate of fixation times the
time until the carbon decomposes into the atmosphere again.
That's in the long run.  In the short run if things a re 
getting sweaty going for a high fixation rate might make sense.
Ignoring carbon in the rocks and such, the carbon that does
not contribute to the greenhouse effect is that in organic
matter whether living, preserved, or rotting.  

∂19-Jul-89  1526	Mailer 	re: greenhouse effect 
To:   JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

[In reply to message from JMC rcvd 19-Jul-89 00:11-PT.]

Sure, fertilizer can be used, and the poor can eat cake,
but are these solutions cost-effective?

∂19-Jul-89  1536	RWF 	re: Nuclear waste disposal    
[In reply to message rcvd 18-Jul-89 23:47-PT.]

Ah, I was wondering about that.  But isn't there a bound on the 
energy transfer to the planet, and isn't that bound smallish?
My guess is that the best you can do is to reduce the speed by
the speed of an object orbiting the planet.  Much less than the planet's own
speed.  

∂19-Jul-89  1617	RWF 	re: Nuclear waste disposal    
[In reply to message rcvd 19-Jul-89 16:14-PT.]

I agree, at least qualitatively.  I think planetary velocities
are large compared to escape velocities from the planets, and that
this must limit the usefulness of the rebound effects.

∂19-Jul-89  1608	RWF 	re: greenhouse effect    
[In reply to message rcvd 19-Jul-89 16:02-PT.]

I guess what I'm asking is:
* What is the minimum ratio of phosphate (say) to carbon?
* What is the minimum cost of phosphate, by weight?
The product gives a lower bound on cost of carbon fixation
per unit weight when nutrients must be provided.

∂19-Jul-89  1920	cloutier@sierra.STANFORD.EDU 	Appointment    
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 19 Jul 89  19:19:55 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Wed, 19 Jul 89 19:19:57 PDT
From: cloutier@sierra.STANFORD.EDU (Mary Cloutier)
Date: Wed 19 Jul 89 19:19:56-PDT
Subject: Appointment
To: jmc@Sail
Cc: cloutier@SIERRA
Message-Id: <616904396.0.CLOUTIER@SIERRA>
Mail-System-Version: <SUN-MM(219)+TOPSLIB(128)@SIERRA>


Dr. McCarthy:

I have put you down for an appointment with Dr. Gibbons on Friday,
August 4, l0:00 a.m.  He has a 9:00 and an ll:00 so it will only be
for one hour.
 
See you in Terman 214 on Friday, the l0th.  Please confirm that this
is all right.

					Mary

-------

∂20-Jul-89  0601	cliff%unix.computer-science.manchester.ac.uk@NSFnet-Relay.AC.UK 	TESTING        
Received: from NSFnet-Relay.AC.UK by SAIL.Stanford.EDU with TCP; 20 Jul 89  06:00:56 PDT
Received: from unix.computer-science.manchester.ac.uk by NSFnet-Relay.AC.UK 
           via Janet with NIFTP  id aa04879; 20 Jul 89 13:31 BST
From: Cliff Jones <cliff%unix.computer-science.manchester.ac.uk@NSFnet-Relay.AC.UK>
Date: Thu, 20 Jul 89 11:53:44
To: JMC@sail.stanford.edu
In-reply-to: John McCarthy's message of 19 Jul 89  1055 PDT <nv9V7@SAIL.Stanford.EDU>
Subject: TESTING    

ACK

(no more e-mail from me before Aug 7th - off to France!)

cliff

∂20-Jul-89  1140	RWF 	re: Nuclear waste disposal    
[In reply to message rcvd 19-Jul-89 18:07-PT.]

Thanks for the explanation.  I guess that for the purpose of eventually 
shooting the sun, you rank the planets by the energy it takes to get there
and lose the tangential velocity component, so going via mars might
be correct.  Oh, you also have to be sure the radial component isn't
outward...no, I guess that doesn't hurt.  Might one use two planets
to get to the sun?

∂20-Jul-89  1447	pcohen@ai.sri.com 	speech acts
Received: from Warbucks.AI.SRI.COM ([128.18.3.1]) by SAIL.Stanford.EDU with TCP; 20 Jul 89  14:47:37 PDT
Received: from stinson.ai.sri.com by Warbucks.AI.SRI.COM with INTERNET ;
          Thu, 20 Jul 89 14:48:03 PDT
Received: by stinson.ai.sri.com (4.1/4.16)
	id AA08014 for jmc@sail.stanford.edu; Thu, 20 Jul 89 14:49:05 PDT
Date: Thu 20 Jul 89 14:49:03-PDT
From:     PCOHEN@Warbucks.AI.SRI.COM (Phil Cohen)
Subject: speech acts
To:       jmc@sail.stanford.edu
cc:       pcohen@Warbucks.AI.SRI.COM
Message-Id: <616974543.0.PCOHEN@STINSON.ARPA>
Mail-System-Version: <SUN-MM(229)+TOPSLIB(128)@STINSON.ARPA>
Reply-To: pcohen@Warbucks.AI.SRI.COM (Phil Cohen)

John: 

Ray gave me your recent paper, and invited me to your meeting
yesterday.  Unfortunately, I couldn't make it, but would enjoy
discussing speech acts, and your paper.  Would you be interested?

Also, I've got a recent paper providing definitions (and semantics)
for speech acts in terms of a theory of mental states (esp.
intention/commitment), and arguing why Searle's theories are 
scientifically uninteresting (they're taxonomic and stipulative, and
most everything of interest follows from a theory of rational interaction).
At any rate, if you're interested, I could send you some papers so you
can see where I'm coming from...

Best,

Phil Cohen
-------

∂20-Jul-89  1705	H.HARRY@Macbeth.Stanford.EDU 	Prof. McCarthy's third-world problem [was Re: When should I call JMC racist?]   
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Jul 89  17:05:37 PDT
Date: Thu 20 Jul 89 17:04:46-PDT
From: Harinder Singh <H.HARRY@Macbeth.Stanford.EDU>
Subject: Prof. McCarthy's third-world problem [was Re: When should I call JMC racist?]
To: kanakia@Polya.Stanford.EDU
cc: su-etc@shelby.stanford.edu, jmc@sail.Stanford.EDU
In-Reply-To: <10767@polya.Stanford.EDU>
Message-ID: <12511609699.77.H.HARRY@Macbeth.Stanford.EDU>


	What I find even more amusing is that Prof. McCarthy sees
third-world conspiracies every time a foreign student from a third-world 
country differs with him.

	In this most recent scrap he brought up my third-world origins
as if that were a material parameter! Had I responded to that, as he
kept asking me to, it would have sidetracked the discussion instantly.
``Mr. Singh's country of origin...anti-American foreign policy... blah,
blah...'' - none of which had anything to do with the Alice Walker
article.

	As the record will show, I'd hardly even expressed an opinion
on the Alice Walker piece; I was content simply to challenge JMC's
`sleight-of-hand' dismissal of the article with some questions of
my own.

	I'm glad I didn't let his life-expectancies-of-black-women-etc
dodge slide. It enabled us to see his true intent behind asking that
question, by his own subsequent admission.

	When will Prof. McCarthy stop stereotyping third-world students
and dragging that into every discussion in which they differ from him?
He has already had to apologize publicly for it a few months ago and
one hopes that the sincerity of that apology will manifest in his
words and actions at some point.

		Inder

-------

∂21-Jul-89  0154	Mailer 	Reply and award for Prof. McCarthy [was re: Prof. McCarthy's third-world problem] 
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Jul 89  01:54:22 PDT
Date: Fri 21 Jul 89 01:53:40-PDT
From: Harinder Singh <H.HARRY@Macbeth.Stanford.EDU>
Subject: Reply and award for Prof. McCarthy [was re: Prof. McCarthy's third-world problem] 
To: JMC@SAIL.Stanford.EDU
cc: su-etc@SAIL.Stanford.EDU
In-Reply-To: <6vcwd@SAIL.Stanford.EDU>
Message-ID: <12511705983.76.H.HARRY@Macbeth.Stanford.EDU>


	Prof. McCarthy objects to the perceived need to `smoke
him out.' The evidence is out there - he had already done his
`sidestep maneuver' and, by his own admission, was being deliberately
insensitive; his subsequent long explanation came in response to 
being smoked out. I can empathise with his sense of discomfort at
that outcome :-)

--->>>	If he wishes to be taken seriously in dialogue, Prof. McCarthy
will have to learn not to twist his questions into manipulative `dismissive'
rhetoric and such. As in this recent case, his questions often cannot be 
taken at face value; in fact I once got a rise out of him simply by
mimicking his style of ending a reply with an altogether irrelevant,
profound-sounding question. [As with the apology recollection, I will
not be surprised if his mileage varies - I could, however, dig these
things out of my archives if his denials are sufficiently vehement.]
If he does not wish to see energy dissipated in smoking him out when
it could be spent on the substantive issues in a civilized discussion,
he should create less need for it by being straightforward in how he
frames his questions.

	Meanwhile I think my deprived third-world-limited life will 
just have to muddle along, somehow, now that it may not have the benefit 
of Prof. McCarthy's erudition in replies. Oh well...

	Prof. McCarthy uses an interesting stunt in his recent 
message to shame those who might agree with him into making their
support public: he says he `fears' they are feeling `intimidated!'
Come on, Professor, give 'em a break!! You can't really expect them
to be dumb enough to fall for that, or do you? I thought you said
you didn't need their company.

	Prof. McCarthy proceeds to pat himself on the back for
exceptional bravery in that ``he can flout public opinion with
more safety'' than his cowardly, silent supporters.

	For displaying such valor I award Prof. McCarthy 3 points
for boobist solidarity and the Goebbels Medal for Excellence in
Disinformation [don't mention it, Professor - happy to return the
favor :-)]

	Happy times,

			Inder 

-------

∂21-Jul-89  0944	pcohen@ai.sri.com 	re: speech acts 
Received: from Warbucks.AI.SRI.COM ([128.18.3.1]) by SAIL.Stanford.EDU with TCP; 21 Jul 89  09:44:28 PDT
Received: from stinson.ai.sri.com by Warbucks.AI.SRI.COM with INTERNET ;
          Fri, 21 Jul 89 09:44:47 PDT
Received: by stinson.ai.sri.com (4.1/4.16)
	id AA08799 for JMC@SAIL.Stanford.EDU; Fri, 21 Jul 89 09:45:53 PDT
Date: Fri 21 Jul 89 09:45:50-PDT
From:     PCOHEN@Warbucks.AI.SRI.COM (Phil Cohen)
Subject: re: speech acts
To:       JMC@SAIL.Stanford.EDU
Message-Id: <617042750.0.PCOHEN@STINSON.ARPA>
In-Reply-To: <jvyqx@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(229)+TOPSLIB(128)@STINSON.ARPA>
Reply-To: pcohen@Warbucks.AI.SRI.COM (Phil Cohen)

OK, how about Thurs, at 1:30?  Meeting at Stanford would be fine.
At Margarget Jacks or Cordura?

Phil
-------

∂21-Jul-89  1429	RPG 	Scherlis  
To:   CLT, JMC    
He was here earlier this week and I gave him a briefing on
the work I am doing for CPL. He seemed happy enough but seemed to
want me to concentrate more on the environment part. What I showed
him was somewhere in between language and environment.

			-rpg-

∂21-Jul-89  1438	RPG 	re: Scherlis   
[In reply to message rcvd 21-Jul-89 14:33-PT.]

He stayed at my house and the purpose of the trip was to
visit Luckham, Green, and Lucid. So I briefed him informally.

I'm trying to revise the X3J13 specification for delivery to ISO
this week and next, but after that we should spend some time chatting.
My new results should interest you as well.

			-rpg-

∂21-Jul-89  1522	alex@jessica.Stanford.EDU 	Orals scheduling decision: Friday August 18th 10am-12noon 
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Jul 89  15:22:11 PDT
Received: by jessica.Stanford.EDU; Fri, 21 Jul 89 15:22:03 PDT
Date: Fri, 21 Jul 1989 15:22:00 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: clt@sail.Stanford.EDU, jmc@sail.Stanford.EDU, rwf@sail.Stanford.EDU,
        dill@amadeus.Stanford.EDU, horning@src.dec.com
Cc: alex@jessica.Stanford.EDU, hemenway@score.Stanford.EDU
Subject: Orals scheduling decision: Friday August 18th 10am-12noon 
Message-Id: <CMM.0.88.617062920.alex@Jessica.Stanford.EDU>

Hello,

Since everybody on my combined (reading-committee + orals-committee)(*) can
make it on FRIDAY 18th AUGUST in the morning, this date is now cast in stone.

I'm asking Sharon Hemenway to reserve a room from 10am to 12noon, because 
I estimate: 60' talk, 10' cookie eating, 50' questioning.

If you have any complaint/request/advice please let me know; if you don't,
please send me a 1-liner msg acknowledging this msg.

In terms of THESIS draft, I intend to give each of you a draft 7 days before
the exam.  If some of you find this time too short and would like to know
what my work is before then, then I can give you right now a techreport and
2 conference papers (one presented this past June, one currently in review) 
which together contain all the TECHNICAL work in my thesis, just not put 
together yet.

Thank you,
				Alex
---
(*) my reading committee only had 2 Academic Council members, therefore
David Dill (with whom I've worked in the past 6 months on my pipelines)
and who is very familiar with my research, has accepted to be the 3rd 
AC member on my orals committee.

∂21-Jul-89  1950	MRC@CAC.Washington.EDU 	re: Shockely's errors..   
Received: from hanna.cac.washington.edu by SAIL.Stanford.EDU with TCP; 21 Jul 89  19:48:56 PDT
Received: from tomobiki-cho.cac.washington.edu by hanna.cac.washington.edu
	(5.61/UW-NDC Revision: 2.1 ) id AA25616; Fri, 21 Jul 89 19:47:22 -0700
Date: Fri, 21 Jul 1989 19:29:33 PDT
From: Mark Crispin <MRC@CAC.Washington.EDU>
Sender: mrc@Tomobiki-Cho.CAC.Washington.EDU
Subject: re: Shockely's errors..
To: John McCarthy <JMC@sail.stanford.edu>
In-Reply-To: <FAu##@SAIL.Stanford.EDU>
Message-Id: <MailManager.617077773.1575.mrc@Tomobiki-Cho.CAC.Washington.EDU>

In <FAu##@SAIL.Stanford.EDU>, John McCarthy writes:
>What about a sterilization bonus of, say, $5,000 available to
>anyone, no need to take any kind of test?
>It would also get some pessimists, people who say it's a crime to
>bring children into this world.
>I would be particularly interested in its effect on declared
>pessimists.  I suspect many would reject it and go through some
>kind of ideological contortion in order to do it.

     I guess I'd be in the "pessimist" category; I used to think I didn't
want any children, now I find that I really don't care.  However, the issue
isn't likely to come up; my SO had one child and doesn't want any more, and is
now nearing the end of her child-bearing years so it will become a dead issue.
I might add, though, that $5K is less than my monthly take-home and so is not
much of a financial motivation.  The perceived advantages or disadvantages of
sterilizations would sway me more than a paltry $5K.

-------

∂22-Jul-89  0700	JMC  
Mike at 10

∂22-Jul-89  0903	RPG 	re: Lisp macros, theory thereof    
[In reply to message rcvd 21-Jul-89 15:05-PT.]

I think your best bet for a ``macros expert'' is one of these two guys:

	Alan Bawden at MIT (alan@ai.ai.mit.edu)
	Gene Kohlbecker (indiana university, ppn unknown)

However, I don't think you'll find any knowledge of what you propose.

			-rpg-

∂22-Jul-89  1519	Mailer 	re: IQ (USA vs Japan) 
To:   JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
From: Les Earnest <LES@SAIL.Stanford.EDU>

JMC says:
. That intelligence (apart from what has been learned) as measured
. by tests was substantially one-dimensional was a surprise,
. discovered maybe around 1910.  The point was that the different
. proposed factors were very highly correlated with each other.
. Someone came up with the g (for general) factor.

Yes, that must have been a surprise, considering that it is an obviously
false conclusion.  For example, rote memory abilities vary widely among
individuals and are only loosely coupled with IQ.  In fact, some idiot
savants have extraordinary memories.

Another mental ability that does not seem to be strongly correlated with
others is what might be called ``awareness'' -- the ability to preceive
and react quickly to events in the immediate environment.  Some people
are very much aware of what is going on around them, while others are
only loosely coupled to the world.  (I'm one of the latter).

	Les Earnest

∂22-Jul-89  1553	Mailer 	re: IQ (USA vs Japan) 
To:   JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
From: Les Earnest <LES@SAIL.Stanford.EDU>

JMC says:
. The examples you cite were known in 1900.  Perhaps you could figure
. out why they might not be relevant for the purposes for which IQ
. tests were designed.

Of course, how stupid of me!  The reason that other forms of intelligence
are not relevant to the purposes for which IQ tests were designed is that
there was a theory that was widely believed around 1900 that was sometimes
called the "Chain of Life."  It held that all of mankind and the "lower"
animals are ranked in a God given canonical order.  In order to fit this
scheme, you need a one dimensional model of human mental abilities.

. Anyway, as Oliver Cromwell wrote to the General Assembly of the Church
. of Scotland on August 3, 1650
. 
.    "I beeseech you, in the bowels of Christ, think it
.    possible you may be mistaken."

Sorry, but I seldom make decisions based on an examination of entrails.
Especially not those of a religious fanatic.

	Les Earnest

∂22-Jul-89  1644	gds@spam.istc.sri.com 	Re: IQ (USA vs Japan) 
Received: from milk7.istc.sri.com by SAIL.Stanford.EDU with TCP; 22 Jul 89  16:43:58 PDT
Received: by milk7.istc.sri.com (4.0/5.00)
	           id AA13128 for JMC@sail.stanford.edu; Sat, 22 Jul 89 16:45:22 PDT
Date: Sat, 22 Jul 89 16:45:22 PDT
From: Greg Skinner <gds@spam.istc.sri.com>
Message-Id: <8907222345.AA13128@milk7.istc.sri.com>
To: JMC@sail.stanford.edu
Subject: Re: IQ (USA vs Japan)
Newsgroups: su.etc
In-Reply-To: <RAyF3@SAIL.Stanford.EDU>
Organization: SRI International, Menlo Park CA

In article <RAyF3@SAIL.Stanford.EDU> you write:
>The selection processes are good enough so that if someone who
>is far below the normal cutoff for selection gets in for some
>reason, he usually has a very hard time and drops out.  This
>happens often enough for a variety of reasons.

What reasons?

Also, did you see my last posting to su.etc?  I was hoping you'd
respond to that.

--gregbo

∂22-Jul-89  1903	Mailer 	re: IQ (USA vs Japan) 
To:   JMC@SAIL.Stanford.EDU, su-etc@SAIL.Stanford.EDU
From: Les Earnest <LES@SAIL.Stanford.EDU>

JMC doesn't seem to want my help in explaining measures of intelligence.
In response to my claim that intelligence is multidimensional, he said:
. That intelligence (apart from what has been learned) as measured
. by tests was substantially one-dimensional was a surprise,
. discovered maybe around 1910.

When I offered examples of forms of intelligence that are not measured by
IQ tests, JMC didn't bother with a refutation.  Instead, he appealed to
authority, saying:
. The examples you cite were known in 1900.  Perhaps you could figure
. out why they might not be relevant for the purposes for which IQ
. tests were designed.

When I responded with a plausible explanation for the "error" in my
analysis in terms of the Chain of Life theory that was widely believed
in 1900, he said:
. the ability to find instant refutations of the results of
. scientists' work without reading it or any exposition of it and on the
. basis of presumptions about what all these individuals must have believed
. is a good start towards becoming a religious fanatic.

That is a nasty thing to say about an avowed antitheist.  Of course, I was
using exactly the kind of argument that John uses all the time -- ascribing
a hidden motive to those with whom I disagree.

Given that John seems to believe that intelligence is one dimensional,
I suggest that he be more direct in supporting his view instead of leaving
it as an exercise for the reader or appealing to ancient authorities.

	Les Earnest

∂22-Jul-89  2352	terman@Portia.stanford.edu 	Re: IQ 
Received: from Portia.stanford.edu by SAIL.Stanford.EDU with TCP; 22 Jul 89  23:52:27 PDT
Received:  by Portia.stanford.edu (5.61/25-eef) id AA20648; Sat, 22 Jul 89 23:49:49 -0700
Date: Sat, 22 Jul 89 23:49:49 -0700
From: Mutant for Hire <terman@Portia.stanford.edu>
Message-Id: <8907230649.AA20648@Portia.stanford.edu>
To: JMC@SAIL.STANFORD.EDU
Subject: Re: IQ
Newsgroups: su.etc
In-Reply-To: <fAd91@SAIL.Stanford.EDU>
Organization: Stanford University
Cc: 

In article <fAd91@SAIL.Stanford.EDU> you write:
>	After reading at least one book, someone could start a
>campaign to have his professorship posthumously withdrawn and
>have Stanford apologize for ever harboring such a person.
>
>	I would oppose such a campaign, because I think he was
>a great man and a great scientist.

And his decendants thank you for that support. To be honest, I haven't studied
much of what he did except in a vague general sense. I have read complaints
that his test was culturally biased, that is, unless you were white middle
class, some of the questions on the test were less understandable than to
people coming from outside that culture.


-- 
Martin Terman                        Disclaimer: She was my cousin! I swear it!
Mutant for Hire                    "Ever go to a wedding, and instead of saying
God in Trainin                   `Don't they make a lovely couple' you'd rather
net.mutant                        say `Sweet Jesus, don't let them have kids!'"

∂23-Jul-89  1407	Mailer 	re: Abortion (what else?)  
To:   JMC@SAIL.Stanford.EDU, SU-ETC@SAIL.Stanford.EDU
From: Dick Gabriel <RPG@SAIL.Stanford.EDU>

[In reply to message from JMC rcvd 22-Jul-89 21:58-PT.]

If the problem with abortion is that the fetus is innocent of crime,
why not allow the fetus to be found guilty by a court of assault with
a deadly weapon or attempted murder? Possibly trespassing would do,
since all the woman desires is eviction.

				-rpg-

∂24-Jul-89  0033	BEDIT@Score.Stanford.EDU 	Summary of June computer charges. 
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Jul 89  00:33:04 PDT
Date: Mon 24 Jul 89 00:15:33-PDT
From: Billing Editor <BEDIT@Score.Stanford.EDU>
Subject: Summary of June computer charges.
To: MCCARTHY@Score.Stanford.EDU
Message-ID: <12512474552.12.BEDIT@Score.Stanford.EDU>

Dear Mr. McCarthy,

Following is a summary of your computer charges for June.

Account     System   Billed    Pct      Cpu    Job   Disk  Print   Adj   Total

JMC         SAIL     2-DMA807  100   393.77  17.78 ***.**  17.61  5.00 2672.34
MCCARTHY    SCORE    2-DMA807  100      .00    .00  30.47    .00  5.00   35.47
jmc         LABREA   2-DMA807  100      .00    .00 105.90    .00  5.00  110.90

Total:                               393.77  17.78 ***.**  17.61 15.00 2818.71


University budget accounts billed above include the following. 

Account     Princip Inv      Title                      Comment             

2-DMA807    McCarthy         N00039-84-C-0211           Task 19, QLISP       


The preceding statement is a condensed version of the detailed summary sheet 
sent monthly to your department. 

Please verify each month that the proper university budget accounts are paying 
for your computer usage.  Please also check the list of account numbers below 
the numeric totals.  If the organizations/people associated with that account 
number should NOT be paying for your computer time, send mail to BEDIT@SCORE. 

Please direct questions/comments to BEDIT@SCORE. 
-------

∂24-Jul-89  0419	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	tough nombres 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 24 Jul 89  04:19:15 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 403583; Mon 24-Jul-89 06:35:21 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 103590; Mon 24-Jul-89 02:46:52 PDT
Date: Mon, 24 Jul 89 02:46 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: tough nombres
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890724094644.7.RWG@TSUNAMI.SPA.Symbolics.COM>

Responding to an email query, I guessed that the term "Mertens's
constant" denoted

   limit     sum   1/p   - ln ln x  ~~ .26152?
  x -> inf  primes
            p < x

based on Hardy & Wright, p 351.

Note the minor miracle that e is the only base of the
outer log for which the limit can be finite.

This constant is a bear to compute.  I don't trust anything
past the .26, although for x = prime(2↑n), n = 16, 17, 18,
it's respectively .26157, .2615291, and .2615271.

This leads to the question:  Which are the most difficult (and
thus mysterious) constants to evaluate?

Chaitin's Omega, of course, has to win for most definitions of
"difficult".  We could measure how many logs you'd need to be
able to give the first digit.  But what would we do about signs,
or the likes of e↑e↑...↑e↑Omega?

Salamin reminds me of the bogus cake-taker "Riemann's constant":

  sum  mu(n)/sqrt(n),
  n>0

where mu is the Moebius function:  0 if a square divides n,
(-1)↑# of prime factors(n) otherwise.

This series converges iff the zeta hypothesis!  However, if
it *does* converge, we can easily guess to what!  1/zeta(1/2)
 ~~ -.684765 .

Thus, a constant's "difficulty" is muddied by the possibility of
a more efficient algorithm for it.  (In the absence of a lower
complexity bound.)

A while back, I posted an even clearer case where we "know" sums,
yet can't prove convergence:

pi↑2 + sum tan pi↑2 k/(πk+1)k = 9.71628 = sum cot(k-1/2)/(k-1/2)(k+1/2),
       k#0                                 k

which depends on pi lacking an infinite progression of prematurely
huge continued fraction terms.  The tameness of my 17M guarantees
that, at some point, the two sums above will concur to 12M digits,
while failing to prove that a first digit even exists, let alone
is 9.

∂24-Jul-89  0517	wri!ilan@uunet.UU.NET 	Re: Riemann Hypothesis
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 24 Jul 89  05:17:29 PDT
Received: from wri.UUCP by uunet.uu.net (5.61/1.14) with UUCP 
	id AA24930; Mon, 24 Jul 89 08:17:59 -0400
From: wri!ilan@uunet.uu.net
Received: by WRI.com (3.2/SMI-3.0DEV3)
	id AA23096; Mon, 24 Jul 89 07:12:29 CDT
Message-Id: <8907241212.AA23096@WRI.com>
Date: Mon, 24 Jul 89 07:12:27 CDT
To: wri-tech@uunet.UU.NET
Subject: Re: Riemann Hypothesis
Cc: ilan@uunet.UU.NET, dek@sail.stanford.edu, rwf@sail.stanford.edu,
        jmc@sail.stanford.edu, gosper@uunet.UU.NET

Speaking of the RH, the following is not altogether correct:

>   Salamin reminds me of the bogus cake-taker "Riemann's constant":
>
>      sum  mu(n)/sqrt(n),
>       n>0

>   where mu is the Moebius function:  0 if a square divides n,
>   (-1)↑# of prime factors(n) otherwise.

>   This series converges iff the zeta hypothesis!  However, if
>   it *does* converge, we can easily guess to what!  1/zeta(1/2)
>    ~~ -.684765 .

Agreed that convergence implies RH. But the converse is clearly
unclear since I think that present knowledge is that RH implies
that M(x) = O(sqrt(x) log↑2(x)), where M(x) = sum_1↑x mu(n).

As a matter of fact, from the bounds of  Odlyzko and te Riele 

 - 1.01 >  liminf M(x)/x,   limsup M(x)/x > 1.06

it is clear that the convergence of the series will happen only if
there is some incredible cancellation (much stronger than RH). 
Actually, maybe one could prove that it doesn't converge.

∂24-Jul-89  0856	ortiz@spam.istc.sri.com 	Recomendation  
Received: from aai0 ([128.18.4.90]) by SAIL.Stanford.EDU with TCP; 24 Jul 89  08:55:54 PDT
Received: from localhost by aai0 (3.2/5.00)
	           id AA15579 for jmc@sail.stanford.edu; Mon, 24 Jul 89 08:57:38 PDT
Message-Id: <8907241557.AA15579@aai0>
To: jmc@sail.stanford.edu
Cc: ortiz@spam.istc.sri.com
Subject: Recomendation
Date: Mon, 24 Jul 89 08:57:37 PDT
From: Charles Ortiz <ortiz@spam.istc.sri.com>


Professor McCarthy,

Hi.  I was in your CS323 Nonmonotonic Reasoning class this last
winter.  I received an A in the class. I am applying to University of
Pennsylvania's PhD program in computer science (in AI) for this fall
and I was wondering if I could trouble you for a recomendation?  

Thanks.

Charlie Ortiz

∂24-Jul-89  1223	Mailer 	re: IQ (USA vs Japan) 
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Jul 89  12:23:28 PDT
Date: Mon 24 Jul 89 12:22:22-PDT
From: ABE DEANDA <A.ABIE@Macbeth.Stanford.EDU>
Subject: re: IQ (USA vs Japan) 
To: JMC@SAIL.Stanford.EDU
cc: dmr@CSLI.STANFORD.EDU, su-etc@SAIL.Stanford.EDU
In-Reply-To: <fAvE#@SAIL.Stanford.EDU>
Message-ID: <12512606866.88.A.ABIE@Macbeth.Stanford.EDU>

JMC writes:

>I haven't heard of anyone taking the position, "Current IQ tests are
>defective in the following way.  If you want to measure intelligence
>properly, you should do it as follows.  Our experiments show ..."

Part of the problem is in defining intelligence as a trait.  What is it
we are looking for?  With IQ tests, we run the risks of circular 
reasoning, ie. "Intelligence is scoring X on this test" and "Scoring X
on this test correlates with intelligence".  There is no meaningful
Turing Test for all people.

Abe
-------

∂24-Jul-89  1702	hoffman.pa@Xerox.COM 	interview    
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 24 Jul 89  17:02:15 PDT
Received: from Semillon.ms by ArpaGateway.ms ; 24 JUL 89 16:59:30 PDT
Date: 24 Jul 89 16:53 PDT
From: hoffman.pa@Xerox.COM
Subject: interview
To: jmc@sail.stanford.edu
cc: hoffman.pa@Xerox.COM
Message-ID: <890724-165930-1974@Xerox>

Sorry for the length of time this took.  Transcribing and editing took
longer than I expected.  The only real editing changes I made to your
answers were grammatical and "style."  For example, I would break an "and"
apart into two sentences, or rephrase it with the same content but less
words.  When I have  comment about your answer (e.g. that you might want to
expand on something), I have delimited it with $$$...$$$.  When I have a
new question which you may or may not want to answer, I put it in with a
%%%....%%%.  You may either have me come down to ask the questions in
person, if you want, and then I will transcribe them, or type them out, or
say that you don't think that they are interesting questions.  I have
included these delimiters so that you may skim to them and then read
seriously.

If you would prefer hard-copy or my raising each issue with you in person,
please let me know and we will set up a time.

			thanks

			reid

Here is the manuscript for your perusal.

The Manuscript:


*Parallel Distributed Processing has gained a lot of popularity recently,
and some people feel it will solve the problems which "traditional,"
symbolic-based approaches of Artificial Intelligence have run up against.
Could you please comment on PDP.

% I looked at the two volumes of the PDP bible and it seemed to me that it
shared many of the defects of the 1950's and 1960's neural pattern
recognition work.  Specifically, it concentrated on unary predicates.
Let's take, as one example, the first paper in the second volume of PDP by
Rumelhart, et al.  In this paper, a network learns to recognize or to
classify the rooms of a house into bedrooms, kitchens, etc. according to
whether or not they do or don't contain certain furniture.  This works for
single rooms.  Now, suppose we wanted to classify either a closet as a very
small room that is next to a bedroom and adjoins only a bedroom or a
butler's pantry as being a room between a kitchen and a dining room.  Now,
in order to do this, one would need is a characterization of both the rooms
of a whole house and the relations among the rooms (such as adjacency
relations.)  This would enormously multiply the number of inputs.  It seems
to me that, if they were to try to solve this problem, they might be able
do it but it would essentially throwing away the network that they had
already trained.  Now, some of the people in PDP recognize these problems.
They recognize the problems associated with symbolic structures (???) and I
haven't followed lately what they have been doing, nor have I heard the
claim that they had solved these problems.  Therefore, PDP currently
strikes me as a very limited approach.

$$$Did the transcription of the second to last sentence come out right???
Maybe you could expand on the point some, or clarify it?$$$

%%%  *Recently, in the Brain and Behavioral Sciences journal, Smolensky
argued or hypothesized that connectionist architecture or "design" was at a
subsymbolic level or a substrate to "symbolic" cognitive science or AI
approach. From this basis, one might argue that these characterizations
"arise" out of a large neural net (e.g. they need not be programmed.)
Could you comment on this a little? %%%

*Could you comment a little on the claim that PDP is the right way to
tackle perception or learning problems, and perhaps that developing a
hybrid machine between PDP approaches and traditional AI approaches would
surmount some of the problems faced by AI.

%  As to learning, there are many kinds of it and only one kind is the kind
where you gradually learn something by reinforcement.  There is a fair
amount of one-shot learning for which PDP seems quite irrelevant.  For
example, take NETTALK which learned to adjust 24000 weights and where the
network took into account a letter of English text and three letters on
each side of it and learned how that letter should be pronounced.  Now,
suppose one adds a statement in the current romanized Chinese that the
letter "q" is pronounced "ch" and the letter "x" is pronounced "sh".  Now,
a human can -- when being told this -- read aloud an English text with
Chinese words in it and he or she certainly doesn't do that by
instantaneously adjusting 24000 weights.  Therefore, it seems to me that,
even in regard to learning, PDP is a limited approach.

I don't have any specific ideas how it might be combined with symbolic AI
approaches, but certainly that's worth thinking about.  

*  The relationship between Artificial Intelligence might seem more
remotely related to Computer Science than other subdisciplines of Computer
Science insofar as Artificial Intelligence has a lot to do with thinking
about knowledge, problem-solving or learning which tends to be the subject
of Philosophers and Psychologists rather than Computer Scientists.  Could
you comment on the relationship of Artificial Intelligence to Computer
Science, Philosophy and Psychology?

%  When one takes Artificial Intelligence, one can consider two approaches
to it.  One is a biological approach and the other is a Computer Science
approach.  The biological approaches take advantage of the fact that
intelligence is realized by human beings--which are animals--and therefore
by studying humans either in their neurophysiology or in their psychology,
one might learn something about intelligence.  The computer science
approach says that what we really want to study both the world and
intellectual strategies that will enable a system, whether it be a
biological system or Artificial Intelligence system, to achieve goals in
the world.  It seems to me that most of the work in Artificial Intelligence
so far has followed the Computer Science approach.  Now, there has been a
worthwhile interaction with psychology, but to a large degree this
interaction has gone the other way.  Namely, the psychologists have gone in
for an information processing model.  In some sense, that's what they
should have been doing all along but they went into this blind alley of
behaviorism around 1910 and it wasn't till they started contemplating
computers that they got out of it.  However, it's not clear to me what was
the specific contribution of AI research to this.  With regard to
philosophy, AI shares some problems because if you want to make an
intelligent computer program -- for example, one for controlling a robot --
then it has some general view of the world.  In order to build it, you have
to understand something about knowledge: what the kinds of knowledge are,
what the kinds of objects are that it should take into account, etc.  These
problems are essentially the same ones that the philosophers have studied
in philosophy of mind, epistemology, and ontology.  Now, it seems to me
that there are some things that AI can get from Philosophy and that there
are also some things that Philosophy can get from AI because of AI
considers these problems in a more concrete way than philosophers have
done.

$$$ In this previous answer (due to the fact this interview is for IEEE),
you might want to comment on how Computer Science AI relates to computer
science.  It's up to you. $$$

%%% Would you give an example of what you mean by an intellectual strategy?
%%%
 
* There seems sometimes to be a dichotomy between engineering AI (Edward
Feigenbaum) and science AI (yourself).  Some claim that most of the science
is done and all that remains is to build a large enough expert system.
Others claim that much basic science remains, and that the claims of these
other technologists are premature.  Could you comment a little on this
dichotomy?

%  I am not sure that I know anybody who is continually taking the extreme
engineering viewpoint that you presented.  That may be a bit of a strawman.
I don't know anyone who claims that there is no science left to be done.  I
would cheerfully argue with such a person if there were.  But in having in
creating this strawman -- let us imagine this strawman for the moment --
and let's take an example with Mycin which is an expert system for
diagnosing bachterial infections of the blood.  It has a large collection
of rules which relate symptoms and the results of tests to diagnoses and
possible treatments.  Now, Mycin does not know about patients really.  You
cannot tell Mycin "I treated a patient yesterday according to your advice
and he died.  I have a patient with the same symptoms."  Mycin will simply
say "unrecognized response."  Mycin doesn't know that bacteria are
organisms that grow inside a person once they get in and that reproduce.
In fact, Mycin doesn't even know anything about processes occuring in time.
Nor can it be told these things.  If you wanted a system that had these
properties, then you would just have to scrap Mycin and start over again.
Now, not all expert systems are quite that limited, but the handling of
general information, the inference to general conclusions, are something
that the current class of expert systems doesn't do.  

$$$ Will Mycin say "unrecognized response" or "unrecognized input"? $$$

*  So, as some science clearly remains to be done, what do you think are
main hurdles for the successful construction of an Artificially Intelligent
machine?

%  Well, I have always thought-- that is always meaning since 1958 -- that
the main or the key problem for AI, is common sense knowledge and reasoning
where common sense knowledge means that we're not working within a fixed
model, but rather we're always free to take a new phenomena into account.
This is in contrast to some topic like Operations Research or any other
formal scientific theory where you first decide what phenomena are to be
taken into account and their relations and then work within the theory.
So, common sense knowledge and reasoning is not bound by that.  (As an
aside, maybe common sense isn't quite the best name for it since "common
sense"--the phrase common sense-- means different things to different
people, but I haven't got any better one right now.)  Furthermore,
commonsense reasoning does not correspond to solving differential
equations.  Nor, it turns out, does it correspond entirely to logical
deduction.  One of the important steps in developing common sense knowledge
and reasoning has been in the last ten years the development of formalized
nonmonotonic reasoning, which uses the tools of mathematical logic but goes
beyond previous mathematical logical systems in the reasoning that it does.

$$$ This last answer may have been a bit garbled on the tape, and I am not
quite sure what you would want to do with it.  You might consider expanding
what you mean by common sense reasoning a little (in a positive sense, as
opposed to what it is not).  I have always thought that your example that
math and other science textbooks are always surrounded by text indicates
that every formal system is tied up in a context or a framework of common
sense. $$$

* What do you think the most significant advances in AI in the last 10-20
years have been, and what do you think--given the current state of AI,
which you may want to elaborate on--will be the advances in the next 10-20
years.?

%  Well, I think the development of expert systems has been a significant
advance.  I didn't think that there would develop a technology on the basis
of the limited science which is currently available, but I think that a
fair amount is being learned from expert systems.  Now, in the strictly
scientific area, I think the development of formalized nonmonotonic
reasoning is key, together with its sort of outriggers (one might say logic
programming and truth maintenance systems).

* What do you think that the current state of AI is, what do you think that
the next 20 years might bring us?

%  That's pretty hard to say, because my view is that some fundamental
breakthroughs are required to reach human level intelligence rather than
simply continuing along a predicted domain.  I don't think for example that
formalized nonmonotonic reasoning was anything that anyone would have
identified as before the middle 70s as being something worth doing and I
suspect that what we have yet to do is also difficult to identify.  Still,
two things which I am presently pursuing, and for which I have good hopes,
is formalization of the notion of context and the formalization of systems
that are described in terms of speech acts.

%%% Could you explain the basic ideas motivating your work in these two
areas, both context and speech acts? %%%

* Could you comment a little on the famous physical symbol systems
hypothesis (by Allen Newell), which can be loosely interpreted as saying
that any machine which has a sufficiently powerful symbol manipulation
device has sufficient and perhaps necessary means for intelligent behavior.
Do you think this is born out by the work in AI?

%  Well, I'm not sure that you've got it right as to what the hypothesis
is, and I am not sure that I understand it precisely as -- in terms of what
Newell meant -- but I can state a version that I support and I suppose that
it agrees with what Newell said, that is that "any intelligent system, that
is that reaches human level intelligence, will be at least substantially
characterizable as a symbol manipulator."  

$$$ Would you like to expand this answer to its relation to AI, or is it a
rather ill-formed question requiring that it be dropped or at least
reformulation. $$$

* Has AI changed this view at all?  While you know much better than I, it
seems to me that it was strongly held that sufficiently powerful (and
realizable) symbol manipulation was the means by which one could implement
intelligence.  

%  Well, the mere idea of symbol manipulation is too weak to do that.
Everything that is done with computer can count as symbol manipulation.
Now, I don't think anybody has refuted that, but it the bare form which I
have described it, it seems like a truism.

$$$ This question's fate is tied with the previous one. $$$

*  Another hypothesis, the knowledge representation hypothesis (by Brian
Smith), claims that any mechanically intelligent device will need to have
two things: first, to have internal structures which we (as outside
observers) can interpret those structures as propositional knowledge about,
and second, that these structures will be causally related to the behavior
of the machine.  Do you think this is an accurate description of the
underlying principles in knowledge representation?  What do you think about
this?

% Well, as far as I understood what you said, this corresponds to something
Daniel Dennett in 1971 called the "intentional stance" (when Smith was but
a pup) and what I gave in my 1979 paper "Ascribing Mental Qualities to
Machines" and what Newell called "The Logic Level."  To put it more or less
in Dennett's terms, that the behavior of many machines, many people, and
other systems, can be best understood by ascribing to them knowledge,
beliefs, and intentions, and to use a phrase of Newell's, they can often be
understood according to a principle of rationality, namely it does what it
thinks will achieve its goals.

$$$ While interesting information, this doesn't seem to quite answer the
question.  When I asked it, I think that I was trying to get at a somewhat
philosophical question: most work in KR is motivated by this sort of clear
semantics where knowledge is represented through structures which can
easily discussed as propositions about the world (reminiscent of Fodor's
mentalese "language of thought").  Further, these structures tie directly
into behavior.  Do you think that this is a good approach?  Do you think
that knowledge really should be structured in this way in a computer?  $$$

* Let's return to a point which you made before in your distinction between
biological/brain/neurological approach to AI and computer science approach
to AI.  Do you think that there will be any helpful interaction between
these two approachs?

% Eventually, but it depends on some progress in neurological research or
possibly in genetics.  Namely, it seems to me that there can be a fairly
intense interaction as soon as the people taking the neuro-physiological
approach are able to say something about how facts and behaviors are
represented in the brain or for that matter in the genes.  Take some animal
that has a rather fixed, but elaborate behavior, like a honey-bee or a
dung-beetle.  Then, as it is fixed, this behavior must be represented
genetically in some way.  I believe that it would be of extreme interest to
AI to learn something about that and I think that the biological methods
are what is required to learn something about that.

*  What do you see as the current-- besides your own projects of
nonmonotonic reasoning and others-- most interesting research programs or
topics in AI?

% Well, I tend to be rather focused on my own projects so I wouldn't really
like to say that what's most interesting is what I happen to be doing.  Nor
would I feel that I have a full view of what other people are doing.  I
think some of the work in heuristics and search recently is pretty good.  I
was favorably impressed by some of the recent work in chess programs, not
merely the fact that they are using faster computers.

* This is a bit of an overgeneralization since researchers trained in
different methodologies move to other schools, but certain methodologies or
approaches to AI have been linked with certain schools.  For instance, the
use of logic at Stanford or problem-solving/cognitive modeling at CMU.
Would you comment on these differences of approach?

% Well, it would be unfortunate if everyone took the same approach.  And
even if it was the best of approach, a lot of different approaches need to
be explored.  

$$$ Would you like to expand this answer or would you prefer that I just
dropped this question? $$$

* You have said before that the idea behind AI is not to achieve human
intelligence, but rather machine intelligence.  Nevertheless, most of the
examples which you give as justifications or reasons for certain designs
involve people.  (For instance, you argue against NETTALK because a human
being does not instantly adjust 24,000 weights.  These sorts of examples
abound in AI.)  Do you think there's any contradiction here?

% Well, what it really means is that we don't have in mind a lot of
intellectual mechanisms which differ substantially from what we can observe
in ourselves.  I suspect that, when we reach human level intelligence, the
major ideas will seem obvious because they will seem to be things which are
apparent from observing ourselves but they will be things which took
someone very smart to formulate.  Let me add another thing.  We do actually
make use in AI of some facilities that we don't have as humans, larger
immediate memories and much faster computation and techniques like hash
tables.  So, I believe that we will find more of such things as well as
things that are based on observations of ourselves as humans.

* There are a number of critics of AI: Winograd and Dreyfus claims AI is
impossible, Searle claims that AI can at best only model the mind (but not
create one), and Weizenbaum claims that AI is immoral.  Do you have
anything that you would like to add to these points?

% Well, it seems to me that every one of the things that you mentioned is
motivated by philosophical views which I have got to consider obselete.  I
think that the phenomenology which motivates Dreyfus and also Winograd and
Flores is bad philosophy and will eventually disappear even when it
confines itself to philosophy rather than taking on AI.  I think that
Searle's ideas about intentionality are based on introspection and the
developments from ordinary language philosophy and ascribe too much
generality to these introspective observations.  One can specifically see
that in Searle's book on speech acts and thinking about speech acts for
learning language.  One wants some of the properties that Searle has
ascribed to Speech Acts but does not want others.  Looking at the ones
which one doesn't want, one doubts whether his observation that they are
characteristic of human Speech Acts is correct.  I am thinking particularly
about his characterization of promises.  Weizenbaum's attack on AI is
associated with a general trend of attack on technology which takes the
form of imagining how any particular scientific or technological advance
may turn out badly because people may misuse it.  Now, that's sort of
associated with the 1960s political radicalism and also environmentalism.
I don't see how they can assess the future when they assess the past so
badly.  

$$$ Should speach acts be capitalized? Would you like to expand the bit
about Searle's thoughts on promises? You might want to make the last two
sentences more continuous, or tell me to do so. $$$

TWO ADDITIONAL QUESTIONS:

%%% Recently, at Stanford, there was an attempt to limit access to certain
unix usenet electronic bulletin boards (rec.humor.funny) due to the racist
nature of some of the jokes posted on the bulletin board. You argued
against this action and through a number of people's efforts this was
changed.  Could you explain the issues at stake here?

%%% There are a number of related social concerns which have arisen with
the advent of computing technology.  Among these are concerns with the ease
of violating privacy rights with computers (such as access to banking
records or the FBI's now defunct scheme of using computers to track
suspicious people.)  Other concerns include the copyrighting of software or
even user interfaces (like the Apple Suit).  Do you have any comments on
any of these social issues?

%%% One of the deep philosophical issues which AI may need to deal with can
be found in what Philosophers call the mind/body problem.  Our mental
states, seemingly unlike physical phenomena, appear to have "intentions:"
they contain information about things.  Similarly, we seem to act on the
basis of that information.  Do you have any thoughts on this?

----- Hoffman.Pa

∂24-Jul-89  2228	CLT 	file 
Please get me a copy of the Springer bid as early as possible
tomorrow so I can talk to the lawyer and get Neil going on it.

Also I would appreciate it if you would find the missing
file folder.  It represents a lot of time and effort on
my part and I am not willing to redo that work.

∂25-Jul-89  0101	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Stutter-free strings    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 25 Jul 89  01:01:50 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 403996; Tue 25-Jul-89 04:02:36 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 103841; Mon 24-Jul-89 23:36:48 PDT
Date: Mon, 24 Jul 89 23:36 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Stutter-free strings
To: ACW@YUKON.SCRC.Symbolics.COM
cc: "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rem@score.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    Math-Fun@YUKON.SCRC.Symbolics.COM
In-Reply-To: <19890724182805.2.ACW@WHIMBREL.SCRC.Symbolics.COM>
Message-ID: <19890725063639.9.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Mon, 24 Jul 89 14:28 EDT
    From: Allan C. Wechsler <ACW@YUKON.SCRC.Symbolics.COM>

    A "stutter" in a string is a substring consisting of any sequence of
    letters repeated twice.  I'm sorry -- I know that is a murky
    explanation.  Maybe examples will clear it up.  The word "banana" has
    two stutters: "anan" and "nana".  The word "stutter" has one stutter:
    "tt".  The word "Mississippi" has four stutters: "ississ", "ss",
    "ssissi", and "pp".  The word "substring" has no stutters -- it is
    "stutter-free".

    The string "headache" is also stutter-free, even though "he" occurs
    twice.  A stutter's two halves have to be adjacent.

    If the alphabet only had two letters, every word longer than three
    letters would have to contain a stutter.  "XOX" is stutter-free, but we
    can't extend it: both "XOXX" and "XOXO" have stutters.

    If the alphabet had three letters, how long would the longest
    stutter-free word be?  It turns out to be fairly easy to grind out very
    long ones.  But is there an upper limit?

This was probably the biggest booboo in HAKMEM, where the problem is
posed in Item 39, and inadvertently solved in Item 122 ("The Parity
Number").

Let P(t) = x(t) + i y(t) be the standard Peano squarefilling map from
the unit interval into the unit square.  Then the only solutions to
P(t) = t+i are theta and its one's complement, where theta is the
"parity number", a binary fraction whose nth bit is the sum mod 2 of
the bits in n base 2.

  theta  =  1/2 - (1/4) Prod 1-2↑-2↑n
                        n>=0

         =  (Prod 1-2↑-2↑n) Sum  2↑-2↑n/(1-2↑-1)...(1-2↑-2↑(n-1))
             n>=0           n>=0

         =  (.6996966996696996...)
                                  16

if there are any hex kinks listening.  (For binary, change the 6s to 0s
and the 9s to 1s.)  The rules are fairly obvious for doubling or even
squaring the length of a power of 2 initial substring.

Now, 2 theta is sutter free radix 4!  Moreover, it remains so if you
identify 1 with 2!

    My brother-in-law claims to have proved that there is no upper limit.
    This may mean that he has an algorithm for generating stutter-free
    strings of arbitrary length from an alphabet of three letters.  Or his
    proof may be non-constructive -- I don't know, because I haven't seen it
    yet.

I'll bet he has a constructive productions scheme.

    My real question is this: does anyone on this list know where this
    problem comes from?  Where would I begin to look in the Mathematical
    Abstracts to see what other work has been done on this?  Combinatorics?
    Coding theory?

I plead scholarly delinquency.  Someone (Hal Abelson?) pointed me at an
article by Marsden Morse, where he solved this as a generalization of
draw-by-repetition in chess!  So it might have been a recreational math
journal.  Worse yet, someone from England sent me a generative solution,
which I have long since mislaid.

It would be interesting if the set of solutions has positive dimension
in the set of ternary reals.  (Which would maybe make the critical radix
e, for some generalization of this problem.)  If not, maybe all solutions
are mutual shifts.  I seem to recall finding some slack in the 2 theta
construction, but I don't recall if the free choices became exponentially
rarer, a la Penrose tiles.  In any case, the slack would provide uncountably
many solutions, unlike the shifts.  Maybe a Cantor set (without rationals).

∂25-Jul-89  0730	JMC  
Neil Springer (Springer Construction) 345-2471

∂25-Jul-89  0755	qphysics-owner@neat.cs.toronto.edu 	request  
Received: from neat.cs.toronto.edu by SAIL.Stanford.EDU with TCP; 25 Jul 89  07:55:19 PDT
Received: by neat.cs.toronto.edu id 10409; Tue, 25 Jul 89 10:53:01 EDT
Received: from nac.no by neat.cs.toronto.edu with SMTP id 10393; Tue, 25 Jul 89 10:52:40 EDT
Received: from [129.241.1.5] by nac.no (5.54/1.15) 
	id AA02267; Tue, 25 Jul 89 14:45:54 +0200
Received: by runix.runit.sintef.no (norunix.EARN) (1.2/6.1)
	id AA25068;  Tue, 25 Jul 89 14:45:42 +0200
Date:	Tue, 25 Jul 89 08:40:00 EDT
From:	Eva Grieg <mseeg%rose.uib.uninett@nac.no>
To:	qphysics@ai.toronto.edu
Message-Id: <316*mseeg@rose.uib.uninett>
Subject: request
Resent-From: qphysics-owner@cs.toronto.edu
Resent-To: qphysics-distribution@cs.toronto.edu
Resent-Reply-To: Eva Grieg <mseeg%rose.uib.uninett@nac.no>
Resent-Message-Id: <89Jul25.105301edt.10409@neat.cs.toronto.edu>
Resent-Date: Tue, 25 Jul 89 10:52:55 EDT

Request: qphysics
topic: members
topic: info
Request:end

∂25-Jul-89  0912	MPS  
John Voelcker, IEEE Spectrum - 212 705 7562
Needs to talk to you on the joke file story.

Pat

∂25-Jul-89  1035	hoffman.pa@Xerox.COM 	also    
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 25 Jul 89  10:35:12 PDT
Received: from Cabernet.ms by ArpaGateway.ms ; 25 JUL 89 10:25:59 PDT
Date: 25 Jul 89 10:19 PDT
From: hoffman.pa@Xerox.COM
Subject: also
To: jmc@sail.stanford.edu
cc: hoffman.pa@Xerox.COM
Message-ID: <890725-102559-3291@Xerox>

If you prefer to have nothing more to do with editing the interview, just
say so.  I just wanted the interview to have the greatest possible impact
with some good ideas.  I also wanted to make sure that I got it right.

thanks
reid


----- Hoffman.Pa

∂25-Jul-89  1038	VAL 	re: meeting on Elephant  
[In reply to message rcvd 24-Jul-89 20:02-PT.]

Thank you, I will come.

∂25-Jul-89  1158	MPS  
Stanley Brown, Physical Review Letters
516 924-5533

Pat

∂25-Jul-89  1400	JMC  
xerox  Lawrence check

∂25-Jul-89  1442	MPS 	Directions to Bell's
As you come from Palo Alto via Foothill Expwy passing Arastradero, Old
Oak Court is on the right as you come to the top of the hill, but
200 meters before you arrive at the stop light at Edith in Los Altos. 
450 is at the end of the cul de sac on this one block dead end street.

∂26-Jul-89  0055	VAL 	Counterfactuals

The assertion

	If the values of the frame fluents F1, F2 in the situation S
	had satisfied P(v1,v2), then Q(S) would have been true

is interpreted as

∀s . [P(val(F1,s),val(F2,s))∧∀f(Frame(f)∧f≠F1∧f≠F2⊃val(f,s)=val(f,S))] ⊃ Q(s).

There should be a metatheorem saying that if P(v1,v2) and Q(S) satisfy this
condition, and the axioms about the effect of some action A entail

	P(val(F1,result(A,S)),val(F2,result(A,S))),

and say nothing about the effect of A on fluents other than F1 and F2, then the
axioms plus the law of inertia entail

	Q(result(A,S)).

This metatheorem will explain how counterfactuals help us reason about the
effects of actions. For instance, imagine that I can prove that if the block
B1 hadn't been on top of the block B2 in the current situation, then B2 would
have been clear. Then, using the metatheorem, I can conclude that B2 will be
clear after the action move(B1,Table).

∂26-Jul-89  0800	JMC  
darpa

∂26-Jul-89  0829	MPS  
Hi,

I called Xerox again and this is what they said.
If it is battery operated, charge the battery and then
use it with the battery instead of plugging it into
the electricity.  The person I talked to just now, said he
was going to double check with the rep from Japan and will
call me back.  

Can you operate this thing with a battery?  That way I will
be able to have more info when he calls back.  Thanks.

Pat

∂26-Jul-89  1014	CLT 	gymboree  
This will be Timothy's first time for this session,
so you should enroll him for the session (give them money).

∂26-Jul-89  1203	VAL 	re: Counterfactuals 
[In reply to message rcvd 26-Jul-89 01:06-PT.]

It seems to me that your proposed replacement

	"If I can get rid of B1, then B2 will be clear"

only superficially looks like an ordinary conditional. Actually, this "if"
means "after", as in "if I move B1 onto the table". So an attempt to make
your replacement precise will give something like

  "After any action that changes the position of B1, B2 will be clear".

But this may not be true, because some available actions may change a
configuration of blocks in a global way (e.g., arrange all blocks in a
tower in a standard order). So I don't see any adequate replacement for
the counterfactual in this example.

In general terms: Reasoning about action may require facts about states close
to the current state, and the language of counterfactuals allows us to
conveniently describe such states by imposing conditions on the frame fluents
whose values may be different.

∂26-Jul-89  1207	BEASLEY@IBM.COM 	Alexander Gorbis  
Received: from IBM.COM by SAIL.Stanford.EDU with TCP; 26 Jul 89  12:06:54 PDT
Date: 26 Jul 89 10:44:01 PDT
From: "Mike R. Beasley"  <BEASLEY@ibm.com>
To:   jmc@sail.stanford.EDU
Message-Id: <072689.104403.beasley@ibm.com>
Subject: Alexander Gorbis

John, I am an IBM corporate recruiter for Computer Science, including
A.I., students at Stanford.   Recently Alexander Gorbis applied to
work at IBM.  He indicated that he is on leave from getting his
PhD in A.I. at Stanford and would like to continue that work part-
time in another year.  He stated that you are his advisor.

Can you tell me something about his capabilities?  Would you recommend
him?  Why is he taking a leave for a year?  What is his area of
specialization?

Thank you for any input you can provide to help us make a decision.

∂26-Jul-89  1306	ACW@ELEPHANT-BUTTE.SCRC.Symbolics.COM 	Stutter-free strings 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 26 Jul 89  13:06:32 PDT
Received: from YUKON.SCRC.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 404625; Wed 26-Jul-89 16:06:57 EDT
Received: from WHIMBREL.SCRC.Symbolics.COM by YUKON.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 476450; Wed 26-Jul-89 16:05:45 EDT
Date: Wed, 26 Jul 89 16:04 EDT
From: Allan C. Wechsler <ACW@YUKON.SCRC.Symbolics.COM>
Subject: Stutter-free strings
To: AndersE@DNA.LTH.SE
cc: "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rem@score.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    RWG@YUKON.SCRC.Symbolics.COM, ACW@YUKON.SCRC.Symbolics.COM, Martin_T._Wechsler%Wayne-MTS@UM.CC.UMich.Edu,
    Math-Fun@YUKON.SCRC.Symbolics.COM
In-Reply-To: <19890725063639.9.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890726200438.8.ACW@WHIMBREL.SCRC.Symbolics.COM>

Dear Anders (and all other CCs):

Here is a brief recent discussion on stutter-free strings.  Other
recipients should understand that Anders is my brother-in-law, and that
he missed the early part of this discussion.

    Date: Mon, 24 Jul 89 23:36 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

	Date: Mon, 24 Jul 89 14:28 EDT
	From: Allan C. Wechsler <ACW@YUKON.SCRC.Symbolics.COM>

	A "stutter" in a string is a substring consisting of any sequence of
	letters repeated twice.  I'm sorry -- I know that is a murky
	explanation.  Maybe examples will clear it up.  The word "banana" has
	two stutters: "anan" and "nana".  The word "stutter" has one stutter:
	"tt".  The word "Mississippi" has four stutters: "ississ", "ss",
	"ssissi", and "pp".  The word "substring" has no stutters -- it is
	"stutter-free".

	The string "headache" is also stutter-free, even though "he" occurs
	twice.  A stutter's two halves have to be adjacent.

	If the alphabet only had two letters, every word longer than three
	letters would have to contain a stutter.  "XOX" is stutter-free, but we
	can't extend it: both "XOXX" and "XOXO" have stutters.

	If the alphabet had three letters, how long would the longest
	stutter-free word be?  It turns out to be fairly easy to grind out very
	long ones.  But is there an upper limit?

    This was probably the biggest booboo in HAKMEM, where the problem is
    posed in Item 39, and inadvertently solved in Item 122 ("The Parity
    Number").

Anders: HAKMEM is an informal anthology of recreational math and
computer science, published by several hackers at the MIT AI lab in the
early 70's.

    Let P(t) = x(t) + i y(t) be the standard Peano squarefilling map from
    the unit interval into the unit square.  Then the only solutions to
    P(t) = t+i are theta and its one's complement, where theta is the
    "parity number", a binary fraction whose nth bit is the sum mod 2 of
    the bits in n base 2.

      theta  =  1/2 - (1/4) Prod 1-2↑-2↑n
			    n>=0

	     =  (Prod 1-2↑-2↑n) Sum  2↑-2↑n/(1-2↑-1)...(1-2↑-2↑(n-1))
		 n>=0           n>=0

	     =  (.6996966996696996...)
				      16

    if there are any hex kinks listening.  (For binary, change the 6s to 0s
    and the 9s to 1s.)  The rules are fairly obvious for doubling or even
    squaring the length of a power of 2 initial substring.

    Now, 2 theta is sutter free radix 4!  Moreover, it remains so if you
    identify 1 with 2!

	My brother-in-law claims to have proved that there is no upper limit.
	This may mean that he has an algorithm for generating stutter-free
	strings of arbitrary length from an alphabet of three letters.  Or his
	proof may be non-constructive -- I don't know, because I haven't seen it
	yet.

    I'll bet he has a constructive productions scheme.

Here is a cleaner one, derived from Gosper's.  Take the parity number in
binary, .01101001100101101001... and write another string under it.
Whenever two adjacent digits are different, write the first digit.
Wherever the two are the same, write a 2.  Thus:

01101001100101101001...
0210120212010210120....

	My real question is this: does anyone on this list know where this
	problem comes from?  Where would I begin to look in the Mathematical
	Abstracts to see what other work has been done on this?  Combinatorics?
	Coding theory?

    I plead scholarly delinquency.  Someone (Hal Abelson?) pointed me at an
    article by Marsden Morse, where he solved this as a generalization of
    draw-by-repetition in chess!  So it might have been a recreational math
    journal.  Worse yet, someone from England sent me a generative solution,
    which I have long since mislaid.

I found it.  The full reference is:

Morse, Marston.  Unending chess, symbolic dynamics, and a problem in
semigroups.  Duke Math. J. vol. 11 (1944) pp. 1-7.  

I didn't actually look at the article.  "Symbolic dynamics" is
apparently something that Morse got several people excited about back
then, and there was another reference that is probably topical:

--.  Symbolic dynamics II.  Sturmian trajectories.  Amer. J. Math. vol.
62 (1940) pp. 1-42.

There was no sign of "Symbolic dynamics I", probably because it predates
the first volume of Mathematical Reviews.
    
    It would be interesting if the set of solutions has positive dimension
    in the set of ternary reals.  (Which would maybe make the critical radix
    e, for some generalization of this problem.)  If not, maybe all solutions
    are mutual shifts.  I seem to recall finding some slack in the 2 theta
    construction, but I don't recall if the free choices became exponentially
    rarer, a la Penrose tiles.  In any case, the slack would provide uncountably
    many solutions, unlike the shifts.  Maybe a Cantor set (without rationals).

I don't know if it is related to this dimensionality woolgathering, but
here is an isomorphic problem in two letters.  Find an infinite string
of 1s and 2s, such that no string whose digits sum to zero mod 3 may be
stuttered.

∂26-Jul-89  1647	eaf@sumex-aim.stanford.edu 	Golub's nomination to NAE  
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 26 Jul 89  16:47:29 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA09950; Wed, 26 Jul 89 16:49:28 PDT
Date: Wed, 26 Jul 1989 16:49:25 PDT
From: Edward A. Feigenbaum <eaf@sumex-aim.stanford.edu>
To: dek@sail.stanford.edu, jmc@sail.stanford.edu, gibbons@sierra.stanford.edu,
        ullman@score.stanford.edu
Cc: quate@sierra.stanford.edu, phy@sail.stanford.edu
Subject: Golub's nomination to NAE
Message-Id: <CMM.0.88.617500165.eaf@sumex-aim.stanford.edu>

The NAE packet of nominee's names with member's comments is due at the NAE
office on Monday. I nominated Gene and his name appears on p.52.

I have reason to believe that his nomination is going to have rough sledding
at the peer committee level. I would deeply appreciate it (and so eventually
will Gene) if you would say supportive and laudatory things about Gene in the
little white box next to his name on p.52, making sure to get this to the NAE
office by Monday.

Thanks,

Ed

∂26-Jul-89  1654	sf@csli.Stanford.EDU 	Matyasevitch visit
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 26 Jul 89  16:54:12 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA19289; Wed, 26 Jul 89 16:55:09 PDT
Date: Wed 26 Jul 89 16:55:07-PDT
From: Sol Feferman <SF@CSLI.Stanford.EDU>
Subject: Matyasevitch visit
To: jmc@sail.stanford.edu, clt@sail.stanford.edu
Message-Id: <617500507.0.SF@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

John, I sent you a copy of a letter concerning the prospective visit of
Matyasevitch in September (if not delayed) under your and Pat's grant.
I should let Martin Davis know if there are any dates in September or
October which you would prefer he avoids when coming to Stanford.
I'll be away the first week in Sept and probably the 2nd week in Oct.
Pat will be away Sept 15-16. How about you and Carolyn?
                                  Sol
-------

∂26-Jul-89  2053	BEASLEY@IBM.COM 	Alex Gorbis  
Received: from IBM.COM by SAIL.Stanford.EDU with TCP; 26 Jul 89  20:52:55 PDT
Date: 26 Jul 89 16:20:07 PDT
From: "Mike R. Beasley"  <BEASLEY@ibm.com>
To:   jmc@sail.stanford.EDU
Message-Id: <072689.162010.beasley@ibm.com>
Subject: Alex Gorbis


Thank you for your quick response to my request.

∂27-Jul-89  1125	wolfe@Polya.Stanford.EDU 	questions...  
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 27 Jul 89  11:25:16 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA19275; Thu, 27 Jul 89 11:25:52 -0700
Date: Thu, 27 Jul 1989 11:25:47 PDT
From: "Michael R. Wolfe" <wolfe@Polya.Stanford.EDU>
To: jmc@sail, wolfe@Polya.Stanford.EDU
Subject: questions...
Message-Id: <CMM.0.88.617567147.wolfe@Polya.Stanford.EDU>

  I'm probably taking 306 in the autumn.  Could I ask you some very
brief questions?

  1) Are you still teaching it?  (The latest version of the schedule
says so, but I want to make sure).   If so:

  2) Is more of a programming or theory course?

  3) Is it very time consuming (eg large progamming projects, lots of
     problem sets?)

  4) How essential/useful is it for future AI work?

feel free to make these as brief as you like.

-Mike

∂27-Jul-89  1159	JMC  
pill

∂27-Jul-89  1205	ullman@nimbin.Stanford.EDU 	Research Prof.   
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 27 Jul 89  12:05:11 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA03136; Thu, 27 Jul 89 12:02:38 PDT
Date: Thu, 27 Jul 89 12:02:38 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8907271902.AA03136@nimbin.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: Research Prof.

There are a few points that I didn't get across, but I think need saying.

1. If a candidate can support himself/herself, and adds a reasonable
amount of lustre to the university, I don't have any trouble with
hiring that person as a prof. (research), even if they are not the
absolutely best in their field.  If they are to be appointed to the
regular faculty, then we really have to establish comparative quality.

2. You seem to depricate the approach theoreticians take to solving
problems, and since in theory, as in all fields, much of what
goes on is worthless, I often feel trapped into agreeing with you.
However, I think there is much of value in the precise and rigorous
style of investigation, and in particular, it is better to solve a
small problem, exhibit your techniques and the scope of what you
have done, publish them and move on, than to address grand issues
but on a level so vague that it is impossible to tell what it is
you have solved, if anything.

3. Just as you seem to have a charicature of theoreticians as
effete snobs worrying about whether an irrelevant problem is NP-complete,
I have a charicature of AI people as fuzzy thinkers, incapable of
achieving anything that can be written down, explained and used by
others.  There is probably some truth in both cases.
What should we do about it?
				---jdu

∂27-Jul-89  1555	sf@csli.Stanford.EDU 	re: Matyasevitch visit      
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 27 Jul 89  15:54:55 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA04816; Thu, 27 Jul 89 15:55:50 PDT
Date: Thu 27 Jul 89 15:55:48-PDT
From: Sol Feferman <SF@CSLI.Stanford.EDU>
Subject: re: Matyasevitch visit   
To: JMC@SAIL.Stanford.EDU
Message-Id: <617583348.0.SF@CSLI.Stanford.EDU>
In-Reply-To: <jC#zx@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

Fine, I'll communicate that.
-------

∂28-Jul-89  0834	CLT 	$    
We have 4k in the bank (after the Springer deposit).
From now til the escrow closes, please don't write any
checks without letting me know the amount.

You had better lean on Nafeh again.
You haven't gotten anything for June or July.

∂28-Jul-89  0900	JMC  
faculty club,schwartz

∂28-Jul-89  0913	VAL 	reply to message    
[In reply to message rcvd 27-Jul-89 17:51-PT.]

That will cost about a dollar, I think...

∂28-Jul-89  1400	hoffman.pa@Xerox.COM 	what we will do   
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 28 Jul 89  14:00:31 PDT
Received: from Cabernet.ms by ArpaGateway.ms ; 28 JUL 89 13:35:12 PDT
Date: 28 Jul 89 13:25 PDT
From: hoffman.pa@Xerox.COM
Subject: what we will do
To: jmc@sail.stanford.edu
cc: hoffman.pa@Xerox.COM
Message-ID: <890728-133512-3789@Xerox>

Professor McCarthy,

Could you please let me know what plan you would like to use vis-a-vis
transforming the interview into final form.  Should I just edit (reduce
it), or should I come in with a hardcopy, or should I call, or should I
wait?

thanks
reid


----- Hoffman.Pa

∂28-Jul-89  1402	GRAHAM%UMKCVAX2.BITNET@Forsythe.Stanford.EDU 	Concepts as objects ... 
Received: from Forsythe.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Jul 89  14:01:52 PDT
Received: by Forsythe.Stanford.EDU; Fri, 28 Jul 89 14:00:26 PDT
Date:     Fri, 28 Jul 89 16:03 CDT
From:     <GRAHAM%UMKCVAX2.BITNET@Forsythe.Stanford.EDU>
Subject:  Concepts as objects ...
To:       jmc@sail.stanford.edu
X-Original-To:  jmc@sail.stanford.edu, GRAHAM

Dr. McCarthy,
        This past summer, I've been teaching a course on knowledge
representation.  Part of the readings for the course were your articles:
        Epistemological Problems of Artificial Intelligence
and
        First Order Theories fo Individual Concepts and Propositions

        Your treatment of "concepts as objects" intrigued me, particularly
with regard to sentences such as "I thought that your yacht was longer than
it is."  I was wondering if you (or anyone else) have followed up on your ideas
and developed them further.  Could you suggest any references?  Does this area
still appear to be worth investigating?

Thank you for considering these requests.

Steven Graham
graham@umkcvax1.bitnet
UMKC Computer Science Telecommunications Program
4747 Troost Ave.
Kansas City, Missouri 64110

∂28-Jul-89  1509	Mailer@Score.Stanford.EDU 	Message of 25-Jul-89 14:58:01    
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Jul 89  15:09:09 PDT
Date: Fri 28 Jul 89 15:08:08-PDT
From: The Mailer Daemon <Mailer@Score.Stanford.EDU>
To: JMC@Sail.Stanford.EDU
Subject: Message of 25-Jul-89 14:58:01

Message undeliverable and dequeued after 3 days:
pat@IMSSS.#Pup: Cannot connect to host
	    ------------
Received: from csli.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Tue 25 Jul 89 14:58:02-PDT
Received: from SAIL.Stanford.EDU by csli.Stanford.EDU (4.0/inc-1.0)
	id AA00574; Tue, 25 Jul 89 14:59:22 PDT
Message-Id: <1NCsp0@SAIL.Stanford.EDU>
Date: 25 Jul 89  1458 PDT
From: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Matyasevich    
To: suppes@CSLI.STANFORD.EDU

I expect to be away Sept 25-27.  Otherwise, I'll be here.
Please acknowledge to verify that you do get email at this address.

-------

∂28-Jul-89  1557	VAL 	Design stance  

When we think of a system in terms of its input-output relation,
is this an instance of the design stance, or is it something different
(the "black box stance")?

∂28-Jul-89  1612	VAL 	re: Design stance   
[In reply to message rcvd 28-Jul-89 16:02-PT.]

The "black box stance" is my invention--in case we want to distinguish it from
the design stance.

On another subject: you write on p. 185 of the book I'm proofreading that we
don't want a computer that loses its temper. What about the behavior of the
operating system or an ATM when you give a wrong password 3 times in a row?
An ATM will actually take your card away! It seems to me that the useful role
of this mechanism is the same as in humans. A compiler that stops printing error
messages when there are too many of them is in the same category.

∂29-Jul-89  0147	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	pi, yi, yi.   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 29 Jul 89  01:47:44 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 405495; Sat 29-Jul-89 04:47:10 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104483; Sat 29-Jul-89 01:40:17 PDT
Date: Sat, 29 Jul 89 01:39 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: pi, yi, yi.
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890729083952.4.RWG@TSUNAMI.SPA.Symbolics.COM>

The high-water marks in pi's first 17001303 partial quotients are

       n          a_n  [k n ln n]

       1st         3         0
       2nd         7         6
       3rd        15        15
       5th       292        37
     308th       436      8291
     432nd     20776     12316
   28422nd     78629   1369336
  156382nd    179136   8787051
  267314th    528210  15693583
  453294th  12996958  27736867
11504931st 878783625 878783625

with k (= 4.698) fudged to fit the last term.  If I got it straight,
Gregory Chudnovsky told me that the least monotone function majorizing
the cf sequence, when reciprocated and summed over the positive integers,
converges or diverges as a sensitive measure of pi's (or whosever)
"degree of irrationality".

Note that n ln n is convex (supralinear), while you might expect a
"random" integer sequence to spike up to new highs more timidly than
linear.  And even the cfs for numbers like e and e↑2 increase just
linearly, their continued fractions converge qualitatively faster
than pi's, because new highs arrive with nondecreasing frequency.

Gregory also told me a bit more about how they managed to eliminate the
gcds from their matrix elements.  They apparently have some fancy
theoretical congruences.  But just fooling around, I have noticed that
it is possible to *exactly characterize* the prime decomposition of three
of the four matrix elements, after all four elements (one is 0) have been
fully reduced to lowest terms!  Thus, you only need to store one of the
elements (b), and it is mucho smaller, since you can exactly characterize
how far it can reduce!  I should have gotten 69M instead of 17.5!!

The key observation is that, at least in these Ramanujanoid series, the
reduction by g := gcd(a1,d2) in

      a1/g  b1   a2   b2
     (        ) (        )
       0    d1    0  d2/g

homogeneously scales the product, and is sufficient to fully reduce
any product starting from term 0.  (Some gubbish builds up in products
over non-initial intervals, but disappears when merged into the initial
one.)  But the unreduced a and d elements are expressible in terms of
powers and factorials, whose prime decompositions (hence gcd) have
effectively linear cost!  Damn!  This should put the complexity crossover
for Salamin's iterated mean out near neutron-star machines.  If Kanada
(using means) cracks a billion before the Chudnovskys (using series), it
is testimony to heroic programming and staggering Hitachi "Crayola"
power.

∂29-Jul-89  1338	weening@gang-of-four.Stanford.EDU 	Forsythe down Saturday   
Received: from gang-of-four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Jul 89  13:38:40 PDT
Received: from localhost by gang-of-four.Stanford.EDU (5.0/25-eef) id AA00842; Sat, 29 Jul 89 13:39:34 PDT
Resent-Message-Id: <8907292039.AA00842@gang-of-four.Stanford.EDU>
Return-Path: <@SAIL.Stanford.EDU:roberts@jessica.Stanford.EDU>
Received: from SAIL.Stanford.EDU by gang-of-four.Stanford.EDU (5.0/25-eef) id AA00710; Sat, 29 Jul 89 12:57:51 PDT
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Jul 89  12:56:35 PDT
Received: by jessica.Stanford.EDU; Sat, 29 Jul 89 12:56:01 PDT
Date: Sat, 29 Jul 1989 12:55:59 PDT
From: Ron Roberts <roberts@jessica.stanford.edu>
To: nethax@jessica.Stanford.EDU
Subject: Forsythe down Saturday
Message-Id: <CMM.0.88.617745359.roberts@Jessica.Stanford.EDU>
Resent-To: jmc@sail
Resent-Date: Sat, 29 Jul 89 13:39:33 PDT
Resent-From: Joe Weening <weening@gang-of-four.Stanford.EDU>

Connectivity to Forsythe and Lindy is down due to a power shutdown in
Forsythe Hall.  The main building power feed has been tripping its 1200A
breaker, so they have some electrical types here attempting to diagnose
the problem.  Power is expected to be down until 1400, possibly longer
depending on the problem(s) found.....					/Ron

∂29-Jul-89  1500	wri!ilan@uunet.UU.NET 	Re: Sum of MoebiusMu[n] / Sqrt[n]    
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 29 Jul 89  14:59:41 PDT
Received: from wri.UUCP by uunet.uu.net (5.61/1.14) with UUCP 
	id AA00298; Sat, 29 Jul 89 17:59:41 -0400
From: wri!ilan@uunet.uu.net
Received: by WRI.com (3.2/SMI-3.0DEV3)
	id AA07211; Sat, 29 Jul 89 13:56:38 CDT
Message-Id: <8907291856.AA07211@WRI.com>
Date: Sat, 29 Jul 89 13:56:37 CDT
To: gosper@uunet.UU.NET
Cc: jmc@sail.stanford.edu, rwf@sail.stanford.edu, dek@sail.stanford.edu,
        dbailey@ew11.nas.nasa.gov, wri-tech@uunet.UU.NET
In-Reply-To: Bill Gosper's message of Sat, 29 Jul 89 01:39 PDT <19890729083952.4.RWG@TSUNAMI.SPA.Symbolics.COM>
Subject: Re: Sum of MoebiusMu[n] / Sqrt[n]

By using partial summation on the tails of Sum MoebiusMu[n]/Sqrt[n] it
follows from the classical fact that the Sum from 1 to x of
MoebiusMu[n] is not o(Sqrt[x]) (see Titchmarsh, Theory of the Riemann
Zeta Function, or the lower bound of Odlyzko and te Riele) that the
infinite sum does not converge.

∂30-Jul-89  0325	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	russian intransigence [was Pi]    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 30 Jul 89  03:24:54 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 405605; Sun 30-Jul-89 06:26:13 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104589; Sun 30-Jul-89 02:44:13 PDT
Date: Sun, 30 Jul 89 02:43 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: russian intransigence [was Pi]
To: "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: esc@RUSSIAN.SPA.Symbolics.COM, cgay@RUSSIAN.SPA.Symbolics.COM
In-Reply-To: <8907291309.AA15360@ew11.nas.nasa.gov>
Message-ID: <19890730094357.6.RWG@TSUNAMI.SPA.Symbolics.COM>
bcc: "tk@ai.ai.mit.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM

    Date: Sat, 29 Jul 89 06:09:12 -0700
    From: dbailey@ew11.nas.nasa.gov (David Bailey)

     . . .

    I note that e-mail from you always has the return address
    rwg@russian.spa.symbolics.com.

That should work, but . . .
				    However, whenever I try to reply to
    your messages from this address, the mail is bounced.

JMC@Stanford and TK@mit complain similarly.
							   Also, if I try
    to send directly to this address, no luck.  Even a ping to this
    address, although our name server is able to resolve it, shows all
    packets lost.  When I mail to you at yukon.scrc.symbolics.com, though,
    it usually works just fine.

How about just rwg@symbolics.com?  (Or maybe gosper@symbolics.com?)

				 What is so peculiar about the russian
    system?

Nothing, in theory, although it has been occasionally threatened with
excommunication for running obsolete namespace data.

	     Is it only turned on part of the time?

It's up a good 98%.  A bleep of a lot better than Symbolics's
East/West link, which goes out many nights and weekends.

The problem is probably in our namespace.  When I went on no fee
status, they reamed me off of most databases and lists, then
piecemeal restored me, then reamed me a little more, then restored
me a little more.  I'll ask some other people who get their mail
through russian if they have similarly frustrated correspondents.

Have you ever responded to mail from Wolfram research?  I never
succeeded with the supplied uunet or University of Illinois return
addresses, but win every time with wri.com.

∂30-Jul-89  0649	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	hot series for ln2 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 30 Jul 89  06:49:13 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 405612; Sun 30-Jul-89 09:46:52 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104612; Sun 30-Jul-89 05:07:52 PDT
Date: Sun, 30 Jul 89 05:07 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: hot series for ln2
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890730120736.8.RWG@TSUNAMI.SPA.Symbolics.COM>

              [          1                    25 ]
         11   [     (K + -) (K + 1)    17 K + -- ]
        /===\ [          2                    2  ]
         ! !  [ - -------------------  --------- ]
(D1012)  ! !  [            5       7      18     ] = 
         ! !  [   288 (K + -) (K + -)            ]
        K = 0 [            4       4             ]
              [                                  ]
              [           0                1     ]


             [                1                  3944416672720412374012269665852437 ]
             [ --------------------------------  ---------------------------------- ]
             [ 31075098250364876952254571724800  5690590372933484324170961001185280 ]
             [                                                                      ]
             [                0                                  1                  ]

(C1013) BFLOAT(EXP(RHS(%)[1,2]));
Time= 584 msecs

(D1013)                  1.99999999999999999999999999999923B0

I.e., > 8 bits/term.  Also note unit numerator in u.l. element,
corresponding to next "term" of series.  (Actual term is also
multiplied by (17k+25/2)/18.)

Stoopid pFq notation would make this a 4F3, but it is really
a fractionally shifted 3F2, and morally a 2F1.

In Stigma notation,

	 inf
	 ====
	 \			   34 K + 25
(D1020)   >    -------------------------------------------------- = 6 LOG(2)
	 /				                        K
	 ====  (4 K + 3) (4 K + 2) BINOMIAL(4 K + 1, 2 K) (- 18)
	 K = 0

which, I must admit, makes obvious the reciprocal integer in the matrix.

∂31-Jul-89  0529	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	ln(2), pi
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 31 Jul 89  05:28:54 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 405689; Mon 31-Jul-89 08:25:53 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104767; Mon 31-Jul-89 04:12:10 PDT
Date: Mon, 31 Jul 89 04:11 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: ln(2), pi
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "tk@ai.ai.mit.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890731111150.1.RWG@TSUNAMI.SPA.Symbolics.COM>

More generally

        LN(Y)
(D1104) ----- = 
        Y - 1

              [          1                 4                                         ]
         inf  [     (K + -) (K + 1) (Y - 1)            2                2            ]
        /===\ [          2                       K (2 Y  + 12 Y + 2) + Y  + 10 Y + 1 ]
         ! !  [ - -----------------------------  ----------------------------------- ]
         ! !  [           5       7           2              6 Y (Y + 1)             ]
         ! !  [   16 (K + -) (K + -) Y (Y + 1)                                       ]
        K = 0 [           4       4                                                  ]
              [                                                                      ]
              [                0                                  1                  ]

(Notice the palindrome polynomials in the u.r., necessary because you expect essentially
 the same matrix for ln(1/y).)

So you can get arbitrarily rapid *pairs* of series for logs of rationals, because, e.g.,

   ln 2  =  - 2 ln 3/4 - ln 8/9,

and the series matrix for ln 3/4 is

                      [          1                            ]
                      [     (K + -) (K + 1)                   ]
                      [          2                194 K + 145 ]
                      [ - --------------------  - ----------- ]
(D1105)               [             5       7         504     ]
                      [   9408 (K + -) (K + -)                ]
                      [             4       4                 ]
                      [                                       ]
                      [           0                   1       ]

(~ 4 digits/term) and for 8/9

                     [           1                              ]
                     [      (K + -) (K + 1)                     ]
                     [           2                 1154 K + 865 ]
                     [ - ----------------------  - ------------ ]
(D1106)              [               5       7         7344     ]
                     [   332928 (K + -) (K + -)                 ]
                     [               4       4                  ]
                     [                                          ]
                     [            0                    1        ]

(~ 5.5 digits/term).  (3/4 and 8/9 being approximations to 1, where convergence
is instantaneous.)  For ln 243/256

              [                 1                                      ]
              [      28561 (K + -) (K + 1)                             ]
              [                 2                 12943658 K + 9706645 ]
              [ - ----------------------------  - -------------------- ]
(D1107)       [                     5       7          186250752       ]
              [   247837667328 (K + -) (K + -)                         ]
              [                     4       4                          ]
              [                                                        ]
              [               0                           1            ]

and now we begin to see a problem.  Arbitrarily rapid neq arbitrarily
efficient!  Since 84/53 is an approximant to (ln 3)/(ln 2) = 1 1 1 2 2 3 1 5 2 23 . . .

(C1108) 3↑53/2↑84;

                              19383245667680019896796723
(D1108)                       --------------------------
                              19342813113834066795298816

but clearly, 1 - % would be "too cumbrous to be of any importance".

The same series acceleration applies to arctan:

                        [            1           4                                  ]
                   inf  [       (K + -) (K + 1) Z                2           2      ]
                  /===\ [            2                  Z (K (2 Z  + 4) + 2 Z  + 3) ]
                   ! !  [ - --------------------------  --------------------------- ]
(D1109) ATAN(Z) =  ! !  [          5       7    2                   2               ]
                   ! !  [   4 (K + -) (K + -) (Z  + 1)          3 (Z  + 1)          ]
                  K = 0 [          4       4                                        ]
                        [                                                           ]
                        [              0                             1              ]

Remembering Machin's (MAYchins or mahSHaNs?) formula, pi/4 = 4 atan 1/5 - atan 1/239,
the matrices are

                        [          1                         ]
                        [     (K + -) (K + 1)                ]
                        [          2              102 K + 77 ]
                        [ - --------------------  ---------- ]
(D1110)                 [             5       7      390     ]
                        [   2600 (K + -) (K + -)             ]
                        [             4       4              ]
                        [                                    ]
                        [           0                 1      ]

(~ 3.4 digits/term), and 

                 [              1                                   ]
                 [         (K + -) (K + 1)                          ]
                 [              2                 228486 K + 171365 ]
                 [ - ---------------------------  ----------------- ]
(D1111)          [                    5       7       40956474      ]
                 [   13051463048 (K + -) (K + -)                    ]
                 [                    4       4                     ]
                 [                                                  ]
                 [               0                        1         ]

(zowie).

The Ramanujan series I used for the 17.5M turned ~8 digits/term, but
was >1.5 times hairier, so the atan 1/5 should be regarded as > 5.1
digits/term, when compared with cubic term ratios.  The 1/239
series roughly corresponds in difficulty to the sqrt 2 required by
Ramanujan, so he still beats souped up Machin, but not badly,
especially given the unit in the u.l. partial Machin products.

On the other hand, the Chudnovsky series about doubles Ramanujan.
Gleep.

∂31-Jul-89  0906	cross@vax.darpa.mil 	DARPA Project Summary   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 31 Jul 89  09:06:04 PDT
Received: from sun46.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA15296>; Mon, 31 Jul 89 11:57:05 -0400
Posted-Date: Mon 31 Jul 89 12:06:56-EDT
Received: by sun46.darpa.mil (4.1/5.51)
	id AA06569; Mon, 31 Jul 89 12:06:57 EDT
Date: Mon 31 Jul 89 12:06:56-EDT
From: Steve Cross <CROSS@DARPA.MIL>
Subject: DARPA Project Summary
To: englemore@sumex-aim.stanford.edu, lesser@cs.umass.edu,
        mcgregor@vaxa.isi.edu, rolf%lockheed.com@relay.cs.net,
        msf@isl1.ri.cmu.edu, jmc@sail.stanford.edu, mcdermott-drew@yale.edu,
        gjs@ai.ai.mit.edu
Cc: cross@vax.darpa.mil
Message-Id: <617904416.0.CROSS@SUN46.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@SUN46.DARPA.MIL>

During the next two weeks this office will engage in its annual budget 
blood bath.  The only ammunition I will have will be the project summary 
you indicated you would sned me by Jul 28th. I need your help and would 
appreciate your immediate support. Thanks in advance. Steve Cross
-------

∂31-Jul-89  1000	JMC  
Shankar

∂31-Jul-89  1011	cloutier@sierra.STANFORD.EDU 	Rescheduling of meeting  
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 31 Jul 89  10:11:43 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Mon, 31 Jul 89 10:11:07 PDT
From: cloutier@sierra.STANFORD.EDU (Mary Cloutier)
Date: Mon 31 Jul 89 10:11:05-PDT
Subject: Rescheduling of meeting
To: jmc@Sail
Cc: cloutier@SIERRA
Message-Id: <617908265.0.CLOUTIER@SIERRA>
Mail-System-Version: <SUN-MM(219)+TOPSLIB(128)@SIERRA>



Dr. McCarthy:

Unfortunately I am having to reschedule your meeting with Dr. Gibbons
on Friday, August 4 at l0:00 a.m.  Dr. Gibbons will be out of town
most of the following week so we will have to look at the week of
the l4th.  How about l0:00 a.m. on Wednesday, the l6th, 9:00 a.m. on
the l7th or if neither of these times will work, I might be able to
clear some time on  Friday morning, August l8th.  Please let me know
which date you would prefer.

Sorry for the cancellation.

					Mary

-------

∂31-Jul-89  1053	CLT 	DARPA Project Summary    
I can send it now with guesses or I can wait til Wednesday
and make some more realistic projections as to when the next
increment of money is due.

∂31-Jul-89  1137	CLT 	DARPA Project Summary    
Vladimir did that (or at least said he was going to)

∂31-Jul-89  1408	hoffman.pa@Xerox.COM 	sudden deadline   
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 31 Jul 89  14:08:39 PDT
Received: from Semillon.ms by ArpaGateway.ms ; 31 JUL 89 14:09:15 PDT
Date: 31 Jul 89 13:56 PDT
From: hoffman.pa@Xerox.COM
Subject: sudden deadline
To: jmc@sail.stanford.edu
Message-ID: <890731-140915-8012@Xerox>

When I arrived at work this morning, I was handed a sudden deadline of work
to be done by tomorrow noon so I couldn't call you today.  I will call
tomorrow morning.  

sorry 
reid


----- Hoffman.Pa

∂31-Jul-89 1418 GOODMAN%uamis@rvax.ccit.arizona.edu rumor has it
that these folks are developing an argument against our Received:
from rvax.ccit.arizona.edu by SAIL.Stanford.EDU with TCP; 31 Jul
89 14:18:21 PDT Date: Fri, 28 Jul 89 08:52 MST From:
GOODMAN%uamis@rvax.ccit.arizona.edu Subject: rumor has it that
these folks are developing an argument against our position of
commodity-determined decontrol. will keep you posted.  To:
DUANE.ADAMS@C.CS.CMU.EDU, MBLUMENT@NAS.BITNET,
DONGARRA@ANL-MCS.ARPA, GANNON%RDVAX.DEC@DECWRL.DEC.COM,
JAHIR@ATHENA.MIT.EDU, HEARN@RAND-UNIX.ARPA,
JLH@SIERRA.STANFORD.EDU, JMC@SAIL.STANFORD.EDU,
MCHENRY@GUVAX.BITNET, CWEISSMAN@DOCKMASTER.ARPA, TROYWIL@IBM.COM,
RALSTON@MCC.COM X-VMS-To: @NAS

	From:  Jnet%"MBLUMENT@NAS" "Marjory Blumenthal"
28-JUL-1989 08:32:47.32 To:  <goodman@arizmis> CC:  Subj:
commodity arguments

	Received: From NASVM(MAILER) by MRVAX with Jnet id 9781
for GOODMAN@ARIZMIS; Fri, 28 Jul 89 08:32 MST Received: by NASVM
(Mailer R2.03B) id 9780; Fri, 28 Jul 89 10:32:55 EDT Received: by
NAS (PC Mail Gateway) id 9577; Fri, 28 Jul 89 10:32:54 EDT Date:
Fri, 28 Jul 89 10:05 EDT From:  Marjory Blumenthal <MBLUMENT@NAS>
Subject:  commodity arguments To:  <goodman@arizmis> *To: Tom
Snitch cc: BITNET at NRC/BITNET *cc: Mitchel Wallerstein *cc: Meg
Knemeyer Tom, thanks for the fast turnaround.  I hope your launch
is going well.  --Marjory (cc: Sy Goodman) ========== Here is
your answer- "The Center for Security Policy is a not-for-profit,
non-partisan educational corporation established in 1988. It
exists to assist those in the executive and legislative branches
of The United States government, in the press and among the
general public concerned with foreign and defense issues in
promptly understanding and addressing strategically significant
security developments".  Text from a recently released study on
alternative approaches for a national security review.  Director
of the center is Frank Gaffney.  Board of Directors includes;
Elliot Abrams Steve Bryen Tom Clancey Colin Gray Jim Hackett Fred
Ikle Richard Perle and the rest of the gang.  Tom 0950 28 July

∂31-Jul-89  1612	CLT 	summer pay
Are you getting it?
There seems to some confusion on the second floor.

∂31-Jul-89  1709	cloutier@sierra.STANFORD.EDU 	re: Rescheduling of meeting   
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 31 Jul 89  17:09:40 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Mon, 31 Jul 89 17:09:06 PDT
From: cloutier@sierra.STANFORD.EDU (Mary Cloutier)
Date: Mon 31 Jul 89 17:09:01-PDT
Subject: re: Rescheduling of meeting   
To: JMC@SAIL.Stanford.EDU
Message-Id: <617933341.0.CLOUTIER@SIERRA>
In-Reply-To: <1pF9TH@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(219)+TOPSLIB(128)@SIERRA>

Dr. McCarthy:

I have you down on the l6th at l0:00 a.m.  Thanks for changing.					Mary

-------

∂01-Aug-89  0336	J.JBRENNER@Macbeth.Stanford.EDU 	re: Higgins and US policy against terrorism    
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Aug 89  03:36:39 PDT
Date: Tue 1 Aug 89 03:35:07-PDT
From: Joe Brenner <J.JBRENNER@Macbeth.Stanford.EDU>
Subject: re: Higgins and US policy against terrorism    
To: JMC@SAIL.Stanford.EDU
cc: : ;
In-Reply-To: <nFDT6@SAIL.Stanford.EDU>
Message-ID: <12514608035.80.J.JBRENNER@Macbeth.Stanford.EDU>


I'd like to suggest that as a practical matter it probably 
doesn't make a lot of sense for the US to abandon it's 
"moral highground", and go in for blood thirsty revenge tactics
as a method of suppressing terrorism (however effective it might be
for dealing with the immediate problem):

(1) Our strategy has been to try and attract converts by at least 
appearing to be the Good Guys (though admittedly we keep messing 
up by supporting unpopular "anti-communist" dictatorships).  

(2) Our politicians are likely to be afraid of having to defend 
such actions, especially if the actions fail.  Reagan may be 
looked on fondly for Grenada and Libya, but Carter is sneered at
for the Iranian hostage rescue mission.

(3) There's the danger that we could be manipulated into attacking 
the wrong party in the future by a forged claim of responsibility 
for some atrocity.

(Uh oh, I'm numbering my points.  I guess my clockwise 
brains are showing.)

The main course of "action" that I most prefer is to simply ignore 
anyone who takes hostages.  If you never respond to blackmail, 
it's much less likely anyone will try to blackmail you in the 
future.  

Notice that I'm still using what you might call "conservative"
reasoning:  I'm arguing for trading some innocent lives today for 
a greater number of innocent lives in the future.  There are
advantages in this kind of passivity, though:  (1) At least the 
appearence of "purity" is maintained.  We want to be sufferring 
martyrs rather than vicious barbarians. (2) I think it's easier for 
the political process to achieve inactivity rather than intelligent 
action.  

As a secondary course of action, I would consider things like 
"heroic rescues" or "military retaliation", but I think it
has to at least look like we've made an attempt at striking 
at the guilty and sparing the innocent (intentionally atttacking 
the families of the guilty would look too ruthless).  I think 
that side effects like the loss of international good will can 
be minimized this way.  It's easier for the politicians to 
sell as well. 

This is especially true if you can follow Reagan's example, 
and keep the press off of the scene.  The American people can 
be pretty gullible, and are likely to believe in myths like 
the "surgical strike" if they don't see children bleeding to 
death on television.  

-- Joe B. 

-------

∂01-Aug-89  0432	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	computer algebra challenge integrals   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 1 Aug 89  04:32:16 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 406007; Tue 1-Aug-89 07:29:58 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104949; Tue 1-Aug-89 02:54:21 PDT
Date: Tue, 1 Aug 89 02:54 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: computer algebra challenge integrals
To: macsyma-i@RUSSIAN.SPA.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890730120736.8.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890801095403.2.RWG@TSUNAMI.SPA.Symbolics.COM>


                                                2
                                         (a + 1)
                                       - --------
                                           4 b                            a + 1
    /          2           Sqrt(pi) i e           Erf(i Sqrt(b) (Log(x) + -----))
    |  a  b Log (x)                                                        2 b
    | x  e          dx = - -----------------------------------------------------
    |                                            2 Sqrt(b)
    /

and, differentiating w.r.t. a,

   /                 2
   |  a         b Log (x)
   | x  Log(x) e          dx = 
   |
   /

                                2
                         (a + 1)
                       - --------
                           4 b                           a + 1                  2
   Sqrt(pi) i (A + 1) e          Erf(i Sqrt(b) (Log(x) + -----))    a + 1  b Log (x)
                                                          2 b      x      e
   ------------------------------------------------------------- + -----------------
                                 3/2                                      2 b
                              4 b

etc, to get any (non-negative) integer power of the log.  (Differentiating wrt a
twice = wrt b once!)  If we could integrate e↑(x↑2+ax) erf(x), we could do at least
one negative power of the log.

∂01-Aug-89  0436	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	stigidity
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 1 Aug 89  04:35:57 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 406009; Tue 1-Aug-89 07:34:15 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104952; Tue 1-Aug-89 03:12:40 PDT
Date: Tue, 1 Aug 89 03:12 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: stigidity
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890801101224.3.RWG@TSUNAMI.SPA.Symbolics.COM>

Schroeppel points out that, while it is obvious that reversing
the digits of a decimal integer n preserves its divisibility by
3, it is slightly less obvious for octal and binary, where
reversal preserves |n mod 3|, though not necessarily n mod 3.

More generally, reversal preserves divisibility by d in base kd+-1.

∂01-Aug-89  0622	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	[ESC@STONY-BROOK.SCRC.Symbolics.COM: russian intransigence [was Pi]] 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 1 Aug 89  06:22:05 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 406026; Tue 1-Aug-89 09:23:13 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 104969; Tue 1-Aug-89 04:36:28 PDT
Date: Tue, 1 Aug 89 04:36 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: [ESC@STONY-BROOK.SCRC.Symbolics.COM: russian intransigence [was Pi]]
To: "tk@ai.ai.mit.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Included-msgs: <19890731140633.5.ESC@DJINN.SCRC.Symbolics.COM>,
               The message of 31 Jul 89 07:06 PDT from ESC@STONY-BROOK.SCRC.Symbolics.COM,
               The message of 31 Jul 89 07:06 PDT from Eric S. Crawley
Included-References: <19890730094357.6.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890801113612.7.RWG@TSUNAMI.SPA.Symbolics.COM>

[I bcc'ed you for some reason.  Here is how to reach me.]
Date: Mon, 31 Jul 89 07:06 PDT
From: Eric S. Crawley <ESC@STONY-BROOK.SCRC.Symbolics.COM>
Subject: russian intransigence [was Pi]
To: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
cc: dbailey@ew11.nas.nasa.gov, cgay@RUSSIAN.SPA.Symbolics.COM
In-Reply-To: <19890730094357.6.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Sun, 30 Jul 89 02:43 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

	Date: Sat, 29 Jul 89 06:09:12 -0700
	From: dbailey@ew11.nas.nasa.gov (David Bailey)

	 . . .

	I note that e-mail from you always has the return address
	rwg@russian.spa.symbolics.com.

    That should work, but . . .

No, it won't work all the time because Russian is not connected to the
Internet.  It is only connected to the Symbolics chaosnet.

					However, whenever I try to reply to
	your messages from this address, the mail is bounced.

    JMC@Stanford and TK@mit complain similarly.
							       Also, if I try
	to send directly to this address, no luck.  Even a ping to this
	address, although our name server is able to resolve it, shows all
	packets lost.  When I mail to you at yukon.scrc.symbolics.com, though,
	it usually works just fine.

    How about just rwg@symbolics.com?  (Or maybe gosper@symbolics.com?)

				     What is so peculiar about the russian
	system?

    Nothing, in theory, although it has been occasionally threatened with
    excommunication for running obsolete namespace data.

The link between SCH and SPA is chaos-only and does not forward IP
packets.  There is a local Internet in SPA but it is not connected to
the real Internet.  The domain system has no concept of non-contiguous
networks so the internet address for Russian on the SPA Internet gets
handed out when you resolve the host.

		 Is it only turned on part of the time?

    It's up a good 98%.  A bleep of a lot better than Symbolics's
    East/West link, which goes out many nights and weekends.

The link between SCRC and MIT is up 100% of the time.  We sometimes have
problems with the link between MIT and BBN now that MIT is not directly
connected to the ARPAnet.  The cross-country link between SCH and SCRC
is up most of the time (85%) modulo software and hardware problems with
the gateways.

    The problem is probably in our namespace.  When I went on no fee
    status, they reamed me off of most databases and lists, then
    piecemeal restored me, then reamed me a little more, then restored
    me a little more.  I'll ask some other people who get their mail
    through russian if they have similarly frustrated correspondents.

    Have you ever responded to mail from Wolfram research?  I never
    succeeded with the supplied uunet or University of Illinois return
    addresses, but win every time with wri.com.

You are defined on all the mailers in Symbolics so there should be no
problem forwarding mail to you once it gets to a mailer here.  I would
suggest that you find the ZMail hack that puts the REPLY-TO: field on
all your mail and has an address on Stony-Brook, Yukon or some SCRC file
server.  You might check with some of the support folks in SCH for the
hack.

One additional note:  rwg@Symbolics.COM will work only part of the time
because of braindamage in our mailers.  Your best bet is to use a full
host description such as rwg@Stony-Brook.SCRC.Symbolics.COM or
rwg@Yukon.SCRC.Symbolics.COM.

∂01-Aug-89  1000	JMC  
Shankar

∂01-Aug-89  1644	pjd@riacs.edu 	Your submission to CACM. 
Received: from icarus.riacs.edu by SAIL.Stanford.EDU with TCP; 1 Aug 89  16:44:18 PDT
Received: from hydra.riacs.edu by icarus.riacs.edu (5.59/2.1G)
	   id AA18114; Tue, 1 Aug 89 16:45:30 PDT
Received: by hydra.riacs.edu (4.12/2.0N)
	   id AA24988; Tue, 1 Aug 89 16:45:33 pdt
Message-Id: <8908012345.AA24988@hydra.riacs.edu>
Date: Tue, 1 Aug 89 16:45:33 pdt
From: Peter J. Denning <pjd@riacs.edu>
To: jmc@sail.stanford.edu
Subject: Your submission to CACM.
Cc: jimm@acmvm.bitnet


John,

Jim Maurer has forwarded a copy of your short piece (Networks
considered harmful -- for electronic mail) which you submitted
to the CACM.  As it stands this piece fits perfectly into the
ACM Forum.  I am prepared to publish it there, but I notice from
your cover letter that you reserved the last word on this option.

If you would like to publish something more substantial on this
subject -- but not more than 2000 words -- there is another
mechanism that then becomes suitable, the signed Viewpoint
commentary.  I offer you this option, but to exercise it,
you'll have to add some more material to the piece.

Please let me know what you'd like to do.

Thanks for your support of the CACM.

Peter

∂02-Aug-89  0640	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	After he's through canning Rose,  
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 2 Aug 89  06:40:25 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 406542; 2 Aug 89 08:59:41 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 105162; Wed 2-Aug-89 04:24:00 PDT
Date: Wed, 2 Aug 89 04:23 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: After he's through canning Rose,
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890802112339.1.RWG@TSUNAMI.SPA.Symbolics.COM>

do you think we can prevail upon commissioner Giamatti to rename it
the World Sequence?

Elevating the discussion only slightly,

                   inf                         1 2 n + 1
                   ====  Binomial(2 n, n) (y - -)
                   \                           y
           ln(y) =  >    ------------------------------- ,
                   /                            n
                   ====         (4 n + 2) (- 16)
                   n = 0

which is a generating function of the not-quite-Catalan numbers.
Pa(ren)thetically, when is Binomial(2 n, n)/(4 n + 2) an integer?
The first n is 97.  Define Fubar's constant as the probability for
"random n". . .   What would you bet it exists?  If so, maybe = 0?

The density increases!  136 178 192 199 292 313 332 448 467 472 478
487 535 542 577 604 617 697 773 790 797 852 885 940 962 967 997 1017
1045 1096 . . .  One thing for sure, 2n+1 must be composite, since
Binom(2 n,n) mod 2 n + 1 = (-1)↑n for 2 n + 1 prime (and conversely,
until n = 2953 !)  And there are twin n at 1291 and 1292 !

*Now* what would you bet for Fubar's constant?

∂02-Aug-89  0651	P.DRAVID@GSB-WHY.Stanford.EDU 	re: Higgins   
Received: from GSB-WHY.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Aug 89  06:51:28 PDT
Date: Wed 2 Aug 89 06:48:17-PDT
From: Ajay Dravid <P.DRAVID@GSB-WHY.Stanford.EDU>
Subject: re: Higgins 
To: JMC@Sail.Stanford.EDU
In-Reply-To: <NG0ez@SAIL.Stanford.EDU>
Message-ID: <12514905345.139.P.DRAVID@GSB-WHY.Stanford.EDU>

Prof. McCarthy:
I'm worried -- this is perhaps the first time I find myself agreeing with you on 
something!
-------

∂02-Aug-89  0800	JMC  
bicycle, eat here banana

∂02-Aug-89  0814	chandler@Polya.Stanford.EDU 	Pat Simmons......    
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Aug 89  08:14:40 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA11426; Wed, 2 Aug 89 08:15:19 -0700
Date: Wed, 2 Aug 1989 8:15:18 PDT
From: "Joyce R. Chandler" <chandler@Polya.Stanford.EDU>
To: jmc@sail, sloan@score
Cc: davis@score, tlc@sail
Subject: Pat Simmons......
Message-Id: <CMM.0.88.618074118.chandler@Polya.Stanford.EDU>

Will not be in today.  She called me this morning and told me to tell you all
that she isn't feeling well.

∂02-Aug-89  0900	JMC  
Gerald Gillespie 3-3063 856-9580

∂02-Aug-89  0928	pjd@riacs.edu 	re: Your submission to CACM.  
Received: from icarus.riacs.edu by SAIL.Stanford.EDU with TCP; 2 Aug 89  09:28:22 PDT
Received: from hydra.riacs.edu by icarus.riacs.edu (5.59/2.1G)
	   id AA18921; Wed, 2 Aug 89 09:29:11 PDT
Received: by hydra.riacs.edu (4.12/2.0N)
	   id AA03277; Wed, 2 Aug 89 09:29:15 pdt
Message-Id: <8908021629.AA03277@hydra.riacs.edu>
Date: Wed, 2 Aug 89 09:29:15 pdt
From: Peter J. Denning <pjd@riacs.edu>
To: JMC@SAIL.Stanford.EDU
Subject: re: Your submission to CACM.
Cc: jimm@acmvm.bitnet

John,

We aim to get the signed Viewpoints on two printed pages.  If you
shoot for about 1500-1600 words it will fit fine.

Yes electronic submission (straight text) is very helpful.  Send
copies to both me and Jim Maurer (cf his address in the CC field
of this note).

Peter

∂02-Aug-89  0938	chandler@Polya.Stanford.EDU 	re: Pat Simmons......     
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Aug 89  09:38:39 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA18343; Wed, 2 Aug 89 09:39:15 -0700
Date: Wed, 2 Aug 1989 9:39:15 PDT
From: "Joyce R. Chandler" <chandler@Polya.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: Pat Simmons...... 
In-Reply-To: Your message of 02 Aug 89 0923 PDT 
Message-Id: <CMM.0.88.618079155.chandler@Polya.Stanford.EDU>

The "tlc" message bounced and I resent it correctly.  Thanks for catching
that, John.

∂02-Aug-89  1718	RQ.WSB@Forsythe.Stanford.EDU 	hoover press paper  
Received: from Forsythe.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Aug 89  17:18:33 PDT
Date:      Wed,  2 Aug 89 16:29:43 PDT
To:        jmc@sail.stanford.edu
From:      "William Bonnett" <RQ.WSB@Forsythe.Stanford.EDU>
Subject: hoover press paper




            TERMS FOR SOVIET ACCESS TO WESTERN COMPUTER TECHNOLOGY

                                 John McCarthy

     This essay proposes that the United States and its allies undertake to set
terms for substantially increasing Soviet access to Western computer and other
electronic technology. These terms should ensure that this increased access
does not decrease our security against attack or increase our defense costs.
If we do not set such terms, we may find that the restrictions have disappeared
and that we have gotten nothing in return.  To get the Soviets to accept our
terms, we will also have to sell them on changing their ways of importing
technology and on the advantages of paying the high price we should ask.*

Soviet Problems with Computer Technology

     The Soviets now lag far behind the West in computer technology. They have
been behind ever since the computer industry started, and they are not catching
up.
     Through the Coordinating Committee for Multilateral Export Controls
(CoCom),the Western countries restrict, for defense reasons, the export of
technology to the Soviet Union. In the computer area the newest technologies
are restricted, but the restrictions on any particular technology are usually
removed after some years.

_____________
     *This paper complements the recent report, Global Trends in Computer
Technology and Their Impact on Export Control, prepared by a committee of the
National Research Council, of which Sy Goodman was chairman.  I had a small
part in composing this report.  In my opinion, it does not sufficiently
consider the advantages to the Soviet Union of normal commercial relations with
equipment suppliers, in contrast to the ability to acquire units of a
particular kind of equipment only once.



    The Soviets import some Western computer technology in compliance with the
restrictions.  They steal other technology, chiefly by  getting unscrupulous
Western businessmen to set up dummy companies, purchase the computers, and
smuggle them into the USSR. Every so often Western countries catch and arrest
someone doing this. Only small numbers of computers are illegally imported.
They are probably imported primarily to be copied rather than used. In the
mid-1960s the Soviets started copying  the IBM 360/370 line of computers but
with only moderate success. In the mid-1970s they began making computers
intended to be compatible with the DEC PDP-11 and, more recently, the VAX.
These undertakings haven't been very successful either.

    Theft as a means of obtaining technology has serious disadvantages for the
Soviets, as emigrants from the Soviet Union and other Soviet bloc countries
report. One of these disadvantages is the absence of the continuous
communication between the users of the hardware or software and the suppliers
that normal use of computer technology involves. This communication involves
correcting users' mistakes, resolving ambiguities and incompleteness in the
documentation,eliminating bugs that have affected the users' work, and getting
information about projected improvements in the software and hardware. The
Soviets and their allies have no reliable way of communicating with the
suppliers of the technology they steal. As a result, the initial compatibility
of the stolen technology with the foreign technology on the market is often
lost when the Soviets have to improvise a solution to a problem that arises,
for that solution is likely to be incompatible with the supplier's solution to
the problem. As a consequence, software or auxiliary equipment acquired later
requires expensive and time-consuming modification.

   Another disadvantage is that documentation obtained by theft often doesn't
correspond to the hardware it's supposed to describe and explain. One might
suppose that documentation about hardware is even easier to steal than
hardware, but apparently the KGB isn't equipped to steal exactly what the
Soviet users of Western technology want. Or is may be difficult for those users
to determine exactly what it is they want or need. Documentation obtained
opportunistically is especially likely to be incompatible with equipment
obtained earlier.

   The KGB, or whoever steals the technology, insists that the technology be
kept under wraps, which  interferes with communication within the Soviet Union.
Institutions with stolen technology are also restricted in their communication
with foreigners.

   It is likely that the KGB doesn't understand the difficulties their methods
create for their Soviet customers. They are probably proud of their
intelligence coups. Our intelligence people are chagrined at the KGB's
successes but may not be in a position to analyze accurately how much use such
theft is to Soviet industry.

   The West's Cocom restrictions have important effects in limiting the
availability of computer technology in the Soviet Union. Requiring the firms'
selling computers to the USSR to obtain separate export licenses for
maintaining and servicing those computers doubtless plays a role. However,
Western restrictions often hinder the Soviet's legal acquisition and use of
Western technology less than do the Soviets' own restrictions.

   For example, the Soviets have restricted foreign travel by their own
scientists and engineers far beyond the restrictions imposed by their shortage
of foreign currency, and although their central libraries receive all the
important foreign scientific journals, distribution of these journals
throughout the country is poor and few individuals subscribe to them. The
restrictions on travel and on access to scientific journals from abroad makes
use of foreign ideas Soviet scientists difficult and encourages complacency
about how well they are doing. Whether perestroika is changing this situation
fundamentally is still not clear.

   When the Soviets do buy foreign computers legally, they usually restrict
their contacts with the service organizations of the companies from which they
purchase them, often not permitting these companies to set up service
organizations within the Soviet Union. This prohibition makes service calls
very difficult.

Encouraging the USSR to Rejoin the World Technologically

   Mikhail Gorbachev's reforms, however, are leading people all over the world
to view the Soviet Union more favorably. Many people didn't see the defense
importance of observing CoCom restrictions in the past, and others are likely
to join them. The number of countries that can supply computer equipment and
software has grown, and many are not members of CoCom. Thus the CoCom system
may become considerably weaker in the future.

    Some Westerners favor relaxing the CoCom restrictions unilaterally to
encourage Gorbachev, to promote peace, to save what they consider wasted
effort, or to make a profit.  Their views prevail from time to time, but there
is no reason to suppose they will succeed in getting the West to abandon the
restrictions to an extent that would remove them as a hindrance to Soviet
technology.  Therefore, the West has bargaining power.

   It would increase the Soviet standard of living considerably over the long
term to induce the West to relax or abandon its restrictions on technology
transfer. (This essay assumes that our motivation for our restrictions is
security and that we have no desire to keep down the Soviet standard of living
either from dislike or to try to force changes in their society.) The Soviet
standard of living has not been a dominant consideration with the Politburo in
the past, and it isn't obvious how important it is today relative to military
advantage; it would nonetheless be to the Soviets' advantage to pay a
considerable price for the West's relaxation of its restrictions. A technology
race with the rest of the world is a race the Soviets can't win.

   Thus to the extent that the West is agreeable, the best Soviet strategy is
to rejoin the world technologically, which means buying Western products and
technology from a variety of countries and using the same commercial practices
as do Western countries. It means motivating Western companies to set up sales
and service organizations within the Soviet Union. It means letting Soviet
organizations deal directly with foreign companies, rather than only through
the Ministry of Foreign Economic Relations. It means letting Soviet engineers
and scientists subscribe freely to foreign publications and travel abroad
freely to advance their work.

   The advantages of rejoining the world technologically will not be obvious to
organizations like the KGB, proud of their successes in stealing technology. It
also goes against the tendency of the party to control everything. However, the
Gorbachev regime has been taking some steps in this direction.

   What the West needs to do is figure out how to sell the Soviets on the
advantages of rejoining the world. Otherwise, the attempt to negotiate the
terms and conditions for greater Soviet access to Western computer technology
will fail because the Soviet government and communist party won't find the
price worth paying. Perhaps we should even consider doing more than making
normal commercial relations possible in the computer field. We should consider
how we might help the Soviets. Naturally, actual help should command a higher
political and defense price than just easing restrictions.

The Terms of an Agreement

    The scope of this essay does not allow for a comprehensive treatment of the
price we should ask. Presumably people will differ about this along the same
lines that they have differed in the past about relations with the Soviet
Union. However, our main criterion--a point that many will support--should be
that any agreement should not make us less secure from attack or increase
defense costs.

   The Soviets might prefer agreements restricted to an exchange of our
technology for theirs. However, we have so much more to offer than they that
purely technological exchanges can have only limited scope if they are to be
even. The area where such exchanges would be most reasonable is in software
involving advanced mathematics produced by an individual or small group. That's
what the Soviets do best because their system inhibits individual achievement
less than it inhibits work that requires coordination, communication, and
motivation of many people.

   We need, then, to imagine a sequence of successively more comprehensive
agreements. Reduction of Soviet territorial and industrial secrecy should be
part of the price. Giving up some of the military advantage this secrecy gives
them will make disarmament agreements more verifiable. (Some of this secrecy
was pointless in any case, and we are getting some concessions for nothing. For
example, the Soviets have now promised the Soviet public that they will publish
correct maps, including a road atlas of the Soviet Union.)

   But the Soviets'main concessions have to be in the military area. They need
to give up some of their conventional armaments advantage in Europe and perhaps
their military draft. Certainly we should make sure that an agreement reducing
restrictions contains enough such concessions that our own military costs don't
increase.

   Experts frequently have said, "The Soviets will never give up X," only to be
confounded by events. We should enter negotiations without preconceptions about
what they might agree to. According to Arkady Shevchenko, the U.N. under
secretary who defected to the United States in 1985, the Soviets have never
feared an unprovoked Western attack. This suggests that also we make no
presumptions about their state of mind on the basis of ideas of symmetry
between their situation and ours.

   The single most effective way for the Soviets to improve their standard of
living is to reduce military and police expenditures, and the fastest way for
them to do it is unilaterally. However, this approach presents institutional
difficulties, the extent of which we cannot predict.



Summary

     1. The West should decide on terms for reducing restrictions on the sale
of computer technology to the Soviet Union. Otherwise, these restrictions may
just evaporate with no corresponding gain in Western security, which would
reinforce the KGB doctrine that the West consists of villains and suckers.

     2. We need to persuade the Soviets of the advantages for their standard of
living of rejoining the world technologically. These advantages will motivate
them to pay the price we must ask, which should be such that our security is
enhanced and our defense costs reduced.

     3. Further, we might consider actual help to the Soviet Union and what
price we should charge for it.


Postscript

     I made a two-week visit to Moscow in April 1989 that led to the following
elaborations but no fundamental changes. During my visit I showed the above to
many people, from programmers in cooperatives to high officials of the Academy
of Sciences, all of whom agreed with the main points about the Soviet computer
situation and the importance of normal commercial relations.

   Soviet economic weakness arises out of a lack of commercial organization
rather than a lack of technology, but much of the enthusiasm for computers
among Soviet administrators and their scientific advisers is aimed comes from
their desire to use computers instead of freer markets to reduce irrational
economic behavior. I don't think this will work because the irrational
economics occur much less from an inability to compute correct prices, for
example, than from various groups using their power and influence to optimize
their own positions. Computerization won't prevent people with power from
abusing it nor will it prevent empire building and semi-feudal hierarchies.

   The partial freeing of some aspects of the economy via cooperatives has led,
on occasion, to profit making that seems enormous by Soviet standards. For
example, in 1988, when cooperatives were first permitted to make barter deals
with foreign companies, one cooperative arranged to barter twigs and branches
from the Soviet lumber industry for personal computers from a Japanese firm
that made wallboard out of the scrap. The cooperative then sold the computers
in the Soviet market for high prices and made large profits. These profits led
to envy (an important characteristic of Soviet ideology),complaints, a decree
that cooperatives could barter only their own products and only for foreign
goods that they used directly, and the freezing of the bank accounts of some
cooperatives. This reaction was precisely counterproductive because there are
many opportunities for barter deals with foreign companies,and letting hundreds
of cooperatives barter would benefit the economy and, through competition,
bring down the profits. This example also demonstrates that innovative trading
activity would in itself benefit the Soviet economy.

  As the following shows, Soviet internal prices for Western computers and
auxiliary equipment are 30 to 100 times Western prices, given the official
exchange value of the ruble (about 1.75 dollars per ruble at the time of my
visit).

     A Soviet-made 8086-compatible 8-megahertz(mhz) computer with 256K
random-access memory (RAM), a 350-kilobyte floppy disk from Bulgaria, a
5-megabyte (mb) hard disk with 170 milisecond access time, and a text-only
monochrome video card costs 26,000 rubles.

     An East German Robotron 100-character-per-square-inch dot matrix printer
costs 26,000 rubles.

   The following prices, said to be from a Gosnab price list, are what
government institutions are allowed to pay cooperatives for IBM XT- and
AT-compatible computers.


     An 8- or 10-mhz 8088 with 640K RAM, XT, and color display costs 45,000
rubles.

     A 10-mhz, 80286 machine with a 30-mb hard disk, a 1-mb RAM, and an Aga
monitor plus Epson FX80 printer costs more than 60,000 rubles.

     An 80386, 16-mhz, 80387 coprocessor with a cheap 4-pen plotter, and 1-mb
memory costs 200,000-250,000 rubles; a 20-mhz 386 with a 140-mb hard disk and
an ESDI interface costs 300,000-330,000 rubles. (In the West such a machine
costs from $3,000 to $4,000.)

     A 4-pen plotter costs 100,000 rubles.

     A Hewlett-Packard-compatible laser printer with a 1.5-mb bitmap memory
costs 70,000-80,000 rubles ($2,000 in the West).

     An 80287 coprocessor costs 10,000 rubles (here $100).

     A 1024 x 1024 monochrome display costs 20,000-30,000 rubles (here $1,000).

     The XT computers are exportable under a general license. Other types can
get specific licenses, but the CoCom restrictions are highly effective. Only
two 386 machines had been sold in Moscow as of April 1989.

   By comparison, Soviets can get cars for much smaller premiums, and a
three-bedroom apartment in the center of Moscow costs 25,000 rubles and a
dacha, between 60,000 and 70,000 rubles. A VCR costs 4,000 rubles vs $300 in
the West. Thus the commercial aspects of any agreements with the West on
computer technology will be as important as the technological aspects.
One note of caution: those U.S. commercial firms interested in making deals
should avoid large expenses up front because unexpected bureaucratic delays
could cause large losses.

To:  JMC@SAIL.STANFORD.EDU

∂03-Aug-89  0501	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	surprisingly icky generating function  
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 3 Aug 89  05:01:49 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 406848; 3 Aug 89 08:00:19 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 105317; Thu 3-Aug-89 03:02:10 PDT
Date: Thu, 3 Aug 89 03:01 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: surprisingly icky generating function
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890803100150.4.RWG@TSUNAMI.SPA.Symbolics.COM>

for the reciprocal center binomial coeffs:

                     SQRT(T)         T           ∞
                ASIN(-------) SQRT(-----) + 1   ====          K
                        2          4 - T        \            T
                ----------------------------- =  >    ----------------
                                T               /     BINOMIAL(2 K, K)
                            1 - -               ====
                                4               K = 0

(C1369) TAYLOR(%,T,0,6);
Time= 818 msecs
                    2    3    4    5     6
               T   T    T    T    T     T
(D1369)/T/ 1 + - + -- + -- + -- + --- + --- + . . . = 
               2   6    20   70   252   924

		    2    3    4    5     6
	       T   T    T    T    T     T
	   1 + - + -- + -- + -- + --- + --- + . . .
	       2   6    20   70   252   924




∂03-Aug-89  0651	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	After he's through canning Rose,  
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 3 Aug 89  06:51:08 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 406879; 3 Aug 89 09:49:37 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 105336; Thu 3-Aug-89 05:10:14 PDT
Date: Thu, 3 Aug 89 05:09 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: After he's through canning Rose,
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wilf@central.cis.upenn.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890802112339.1.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890803120953.7.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Wed, 2 Aug 89 04:23 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

    do you think we can prevail upon commissioner Giamatti to rename it
    the World Sequence?

    Elevating the discussion only slightly,

                                                   1 2 k + 1
                       ====  Binomial(2 k, k) (y - -)
                       \                           y
               ln y  =  >    ------------------------------- ,
                       /                            k
                       ====         (4 k + 2) (- 16)
                      k >= 0
(Half line proof:)
                     =  asinh sinh ln y

(sinh ln y = (y-1/y)/2.)  In the same vein, atanh tanh ln y
is an instantaneous derivation of the ln (1+y)/(1-y) speedup trick.
Also,
                                  k   2      1 2 k + 2
                       ====  (- 1)  k!  (y - -)
                2      \                     y
              ln  y  =  >    ------------------------- ,
                       /           2 (2 k + 2)!
                       ====
                      k >= 0

which is interesting, since ln↑2 itself doesn't have a very nice
powerseries expansion--it requires harmonic numbers.

(C1414) TAYLOR(%,Y,1,7);

                                           4            5              6            7
                  2          3   11 (Y - 1)    5 (Y - 1)    137 (Y - 1)    7 (Y - 1)
(D1414)/T/ (Y - 1)  - (Y - 1)  + ----------- - ---------- + ------------ - ----------
                                     12            6            180            10

                                           4            5              6            7
                  2          3   11 (Y - 1)    5 (Y - 1)    137 (Y - 1)    7 (Y - 1)
 + . . . = (Y - 1)  - (Y - 1)  + ----------- - ---------- + ------------ - ----------
                                     12            6            180            10

 + . . .

Why can't you get the ln↑2 series simply by dividing the ln y series by
y and integrating?  (Of course, then you could get ln↑3, ln↑4, . . .)

    which is a generating function of the not-quite-Catalan numbers.
    Pa(ren)thetically, when is Binomial(2 n, n)/(4 n + 2) an integer?
    The first n is 97.  Define Fubar's constant as the probability for
    "random n". . .   What would you bet it exists?  If so, maybe = 0?

    The density increases!  136 178 192 199 292 313 332 448 467 472 478
    487 535 542 577 604 617 697 773 790 797 852 885 940 962 967 997 1017
    1045 1096 . . .

		     One thing for sure, 2n+1 must be composite, since
    Binom(2 n,n) mod 2 n + 1 = (-1)↑n for 2 n + 1 prime (and conversely,
    until n = 2953 !)  And there are twin n at 1291 and 1292 !

. . . 10000 10003 10004 . . .  Does anybody understand Dickman's fcn
enough to say whether there's an asymptotic probability that n will have
no divisors > c sqrt n?  Does it matter?

    *Now* what would you bet for Fubar's constant?

.069 is looking pretty good out near 10000 or so.

∂03-Aug-89  0915	scherlis@vax.darpa.mil 	Financial Information Needed ASAP   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 3 Aug 89  09:14:29 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA26183>; Thu, 3 Aug 89 10:38:30 -0400
Posted-Date: Thu 3 Aug 89 10:39:19-EDT
Received: by sun45.darpa.mil (4.1/5.51)
	id AA01634; Thu, 3 Aug 89 10:39:20 EDT
Date: Thu 3 Aug 89 10:39:19-EDT
From: William L. Scherlis <SCHERLIS@DARPA.MIL>
Subject: Financial Information Needed ASAP
To: SW-PI@vax.darpa.mil
Cc: njacobs@vax.darpa.mil, SCHERLIS@vax.darpa.mil
Message-Id: <618158359.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>

To the Algorithms and Software PIs:
	It is clear that our office-wide attempt to solicit financial
data using the magic of computers and automation has not been as
successful as intended.  The address list for our requests turned out
to be a random collection of email addresses from our mailing lists.
So I'm sending this out again to the Software/Algorithm PI mailing
list that I use.  If you haven't seen this before, please send a note
to Nicole Jacobs (njacobs@vax.darpa.mil) and let her know so the other
list can be fixed.  (It is likely that only one addressee at a given
site received the message rather than the usual group on my file.)
Also reply as soon as possible, please!  If you have seen this before
and responded, then no additional action is required.  Note that
inquiries should go to Nicole Jacobs -- NOT to Louise Lorenzen.
I am very sorry about this mess-up.  
	Thanks,
					Bill

================================================================



MEMO FOR PI OF TASK: <your task>
          TASK CODE: <your task code>

In order to avoid the yearly rush to collect financial data in
support of Fall funding, we are re-instituting last year's 
reporting process with the intention of turning it into a 
routine, quarterly process.  In the future we intend to maintain
the information in a database, and extract update information
automatically from quarterly email reports (which will require 
less information than this initial report).

This report is not a requirement of your contract or grant and does
not take the place of any report required in the contract or grant.  
We are asking for this information in order to do a better job of
managing our funds in order to support you better.

Please send your report to pi-data@vax.darpa.mil, not later than 
28 July 89.  Questions on the report procedure may also be directed
to that address.  If absolutely necessary telephone inquiries may be
....

We try to keep the reporting workload on the research community as
light as possible.  Thank you for helping us support you.  

                DARPA/ISTO Program Management

---------------------------------------------------------------------
Here is the information we need: (Two minus signs denote
comments. In your response, you may annotate with 
comments headed with --. However, please leave the format of the
reporting lines exactly as shown.)

CODE:  Task code from first line of this message (THIS IS IMPORTANT)
TITLE: Task name from first line of this message
CONTRACTOR:  Contractor name (e.g. company or university)
PI-NAME:  PI name
PI-EMAIL:  PI email address
PI-PHONE:  PI phone number
ADMIN-NAME:  Admin name
ADMIN-EMAIL:  Admin email address
ADMIN-PHONE:  Admin phone number
ARPA-ORDER: ARPA Order number
CONTRACT: Contract number
START: Contract start date
EARLYSTART: Date of pre-contract work authorization(omit if none)
END: Planned end date for effort on current contract
VALUE: Contract value (include funded options)
AUTH-RCD: Total spending authority received to date
-- For following line
-- If you know that the agent or others has directly contributed
-- to this contract/task:  show total amount of non-DARPA funding
-- to date.
-- If you know there was jointfunding but don't know how much:
-- include the line JOINTFUNDING with no $$.
-- If there was no joint funding, you are unsure, or don't
-- know what I am talking about: omit the JOINTFUNDING line.
JOINTFUNDING: joint funding 
BILLED: Total amount billed to government as of 1 July 89
-- The following lines are to establish a profile of spending on the
-- project. They are intended to be used by the program manager for the
-- estimation of his budget requirements for this year and for outyears.
-- NOTE: We ask for these numbers because they are likely to have changed
-- since the project was selected for funding.  These numbers will be
-- treated as your estimate of required funding for the project as is
-- currently agreed.  They need not match the profile listed in the
-- contract nor will they be considered contractually binding.  They will
-- form a basis for the managers estimate of nextyear funding. 
-- 
-- Each line represents one calendar year. The four fields represent the
-- four calendar quarters.  (Report in thousands of dollars) There should
-- be one line for each calendar during which this contract/task is
-- active.  These values represent your actual expenses (for past
-- quarters) and your estimate of expenses (for future quarters).  Use
-- RECUR for recurring costs such as labor/travel/maintenance/overhead
-- use NONRECUR for large "one shot" expenses such as equipment
-- purchases.  Generally NONRECUR will be zero except for quarters with
-- major purchases.
RECUR: CY88 ddd ddd ddd ddd
RECUR: CY89 ddd ddd ddd ddd
RECUR: CY90 ddd ddd ddd ddd
RECUR: CY91 ddd ddd ddd ddd
RECUR: CY92 ddd ddd ddd ddd
NONRECUR: CY88 ddd ddd ddd ddd 
NONRECUR: CY89 ddd ddd ddd ddd
NONRECUR: CY90 ddd ddd ddd ddd
NONRECUR: CY91 ddd ddd ddd ddd
NONRECUR: CY92 ddd ddd ddd ddd
-- The following allows planning for options.
-- if there are unselected options that may be invoked at a later
-- time, include them here.
-- Repeat the group of OPTION-NAME and OPTION for each unselected
-- option. (Note: options currently on contract should be covered above).
-- If there are no options then omit this section.
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
OPTION-NAME: name of unfunded option
OPTION: CY90 ddd ddd ddd ddd
OPTION: CY91 ddd ddd ddd ddd
OPTION: CY92 ddd ddd ddd ddd
-- The following is general remarks. Caveats. Comments. Suggestions ....
REMARKS: This is free-form and is terminated by end of message.
------------------------------------------------------------------------
SPECIAL NOTES: 

a. We would like the admin email address to be one which is read
   frequently, whether the PI is available or not.  This will
   provide a backup contact for actions which must be completed
   quickly.

b. All amounts are in thousands of dollars. If any item is zero,
   please provide it anyhow. Punctuation in dollar values is optional:
   dollar signs, commas and "K" will be ignored.

c. Data pertain to current contract or grant identified by the
   task code; please do not report amounts for planned/pending
   renewals. 

d. END should be at or near end of contract period.  If it is 
   after end of contract, you should request a no-cost extension
   so contract covers period of performance.  Through your Program
   Manager separately.

e. For tasking or requirements contracts, data refer to
   just your task, not the whole contract.  Please identify any 
   designator established by the Contracting Officer in CONTRACT. 
   (This will NOT be the same as the Task Code DARPA/ISTO is using 
   to identify your effort.)

f. We plan to input directly to our database from your message, 
   so please use the exact names above, and separate data from 
   description with a colon (THIS IS IMPORTANT). 

g. Please follow this example as closely as possible:

CODE: Z1
TITLE: Advanced Computing Environment
CONTRACTOR: Creative State University
PI-NAME: J.J. Smith
PI-EMAIL: j.smith@cs.creative.edu
PI-PHONE: 123-555-1234
ADMIN-NAME:: M.A. Jones
ADMIN-EMAIL: contract@cs.creative.edu
ADMIN-PHONE: 123-555-4321
ARPA-ORDER: 1432
CONTRACT: N00139-87-C-0876
START: 10 Apr 89
EARLYSTART: 1 Apr 89
END: 9 Apr 92
VALUE: $811K
AUTH-RCD: $250K
BILLED: $136K
RECUR: CY89  0 31 45 50
RECUR: CY90 50 50 50 50
RECUR: CY91 50 50 65 65
RECUR: CY92 50  5  0  0
NONRECUR: CY89 0 120 0 0 
NONRECUR: CY90 0 0 80 0
NONRECUR: CY91 0 0 0 0
NONRECUR: CY92 0 0 0 0
OPTION-NAME: More Creative Stuff
OPTION: CY91 0 0 30 80
OPTION: CY92 120 50 5 0
REMARKS: Spending behind plan due to trouble hiring programmer.
	 We are planning to get new programmer sometime next month.
--END OF MESSAGE----

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

-------

∂03-Aug-89  0921	CLT 	financial projection for AI   
To:   JMC, VAL    

as of jan 90 there will be  460k = 500k(darpa) + 130k(nsf) - 170k debt
    (qlisp,mtc,shoham)
with monthly spending of 32k  (plus travel and jmc summers) this leaves
about 13 months of money

∂03-Aug-89  1001	VAL 	re: bad  not   
[In reply to message rcvd 02-Aug-89 18:25-PT.]

There do that about half of the time, and I've marked all these bad nots.

∂03-Aug-89  1041	MPS  
Good morning

The Xerox people called.  The engineer said you could plug
the battery in to charge it.  Do not plug the xerox in the wall.
This equipment was not made for US consumption.  There is no
guarantee.

I need to know where you have the National Academy Press address.

Thanks.

Pat

∂03-Aug-89  1203	VAL 	Index: First draft  

Ability
Abnormality
Accessibility relation
Action
	deliberate
Advice taker
Airport problem
Alternativeness relation, see Accessibility relation
Anthropomorphism
Approximate theory
Artificial intelligence problem
Artificial intelligence research
	applied
	basic
	deficiences
Aspect
ATM, see Automatic teller
Autoepistemic logic
Automatic teller
Beauty, appreciation of
Behavioral definition, see operational definition
Behaviorism
Belief
Best object
Biology
Blocks world
Body
Brain
Causality
CBCL, see Common Business Communication Language
Cancellation of inheritance
Canning
Cellular automaton
Chess program
Chinese room
Circumscription
	applications of
	collapsible
	domain
	formula
	predicate
	prioritized
Circumscription policy
Circumscriptive inference
Closed world assumption
Cohistorical situations
Combinatorial explosion
Common Business Communication Language
Common sense
Commonsense inertia
Commonsense knowledge
Commonsense physics
Commonsense reasoning
Communication
Communication convention
Computational complexity
Concept
Conciousness
Concurrency
Conjecture
Context
Continuous time
Control
Counterfactual
Creativity
DARPA
Declarative sentence
Deduction
Default
Denotation
Deontic logic
Design stance
Desire
Determinism
Duties of machines
Education by brain surgery
EKL
Elaboration tolerance
Emotion
Empiricism
EMYCIN
Epistemic logic
Epistemic alternative
Epistemological adequacy
Event
Event calculus
Exception
Existence
Expert system
Expert system shell
Extensionality
Fact
Finite automaton
First order language
Fluent
	propositional
FOL
Formal literature
Frame
Frame problem
Free will
Future operator
General Problem Solver
General purpose database
GPS, see General Problem Solver
Hate
Herbrand universe
Heuristic
Heuristic adequacy
History
Horn clause
Humor
Imperative sentence
Incrementalism
Induction, mathematical
Information storage convention
Inheritance
Intellect
Intelligent backtracking
Intention
Intentional stance
Introspection
Is-a hierarchy, see Inheritance
Kempe transformation
Knowledge
	joint
Knowledge engineer
Language
	abstract
	artificial natural
	natural
Law of ability
Learning
Leibniz's law
Lexicographic ordering
Life automaton
LISP
Logic
Logic programming
Love
Map coloring
Martian
Meaning
Mental qualities
Mental state
Metaepistemology
Metaphilosophy, see Metaepistemology
Metaphysical adequacy
Metaphysics
MICROPLANNER
Mind
Minimal inference
Minimal models
Minimal entailment
Minimization
Missionaries and cannibals
Missoury program
Modal function
Modal logic
Modal operator, see Modal logic
Monotonicity, see Nonmonotonic reasoning
Morality
Motor parameter
MYCIN
Name
National Science Foundation
Natural selection
Necessity
Nominalism
Nonmonotonic logic
NP-completeness
NSF, see National Science Foundation
Objectivity
Observation
Occam's razor
Ontology
Operational definition
Parallel programs, see Concurrency
Parameter
Philosophy
	naive
Physical stance
Physics
Plan
PLANNER
Policy
Positivism
Possible worlds
Postponable variable
Predicate calculus, see First order logic
Prediction
Present object
Probability
	and nonmonotonic reasoning
Production system
PROLOG
Priority, see Circumscription, prioritized
Proof checker
Proposition
Psychology
	introspective
Qualification
Qualification problem
Rationalism
Rationality
Realism
Reasoning
	conjectural
	commonsense
	nondeductive
	nonmonotonic
Reasoning program
Referential opacity
Reification
Representation problem
Resolution
Rights of machines
Robot
Rule
Search
Second order language
Self
Self-conciousness
Self-knowledge
Sense
Senses
Separable formula
Sequence extrapolation
Simple abnormality theory
Simultaneous actions, see Concurrency
Situation
Situation calculus
State
Strategy
	feasible
STRIPS
Structural definition
Tense logic
Thermostat
Thought
Time
Time-sharing system
Tool
Transitive closure
Turing test
Ultranatural deduction
Uncertainty
Understanding
Unique names assumption
Unintended model
Variable predicates
Visual images
Will
Yale shooting

∂03-Aug-89  1246	scherlis@vax.darpa.mil 	acks  
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 3 Aug 89  12:45:59 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA00412>; Thu, 3 Aug 89 15:38:56 -0400
Posted-Date: Thu 3 Aug 89 15:29:31-EDT
Received: by sun45.darpa.mil (4.1/5.51)
	id AA01794; Thu, 3 Aug 89 15:29:32 EDT
Date: Thu 3 Aug 89 15:29:31-EDT
From: William L. Scherlis <SCHERLIS@DARPA.MIL>
Subject: acks
To: SW-PI@vax.darpa.mil
Message-Id: <618175771.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>


Nicole Jacobs will return on Monday, at which point acknowledgements
should start to be sent out.  I believe, however, that Nicole has
already contacted every group who has not responded to the earlier
request for technical information.  Thanks,
				Bill
-------

∂03-Aug-89  1426	CLT 	Are you on Scherlis's list SW-PI?  
yes

∂03-Aug-89  1515	pi-data@vax.darpa.mil    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 3 Aug 89  15:15:21 PDT
Received: from sun49.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA01181>; Thu, 3 Aug 89 18:15:50 -0400
Date: Thu, 3 Aug 89 18:17:37 EDT
From: pi-data@DARPA.MIL
Posted-Date: Thu, 3 Aug 89 18:17:37 EDT
Message-Id: <8908032217.AA02889@sun49.darpa.mil>
Received: by sun49.darpa.mil (4.1/5.51)
	id AA02889; Thu, 3 Aug 89 18:17:37 EDT
To: jmc@sail.stanford.edu
Cc: pi-data@vax.darpa.mil

MEMO FOR PI OF TASK: QLISP FOR PARALLEL PROCESSORS
  CORRECT TASK CODE: 5G

The memo you received requesting financial data to support Fall
funding was automatically generated from a list of PIs, Tasks,
Task Codes and ISTO Program Managers.  Unfortunately, the PM for
your task was incorrectly listed as the Task Code as shown below:
 
   MEMO FOR PI OF TASK: QLISP FOR PARALLEL PROCESSORS
             TASK CODE: BOE  <---This was incorrect!!

This was our mistake.  If you responded with the 3 letter code for
PM, don't worry---we will take care of it after parsing your
response. Your real Task Code is "5G", and should be used in
the future if requested.  If you haven't responded yet, please
use the correct 2 letter Task Code.

Our apologies for any inconvenience as we develop this automated
system further.

∂03-Aug-89  1525	pi-data@vax.darpa.mil    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 3 Aug 89  15:25:42 PDT
Received: from sun49.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA01626>; Thu, 3 Aug 89 18:26:15 -0400
Date: Thu, 3 Aug 89 18:28:02 EDT
From: pi-data@DARPA.MIL
Posted-Date: Thu, 3 Aug 89 18:28:02 EDT
Message-Id: <8908032228.AA03195@sun49.darpa.mil>
Received: by sun49.darpa.mil (4.1/5.51)
	id AA03195; Thu, 3 Aug 89 18:28:02 EDT
To: jmc@sail.stanford.edu
Cc: pi-data@vax.darpa.mil

MEMO FOR PI OF TASK: FORMAL REASONING RESEARCH
  CORRECT TASK CODE: 4B

The memo you received requesting financial data to support Fall
funding was automatically generated from a list of PIs, Tasks,
Task Codes and ISTO Program Managers.  Unfortunately, the PM for
your task was incorrectly listed as the Task Code as shown below:
 
   MEMO FOR PI OF TASK: FORMAL REASONING RESEARCH
             TASK CODE: SIM  <---This was incorrect!!

This was our mistake.  If you responded with the 3 letter code for
PM, don't worry---we will take care of it after parsing your
response. Your real Task Code is "4B", and should be used in
the future if requested.  If you haven't responded yet, please
use the correct 2 letter Task Code.

Our apologies for any inconvenience as we develop this automated
system further.

∂03-Aug-89  1603	CLT 	How about pi-data?  
I'm not sure, but I got the msg about mixed up task codes.

∂03-Aug-89  1715	VAL  
It was the Ablex editor, not me, who placed Stanford in Palo Alto!

∂03-Aug-89  1828	RFC 	Prancing Pony Bill  
Prancing Pony bill of     JMC   John McCarthy         3 August 1989

Previous Balance             4.00
Monthly Interest at  1.0%    0.04
Current Charges              4.00  (bicycle lockers)
                             0.30  (vending machine)
                           -------
TOTAL AMOUNT DUE             8.34


PAYMENT DELIVERY LOCATION: CSD Receptionist.

Make checks payable to:  STANFORD UNIVERSITY.
Please deliver payments to the Computer Science Dept receptionist, Jacks Hall.
To ensure proper crediting, please include your PONY ACCOUNT NAME on your check.

Note: The recording of a payment takes up to three weeks after the payment is
made, but never beyond the next billing date.  Please allow for this delay.

Bills are payable upon presentation.  Interest of  1.0% per month will be
charged on balances remaining unpaid 25 days after bill date above.

An account with a credit balance earns interest of  .33% per month,
based on the average daily balance.

Your last Pony payment was recorded on 6/22/89.

Accounts with balances remaining unpaid for more than 55 days are
considered delinquent and are subject to reduction of credit limit.
Please pay your bill and keep your account current.

∂04-Aug-89  0835	deyoung@vax.darpa.mil 	RP3 AVAILABILITY VIA NET   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 4 Aug 89  08:35:47 PDT
Received: from sun38.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA04073>; Fri, 4 Aug 89 10:57:28 -0400
Posted-Date: Fri 4 Aug 89 10:54:35-EDT
Received: by sun38.darpa.mil (4.1/5.51)
	id AA10763; Fri, 4 Aug 89 10:54:37 EDT
Date: Fri 4 Aug 89 10:54:35-EDT
From: Tice DeYoung <DEYOUNG@DARPA.MIL>
Subject: RP3 AVAILABILITY VIA NET
To: ISTO-PI-LIST@vax.darpa.mil
Cc: Steve@hubcap.clemson.edu, wjk@ibm.com
Message-Id: <618245675.0.DEYOUNG@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>


************************************************************************

     The IBM Research Division  and the Defense Advanced Research
     Projects Agency  are announcing that access  to the Research
     Parallel Processor Prototype (RP3) is being offered to qual-
     ified universities,  government laboratories and  commercial
     research organizations interested  in studying shared-memory
     parallel systems.

     To gain access to the  machine, a brief (1500 word) proposal
     describing the  application(s) to  be studied  and outlining
     the experimentation to  be done must be submitted  to IBM at
     the address listed below.   Proposals will be accepted based
     on the following criteria:

     -   technical merit  and novelty  of the techniques  and ap-
         proaches to be used
     -   expertise of the proposer in the application field
     -   potential  benefit  to  research  and  understanding  of
         shared-memory parallel processing
     -   non duplication of methods or approaches already studied
     -   adaptability of  each  application for  execution on the
         RP3 as shown by its ability to execute, in skeletal form,
         on an IBM RT/PC.

     IBM and  the experimenter  will share  the results  of their
     joint experiments, including programming techniques or meth-
     ods exposed  or developed  during their experiments  on RP3.
     DARPA  will  participate in the selection  of all university
     related research activities that use the RP3, with final se-
     lection to be made by IBM.  Brief monthly status reports and
     a final report of results obtained will be required.

     The  IBM Research  Parallel Processor  Prototype (RP3)  is a
     prototype highly  parallel processor intended to  be used to
     gain   insight   into   most  aspects   of   closely-coupled
     shared-memory system organizations.  The RP3 has a rich mem-
     ory hierarchy including cache,  local and global addressing,
     with interleaving  and address hashing available  for global
     memory  references.  The  ratio of  access times  for global
     memory references to  local references is 1.6:1  on a moder-
     ately loaded  interconnection network.  All  partitioning of
     memory  into its  various forms  is controlled  by software.
     The RP3  also has  built in Performance  Measuring circuitry
     capable of nearly unobtrusive measurement of many system ac-
     tivities.

     A modified  version of Carnegie Mellon  University's Mach is
     the operating system for the  RP3.  Mach presents a familiar
     UNIX*  interface to  applications  programmers.  Two  exper-
     imental application  run-time environments are  available to
     researchers at this time.   These environments will be modi-
     fied and improved as further  knowledge of their behavior on
     a machine of this magnitude is understood.

     Proposals for use of the machine (electronic mail is accept-
     able) should be sent to:

               Walter J. Kleinfelder
               IBM Corporation
               T. J. Watson Research Center/Hawthorne (H4-D32)
               P.O. Box 704
               Yorktown Heights, NY 10598

               (914)789-7205
               wjk@ibm.com

     _______________
     *   UNIX is a registered trademark of AT&T Bell Laboratories
************************************************************************

The RP3 was developed under SPAWAR Contract Number N00039-87-C-0122.
The MACH OS was developed under SPAWAR Contract Number N00039-85-C-0134.

-------

∂04-Aug-89  1118	scherlis@vax.darpa.mil 	Important Announcement    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 4 Aug 89  11:18:26 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA04794>; Fri, 4 Aug 89 14:09:58 -0400
Posted-Date: Fri 4 Aug 89 14:00:54-EDT
Received: by sun45.darpa.mil (4.1/5.51)
	id AA02442; Fri, 4 Aug 89 14:00:56 EDT
Date: Fri 4 Aug 89 14:00:54-EDT
From: William L. Scherlis <SCHERLIS@DARPA.MIL>
Subject: Important Announcement
To: SW-PI@vax.darpa.mil
Cc: toole@vax.darpa.mil, boesch@vax.darpa.mil, sowa@vax.darpa.mil,
        schwartz@vax.darpa.mil
Message-Id: <618256854.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>



	We are pleased to announce the appointment of Barry 
	Boehm as ISTO Director effective in November.  Plans 
	for the transition will be announced over the next few 
	weeks.  Most of you know Barry's extensive work on the
	problems of very large-scale software systems development. 
	Barry has recently joined the faculty at UCLA after a 
	long and distinguished career at TRW.
					Bill



-------

∂04-Aug-89  2039	ME 	failed mail returned      
The following message was undeliverable to the address(es) below
because the destination Host Name(s) are Unknown:
siegman@sierra.UUCP

[Use "sierra".]

------- Begin undelivered message: -------
 04-Aug-89  2032	Mailer 	re: When should I call JMC racist? [was Re: black doctors]    
To:   siegman@sierra.UUCP, su-etc@SAIL.Stanford.EDU  
From: John McCarthy <JMC@SAIL.Stanford.EDU>

[In reply to message from siegman@sierra.UUCP sent 5 Aug 89 00:20:27 GMT.]

NSF did a study some years ago of the subsequent performance of people being
considered for NSF fellowships in mathematics.  The only clear correlation
was with age - the younger the better.

This wasn't convincing to the Admissions Committee of the Computer Science
Department this year, which turned down all the very young candidates.

------- End undelivered message -------

∂04-Aug-89  2105	CLT 	tommorrow night
Don't forget dinner at Richard's with Mrs. Takasu -
We should leave around 7.

∂05-Aug-89  0537	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	oopsilon 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 5 Aug 89  05:36:52 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 407670; 5 Aug 89 08:35:55 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 105660; Sat 5-Aug-89 05:29:31 PDT
Date: Sat, 5 Aug 89 05:29 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: oopsilon
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890724094644.7.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890805122905.8.RWG@TSUNAMI.SPA.Symbolics.COM>

(The singular of Uppsala?)
    . . .
    Salamin reminds me of the bogus cake-taker "Riemann's constant":

      sum  mu(n)/sqrt(n),
      n>0

    where mu is the Moebius function:  0 if a square divides n,
    (-1)↑# of prime factors(n) otherwise.

    This series converges iff the zeta hypothesis!  However, if
    it *does* converge, we can easily guess to what!  1/zeta(1/2)
     ~~ -.684765 .

    . . .

Igor, Ilan, and Salamin all corrected me on this.  According to
Igor, convergence requires the indefinite sum of mu(n) be o(sqrt(n))
(i.e. nicer than a random walk), and Titchmarsh sez it ain't.  So
the equivalence is probably with the convergence of

      sum  mu(n)/n↑(1/2+epsilon),
      n>0

for all positive epsilon.  Also, it would not have been safe to bet
on 1/zeta(1/2) at epsilon=0 anyway.  E.g., Salamin points out that,
if
                         inf
                         ====   2 n - 1    4 n - 2    4 n
                         \     x          x          x
               f(x) :=    >    -------- + -------- + ----
                         /     2 n - 1    4 n - 2    4 n
                         ====
                         n = 1

then f(x) = - ln (1-x) for x in (-1,1), yet f(-1) = -(ln 2)/2;  and
at MIT I showed one only slightly weirder that glitched at both ends
of its convergence interval!

∂05-Aug-89  1304	danvy@gang-of-four.Stanford.EDU 	your science-fiction story 
Received: from gang-of-four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Aug 89  13:04:43 PDT
Received:  by gang-of-four.Stanford.EDU (5.0/25-eef) id AA25938; Sat, 5 Aug 89 13:05:24 PDT
Date: Sat, 5 Aug 89 13:05:24 PDT
From: Olivier Danvy <danvy@gang-of-four.Stanford.EDU>
Message-Id: <8908052005.AA25938@gang-of-four.Stanford.EDU>
To: jmc@sail
Subject: your science-fiction story
Cc: danvy@gang-of-four.stanford.edu


	The very reason that extra-terrestrials are detected because
	they start their car 0.5 sec. after the light is green
	calls a couple of comments.

	(1) Who should infer that they are extra-terrestrial?
	And how?  Either the artificial intelligence is able to deduce it
	but then it is not only programmed for regulating lights 
	at cross-roads;
	or a human deduces it after reading the report of the anomaly.
	Then how does the artificial intelligence realize this point?
	Possibility: in the electronic notes (diary?) the human has taken,
	or in some mail he writes to someone; then say that the human
	is killed precisely at a cross-road, to inform the artificial
	intelligence.  However it makes the artificial intelligence
	not only dedicated to regulating cross-roads and book-keeping
	statistics.
	To add some more uncertainty to the end, let us say that the extra-
	terrestrials somehow have abused the sensors of the computer
	to make the zapping look like a regular accident; this would add
	another dimension to the story: doubt.  The artificial intelligence
	would doubt the accuracy of its input, maybe even of its program,
	and eventually of him/itself as an intelligence.
	Maybe there could be extra-terrestrial programmers,
	and for the sake of thrilling, at the 2/3 of the novel,
	the new programmer is an actual extra-terrestrial.
	This in order to make the way out of the novel absolutely thin.

	However, this is another line than yours, which more mechanically
	stick to the fact that there is no such routine in the program
	that takes care of humanity.

	Still this suggests some other issues:
	- the artificial intelligence is actually a specialized version
	of a more general program (cf. inheritance, fatherhood,
	though they are quite anthropocentric);
	- the person the human informs before being killed (zapped or
	in an accident at a crossroad) is either a very close relative,
	say his girlfriend; this would enforce the human side of the story
	with some simple suspense;
	- there could be a chain of persons that know-and-die;
	this suggests an architecture where all the even chapters
	are told eg, subjectively, by an human, until he/she is zapped
	(though this could become expectable and thus monotonous);
	and all the odd chapters (subjectively too) by the artificial
	intelligence that surveys what is going on.
	Still more on the human side: kids, or Zen masters would be
	partly immune to mind-reading either because their mind is not
	developed enough (but then what can kids do: as soon as they grow
	up either they forget (a bit like the religion of their childhood)
	or they are zapped; and Zen masters are too obvious and somehow
	foreign too).
	And again this is too much on the human side.

	However, in parallel to the even chapters, the odd chapters
	could be subjective to -different- artificial intelligences,
	which would add up to the spaghetti aspect of the novel,
	and point out specific communication problems (for a simple example,
	cf. "Who can replace a human", Brian Aldiss, in "Equator", 1964).


	(2) The 0.5 sec. after the light is green make the story possible only
	in countries where the light goes straight from red to green
	(USA, I guess, and France, etc.) in contrast to countries
	where the light becomes orange before getting green, as a symmetric
	warning: red-orange-green-orange-red-orange-green-... (UK, Denmark,
	Sweden, etc.).


	(3) The decisive factor that would save humanity could be
	delayed until the whole humanity is nominally lost,
	and when somehow the very existence of the computers
	is endangered.  But it has the human bias of the "struggle for life".


	(4) Another possible end is to conclude on a very thin chance
	to save humanity.


	(5) Another idea would be to intermingle chapters that are subjective
	to the extra-terrestrials.  However wouldn't it consume the loss
	of your initial idea: tell a humanly-interesting story from
	an artificial intelligence point of view, where the extra-terrestrial
	are only more foreign than the terrestrials to the artificial
	intelligence, and humanity is only saved because there was
	some mechanical humor in the programming, which is interpreted
	in a non-standard way by the program.


	Well, it is an interesting story and it was fun to hear it
	from you and to juggle with it.  Maybe you will find some interest
	to these handful of considerations.

		Regards,			Olivier

∂05-Aug-89  1310	danvy@gang-of-four.Stanford.EDU 	By the way  
Received: from gang-of-four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Aug 89  13:10:19 PDT
Received:  by gang-of-four.Stanford.EDU (5.0/25-eef) id AA25950; Sat, 5 Aug 89 13:11:01 PDT
Date: Sat, 5 Aug 89 13:11:01 PDT
From: Olivier Danvy <danvy@gang-of-four.Stanford.EDU>
Message-Id: <8908052011.AA25950@gang-of-four.Stanford.EDU>
To: jmc@sail
Subject: By the way
Cc: danvy@gang-of-four.Stanford.EDU


	By the way it is "The Two Sides of Tomorrow"
	that was inspired to James Hogan by Marvin Minsky,
	on the idea that "it was time someone write an intelligent
	SF story about computers".  James Hogan lives in California today.

					Olivier

∂05-Aug-89  1916	jim@thrush.STANFORD.EDU 	Re: Blacks and Jews 
Received: from thrush.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 5 Aug 89  19:15:59 PDT
Received: by thrush.STANFORD.EDU (3.2/4.7); Sat, 5 Aug 89 19:14:52 PDT
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: Blacks and Jews 
In-Reply-To: Your message of 04 Aug 89 17:08:00 PDT.
             <NHu4S@SAIL.Stanford.EDU> 
Date: Sat, 05 Aug 89 19:14:50 PDT
From: Jim Helman <jim@thrush.STANFORD.EDU>

My posting (19-Jul-89) must have been delayed quite a bit somewhere.

Unfortunately, I no longer have the article.  But I believe there was
only the one quotation regarding Jewish immigrants.  It was
represented as a typical opinion of the time, but may just have been
some anti-semitic spouting: "see, they're even worse than the blacks."
But stupidity was not the main thrust of the quote, rather it
portrayed them as dirty and lazy ne'er-do-wells.

-jim

∂06-Aug-89  0716	weening@gang-of-four.Stanford.EDU 	Offices   
Received: from gang-of-four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 6 Aug 89  07:16:35 PDT
Received:  by gang-of-four.Stanford.EDU (5.0/25-eef) id AA27468; Sun, 6 Aug 89 07:17:16 PDT
Date: Sun, 6 Aug 89 07:17:16 PDT
From: Joe Weening <weening@gang-of-four.Stanford.EDU>
Message-Id: <8908061417.AA27468@gang-of-four.Stanford.EDU>
To: jmc@sail, clt@sail
Subject: Offices

The office space committee has asked me whether I want to move Dan
into 360 this fall, which would free up 353 as an all-student office.
I'm in favor of the move, since I expect to work a lot with Dan on
Qlisp work.  Do you have any objections to this?  I don't know which
students they would want to move into 353.

∂06-Aug-89  1553	rpg@lucid.com 	Meet 
Received: from lucid.com by SAIL.Stanford.EDU with TCP; 6 Aug 89  15:53:27 PDT
Received: from challenger ([192.9.200.17]) by heavens-gate id AA01228g; Sun, 6 Aug 89 15:50:22 PDT
Received: by challenger id AA05499g; Sun, 6 Aug 89 15:51:39 PDT
Date: Sun, 6 Aug 89 15:51:39 PDT
From: Richard P. Gabriel <rpg@lucid.com>
Message-Id: <8908062251.AA05499@challenger>
To: jmc@sail.stanford.edu
Subject: Meet


I'm now off the hook with respect to X3J13 stuff (at least for a
while). Can we meet Tuesday late morning or early afternoon to discuss
CPL? Also, I want to bring over Sexton and see whether we can finally
hire him. I think I'm pretty on trackto be working on CPL nearly full
time starting Sept 1.

			-rpg-

∂06-Aug-89  1605	ME 	DD display lossage   
 ∂05-Aug-89  1813	JMC 	(on TTY63, at TV-140) display losing    
To:   BUG-E  
On datadisc lines are often garbled.  However, when the line
is visited, i.e. the pointer is moved to it, the display clears
up.  It remains clear when the pointer is moved away from the
line.  Thus the problem seems distinct from earlier display
lossages.

ME - I've wiggled the usual suspect boards in the DD and this problem
seems to have gone away again.

∂07-Aug-89  0922	rpg@lucid.com 	Meet      
Received: from lucid.com by SAIL.Stanford.EDU with TCP; 7 Aug 89  09:21:06 PDT
Received: from challenger ([192.9.200.17]) by heavens-gate id AA01054g; Mon, 7 Aug 89 09:18:00 PDT
Received: by challenger id AA05904g; Mon, 7 Aug 89 09:19:14 PDT
Date: Mon, 7 Aug 89 09:19:14 PDT
From: Richard P. Gabriel <rpg@lucid.com>
Message-Id: <8908071619.AA05904@challenger>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 06 Aug 89  1612 PDT <fItuA@SAIL.Stanford.EDU>
Subject: Meet  


ok, 11 then lunch.

∂07-Aug-89  1000	JMC  
Call about postponing meeting with Dean.

∂07-Aug-89  1033	nilsson@Tenaya.Stanford.EDU 	re: cs306  
Received: from Tenaya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Aug 89  10:33:39 PDT
Received:  by Tenaya.Stanford.EDU (NeXT-0.8/25-eef) id AA02354; Mon, 7 Aug 89 10:32:41 PDT
Date: Mon, 7 Aug 89 10:32:41 PDT
From: Nils Nilsson <nilsson@tenaya.stanford.edu>
Message-Id: <8908071732.AA02354@Tenaya.Stanford.EDU>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: cs306
Cc: stager@score.stanford.edu, nilsson@score.stanford.edu

John,

I presume you are working directly with Roy Jones
and Clare Stager on the matter of who is teaching
cs 306.  Shankar certainly did get a reasonably good
TBP evaluation.

-Nils

∂07-Aug-89  1049	RPG 	Grumble   
I saw this list of ``eminent'' graduates. I daresay I'm more eminent than
Mark Stefik, Randy Davis, Reid Smith, Bill Clancey and a whole raft of others.

			-rpg-

∂07-Aug-89  1538	JONES@vaxb.acs.unt.edu 	researchers in logical reasoning    
Received: from vaxb.acs.unt.edu ([192.12.10.4]) by SAIL.Stanford.EDU with TCP; 7 Aug 89  15:37:37 PDT
Date: Mon, 7 Aug 89 17:38 CDT
From: JONES@vaxb.acs.unt.edu
Subject: researchers in logical reasoning
To: jmc@sail.stanford.edu
X-VMS-To: IN%"jmc@sail.stanford.edu"

Dear Dr. McCarthy,

	Dr. Nilsson responded to some questions of mine indicating that
you would be in a better position to provide answers.  Essentially, I
am in the process of identifying appropriate schools where I can pursue
Ph.d. studies.  My area of interest is logical reasoning.  (I am interested
in several issues and techniques related to the development of logical reasoning
systems.  Several such techniques were covered in Dr. Nilsson's book, Logical
Foundations of Artifical Intelligence.)
	My questions are:  
		1)  who are the top researchers in this field, and at what
                    schools do they teach?
	    and 2)  are you familiar with the work of Dr. Andrew Whinston
                    at the University of Texas (formerly at Purdue)?  Would
                    he be included in this list?  (I ask about him because
                    he indicated to me that he was doing this type of 
                    research, yet I do not find any of his publications in
                    the major journals dealing with AI.)
	I look forward to hearing from you, and will deeply appreciate
any help/advice you may be willing to offer.

                                    Sincerely,
                                    James Jones
                                    Jones@vaxb.acs.unt.edu

∂07-Aug-89  1623	PAF 	Email, Email, wherefor art thou, Email? 
John -

Rather than be lumped into that category of Engineers designing protocols
for future networks who ignored you, I'd like to go over some of the
points you've made -- in person -- and rebut a few (hopefully).

Pick a time and place.

-=paulf

∂07-Aug-89  1640	gregory@Polya.Stanford.EDU 	Re: suggestions solicited  
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Aug 89  16:39:58 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA15811; Mon, 7 Aug 89 16:40:39 -0700
Date: Mon, 7 Aug 89 16:40:39 -0700
From: Gregory R. Whitehead <gregory@Polya.Stanford.EDU>
Message-Id: <8908072340.AA15811@Polya.Stanford.EDU>
To: JMC@SAIL.STANFORD.EDU
Subject: Re: suggestions solicited
Newsgroups: su.computers
In-Reply-To: <1NIyz8@SAIL.Stanford.EDU>
Organization: Stanford University
Cc: 

In article <1NIyz8@SAIL.Stanford.EDU> you write:

                     |------- typo? -------|
>Presumably the same kinds of modems dialers that are used for fax would
>be appropriate but would give better transmission speeds.

Looks good otherwise.  Best of luck getting it published.

-Greg


∂07-Aug-89  2020	PAF 	email
Okay, see you at 2pm in your office on Tuesday...

-=paulf

∂08-Aug-89  0948	WILBER@Score.Stanford.EDU 	Re: suggestions solicited   
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  09:48:32 PDT
Date: Tue 8 Aug 89 09:47:53-PDT
From: Mike Wilber <Wilber@Score.Stanford.EDU>
Subject: Re: suggestions solicited 
To: JMC@Sail.Stanford.EDU
In-Reply-To: <1NIyz8@SAIL.Stanford.EDU>
Affiliation: Rockwell Palo Alto Laboratory
USPS: 444 High Street, Palo Alto, California 94301
Voice: (415)325-7158
Message-ID: <12516510904.9.WILBER@Score.Stanford.EDU>

in your message <1NIyz8@SAIL.Stanford.EDU> of 07 Aug 89  1536 PDT, you write

	    Electronic mail could work the same way, but because of a
    mistake by DARPA about 1970, i.e. making a special-purpose,
    special-politics network the main vehicle for electronic mail, it
    was combined with other network uses that require higher
    bandwith.

actually, the way i remember it is that inter-host mail started in about 1972
as a bbn hack (sndmsg, by name) that was piggybacked onto ftp...what i remember
of darpa's vision statements addressed remote procedure call and opening remote
files, with some occasional remote login...

	    There has been a proliferation of networks and message
    services on a variety of time-sharing utilities.  Some of them
    are commercial and some of them serve various scientific
    disciplines and commercial activities.  The connections between
    these networks require politics and often fail.  A whole industry
    is founded on the technologically unsound ideas of competitive
    special purpose networks and storage of mail on mail computers.
    It is as though there were dozens of special purpose telephone
    networks and no general network.

hear! hear!...

	    Fax has another advantage that needs to be matched.
    Since fax transmits images, fully formatted documents can be
    transmitted.  However, this loses the ability to edit the
    document.  This can be beaten by email, provided there arises a
    widely used standard for representing documents that preserves
    editability.

well, sort of...on the one hand, we might be able to go a long way toward
editing faxed documents if optical character recognition technology could
convert them to character streams (i'm unfamiliar with the current state
of that art, so i don't know how close we are)...on the other hand, that
technology could also be used in conjunction with fax to compress documents
for transmission and storage...(e.g., a protocol might permit one to say,
"here is the text of the document, and here is the difference between the
obvious way to lay it out on the page and the way it's actually laid out
on the page")...

fax also has an advantage of immediacy: you can write things with a pencil
on a piece of paper wherever you are with no special-purpose equipment and
later slap it into the fax and go about your business...while the advantage
is minimal for commercial documents, it could be considerable for personal
communication among the common people...the portability and accessibility
of lap-top computers diminishes this advantage, but they're not cheap (yet,
though they're becoming cheaper), so i'd guess the problem is too urgent
for us to pin our hopes on them...


	    It would be good if the ACM were to set up a committee
    to adopt a telephone electronic mail standard.  However, I fear
    the vested interests would be too strong, and the idea would
    die from being loaded with requirements for features that
    would be too expensive to realize in the near future.

	    Fortunately, there is free enterprise.
    Therefore, the most likely way of getting direct
    electronic mail is for some company to offer a piece of hardware
    as an electronic mail terminal including the facilities for
    connecting to the current variety of local area networks (LANs).
    The most likely way for this to be accomplished is for the makers
    of fax machines to offer ASCII service as well.  This will
    obviate the growing practice of some users of fax of printing out
    their messages in an OCR font, transmitting them by fax,
    whereupon the receiver scans them with an OCR scanner to get them
    back into computer form.

well, as you said in the previous paragraph, it's a competitive world...what
do you suppose might serve as a sufficient inducement to a manufacturer to
provide such a product?...i can think of two opposing trends right here in
the computer market that might be useful in fixing thought: on the one hand,
dec has always pushed vms and deemphasized ultrix (e.g., their long delay in
introducing it in the first place)...the conventional wisdom is that their
motive is to lock customers in to their hardware...on the other hand sun is
hard at work promoting standards (their own (sparc, nfs) and extant ones
(ip/tcp, unix)), ostensibly to expand the size of the market they're selling
into...


btw, have you heard of dasnet?...according to a message of 7 Jul 89 21:15:23
GMT from morgan@Jessica.stanford.edu (RL "Bob" Morgan) to
su-computers@shelby.stanford.edu (of which i can forward you a copy if you
wish), da systems offers a connection between "...more than 20 systems and 
networks, INCLUDING:

  ATT Mail, BIX(TM), DASnet(R) Network, DCMETA, Dialcom(SM), EIES,
  EasyLink(R), Envoy 100(TM), FAX, GeoMail, INET(SM), MCI Mail(R),
  NWI(R), PeaceNet/EcoNet, Portal Communications(TM), The Source(R),
  Telemail(R), ATI's Telemail (Japan), Telex, TWICS (Japan),
  UNISON(R), UUCP, The WELL, Domains e.g. ".COM" and ".EDU".   New
  systems are added all the time."

...the service costs $4.75/mo plus a charge for the volume of traffic,
e.g., peacenet, unison, uucp, .com, .edu: $.10 + .11/1k char, us fax (?!):
$.42 + .37/1kchar, according to that message...

thank you very much for writing this piece...i think you're making some very
important points on a very important issue...i wish you the best of luck in
your endeavor...

-- mike wilber
-------

∂08-Aug-89  1003	P.REDLICH@GSB-WHY.Stanford.EDU 	suggestion   
Received: from GSB-WHY.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  10:02:59 PDT
Date: Tue 8 Aug 89 09:59:35-PDT
From: Warren Redlich <P.REDLICH@GSB-WHY.Stanford.EDU>
Subject: suggestion
To: jmc@Sail.Stanford.EDU
Message-ID: <12516513033.145.P.REDLICH@GSB-WHY.Stanford.EDU>

My suggestion is for political success, not for editing purposes.

I would send a copy of this to Tom Campbell, along with a summary
(Congressmen and their aides are short on time), and letters or
signatures from important people in the area, such as executives
of Silicon Valley companies (David Packard would be a doozy), and
maybe from other important university personnel.

This may have nothing to do with Congress, but Congressmen have a
lot of influence, and they want to serve their constituents.  As 
Silicon Valley probably has the highest use of E-mail in the
country (probably the world), Campbell ought to be particularly
interested.

You might also want to send it to California senators and to
other congressmen in the area.  There might also be similar
interests near Boston (on route 128) and in Austin, as both
have junior Silicon Valleys.

From my perspective, I would never have expensive phone calls
to people who have computers.  A lot of people would probably
appreciate this.

Warren
-------

∂08-Aug-89  1010	P.REDLICH@GSB-WHY.Stanford.EDU 	How do you . . .  
Received: from GSB-WHY.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  10:10:00 PDT
Date: Tue 8 Aug 89 10:06:30-PDT
From: Warren Redlich <P.REDLICH@GSB-WHY.Stanford.EDU>
Subject: How do you . . .
To: jmc@Sail.Stanford.EDU
Message-ID: <12516514293.145.P.REDLICH@GSB-WHY.Stanford.EDU>

How do you get those stories from services like AP?  Is this something
that is common knowledge with common access, or do you have some
special trick or password?

Warren
-------

∂08-Aug-89  1012	wls@nova.npac.syr.edu 	Re:  RP3 AVAILABILITY VIA NET   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 8 Aug 89  10:12:20 PDT
Posted-Date: Tue, 8 Aug 89 12:32:56 EDT
Received: from nova-client.npac.syr.EDU by vax.darpa.mil (5.61/5.61+local)
	id <AA10870>; Tue, 8 Aug 89 12:31:10 -0400
Date: Tue, 8 Aug 89 12:32:56 EDT
From: wls@nova.npac.syr.edu (William L. Schrader)
Received: by nova.npac.syr.edu (4.0/2.1-Northeast Parallel Architectures Center)
	id AA10884; Tue, 8 Aug 89 12:32:56 EDT
Message-Id: <8908081632.AA10884@nova.npac.syr.edu>
To: DEYOUNG@vax.darpa.mil, ISTO-PI-LIST@vax.darpa.mil
Subject: Re:  RP3 AVAILABILITY VIA NET
Cc: Steve@hubcap.clemson.edu, ecs@nova.npac.syr.edu, wjk@ibm.com,
        wls@nova.npac.syr.edu

Tice,
  We are seeing LOTS of interest in the RP3. May we advertise,
so to speak, and encourage lots of users?
Bill

∂08-Aug-89  1031	@NIC.DDN.MIL:sol@NIC.DDN.MIL 	re: your su.computer message  
Received: from NIC.DDN.MIL by SAIL.Stanford.EDU with TCP; 8 Aug 89  10:31:42 PDT
Received: from fs4.NISC.SRI.COM by NIC.DDN.MIL with TCP; Tue, 8 Aug 89 10:25:23 PDT
Received: by fs4.NISC.SRI.COM (4.0/SMI-4.0/LIM-3.2)
	id AA09945; Tue, 8 Aug 89 10:22:21 PDT
Date: Tue, 8 Aug 1989 10:22:20 PDT
From: Sol Lederman <sol@NIC.DDN.MIL>
To: jmc@sail.stanford.edu
Subject: re: your su.computer message
Message-Id: <CMM.0.88.618600140.sol@fs4.NISC.SRI.COM>

Mr. McCarthy,

I enjoyed your message and agree with your views about how e-mail must be
made simpler if it is to survive, especially among the masses. One tiny
gripe: I believe that "ARPANET" is the correct way to "case" the network
name, even though the "NET" part is not an acronym. Two years of experience
working at the Network Information Center at SRI leads me to this belief.

Sol

∂08-Aug-89  1116	nilsson@Tenaya.Stanford.EDU 	re: electronic mail  
Received: from Tenaya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  11:16:09 PDT
Received:  by Tenaya.Stanford.EDU (NeXT-0.8/25-eef) id AA03159; Tue, 8 Aug 89 11:15:11 PDT
Date: Tue, 8 Aug 89 11:15:11 PDT
From: Nils Nilsson <nilsson@tenaya.stanford.edu>
Message-Id: <8908081815.AA03159@Tenaya.Stanford.EDU>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: electronic mail
Cc: nilsson@Tenaya.Stanford.EDU

I like your piece on e-mail and agree wholeheartedly
with it.  How about sending a copy of it to Steve
Jobs (steve_jobs@next.com) or some other entrepreneurs who might be
interested in pursuing the free-enterprise approach
to the problem?  (The NeXT machine is supposed to
someday have a built-in modem/fax service that makes
use of the DSP board.)  Also, what do people like Ralph
Goren and Bill Yundt (and Bob Street) say about it?
If they were in favor, maybe Stanford could play some
kind of official role in pushing for this.  How about
implementing an experimental version around Stanford
using students and others who have computers
and modems in their offices and homes?

I have some friends at EDUCOM. Do you want me
to send them an on-line copy (over the harmful
regular network)?  If so, send me an online copy.

-Nils

∂08-Aug-89  1217	GOTELLI@Score.Stanford.EDU 	Decommissioning of SCORE ***PLEASE READ***
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  12:17:24 PDT
Date: Tue 8 Aug 89 11:52:44-PDT
From: Lynn Gotelli <GOTELLI@Score.Stanford.EDU>
Subject: Decommissioning of SCORE ***PLEASE READ***
To: MCCARTHY@Score.Stanford.EDU
Message-ID: <12516533631.9.GOTELLI@Score.Stanford.EDU>

SCORE is being decommissioned and removed from the CSD-CF Cost Center.
As of September 1st, all users must migrate off of SCORE.

At your request, an account will be opened for you on a different
machine.  This machine will be a UNIX machine (not a machine running
the TOPS-20 operating system).  Note that the commands and programs
available on a UNIX machine will differ from those you are familiar
with on SCORE.  New-account forms are available at the CSD receptionist's
desk (2nd floor, Margaret Jacks Hall).

Please make arrangements soon to move your files off of SCORE.  As
of September 1st, your SCORE account and files will be unavailable
to you.

Please contact George Wheaton (wheaton@athena, 723-5396) if you have
any questions.

Thank you.


-------

∂08-Aug-89  1334	RPG 	I forgot to mention this 
If there were some assurances or even relatively convincing
strong statements that it was something that could happen within
a predictable period of time, that I could be made Professor of Research,
I would almost certainly join Stanford full time immediately and permanently.

			-rpg-

∂08-Aug-89  1614	dupre@csli.Stanford.EDU 	kinds
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  16:14:42 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA04917; Tue, 8 Aug 89 16:15:46 PDT
Date: Tue 8 Aug 89 16:15:45-PDT
From: John Dupre <DUPRE@CSLI.Stanford.EDU>
Subject: kinds
To: jmc@sail.stanford.edu
Message-Id: <618621345.0.DUPRE@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>


The person whose name I forgot was Tyler Burge, and the article that might be
of interest to you is called "Individualism and Psychology", and is in 
Philosophical Review, Jan. 1986.
I'll try and think of anything else that I know that might be useful.

John Dupre
-------

∂08-Aug-89  1816	M.MARTY@Macbeth.Stanford.EDU 	Re: abstract   
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  18:16:51 PDT
Date: Tue 8 Aug 89 18:14:54-PDT
From: martin ross <M.MARTY@Macbeth.Stanford.EDU>
Subject: Re: abstract  
To: JMC@SAIL.Stanford.EDU
In-Reply-To: <1NHrcT@SAIL.Stanford.EDU>
Message-ID: <12516603204.77.M.MARTY@Macbeth.Stanford.EDU>


Frankly, I had no idea what was wanted in a thesis abstract. I'm 
planning my talk to be for a general mathematical audience, so
I certainly think lots of questions will be desirable. To answer briefly
the questions in your two messages:
1. An immersed surface is allowed to run into itself and an
  embedded surface is not. Have you seen pictures of Klein bottles?
  They are immersed.
2. A gr
  By a graph in 3-space, I just mean the set of points (x,y,f(x,y))
  where f(x,y) is a function of two real variables. To say the
  graph is complete means the domain of the function is the
  whole x-y plane.
3. Technically, a minimal graph is one where the function f(x,y)
    satisfies a certain partial differential equation. The
graph, as a set, satisfies the first derivative test to
minimize area (in the sense of calculus of variations). The classical
examples of minimal surfaces are soap films. The start of my talk
wil be a brief introduction to minimal surfaces.

   I hope this is some help. I'm sorry i took  time to answer
  your messages- I don't check my mail that often. I'd be
  happy to answer more questions. I wasn't expecting to
  have an interested chairman.

     Marty Ross

-------

∂08-Aug-89  1826	levinth@sierra.STANFORD.EDU 	problem with permission to hire someone  
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  18:26:00 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Tue, 8 Aug 89 18:25:05 PDT
Date: Tue, 8 Aug 89 18:25:05 PDT
From: levinth@sierra.STANFORD.EDU (Elliott C. Levinthal)
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 08 Aug 89  1129 PDT <1NJp60@SAIL.Stanford.EDU>
Subject: problem with permission to hire someone 

John, I will call you tomorrow on this matter. With whom have you been
dealing on these appointments. Are these Research Associate
appointments?
Elliott

∂08-Aug-89  2000	JMC  
Consumer alert address

∂08-Aug-89  2234	gandalf@csli.Stanford.EDU 	Re: Networks considered harmful - for electronic mail
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Aug 89  22:10:51 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA14061; Tue, 8 Aug 89 22:11:55 PDT
Date: Tue, 8 Aug 89 22:11:55 PDT
From: gandalf@csli.Stanford.EDU (Juergen Wagner)
Message-Id: <8908090511.AA14061@csli.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: Re: Networks considered harmful - for electronic mail

I have read your article and would like to make a few comments:

o  Clearly, a scheme which doesn't require to memorize paths is better than
   one which relies on users when it comes to routing. I think, you should
   mention the fact that in such cases, routing must either be evident from
   the address (like with XNS "clearinghouses" or Grapevine "registries") or
   be done by means of a distributed routing database (as done by Internet
   domain name services).

   Both schemata can be implemented with mere phone numbers (country code,
   area code, number, ...) but in my opinion, a symbolic approach would still
   be preferrable.

   Your arguments against complicated mail paths do not necessarily call for
   abandoning the currect method of name lookups and heterogeneous networks
   which are accessible from each other by means of gateways, they simply
   ask for more intelligent routing (and possibly naming conventions).

o  It seems to me as if you used the term "electronic mail" for "electronic
   mail using UUCP or TCP/IP". I think, the article is lacking the distinction
   between the protocol used by the transport medium and the protocol used by
   the mail/message transfer agents. It is possible to run TCP/IP or XNS over
   standard phone lines, and there is a variety of host architectures with
   quite different operating system types which use X.400.

o  UUCP is a protocol which (as the name states) was developed for UNIX
   to UNIX transfers. Therefore, the commands operate on UNIX-like objects.
   I do not see why this is a bad idea. After all, UNIX is a wide-spread
   operating system which seemed to provide a good model for the specific
   type of message exchange handled by UUCP.

o  The problem of FAX is that it doesn't transmit structured documents. FAX
   transfers images. "E-mail" (which I take to be a term describing at least
   byte-structured messages) is by definition better suited for this job.
   Combining the two would result in a structured document transmission
   protocol, FAX would only be part of.

Having dealt with electronic mail for quite some time, and having installed
Sendmail on several machines myself, I entirely agree with the main point of
your posting: the electronic transmission of messages must become easier and
simpler. However, I see protocols such as FAX only as a partial solution to
the overall problem.

Regards,
Juergen Wagner		   			gandalf@csli.stanford.edu
						 wagner@arisia.xerox.com

∂08-Aug-89  2323	LES 	re: Need opinions on company names.
[In reply to message sent 08 Aug 89 2108 PDT.]

Correct.

∂09-Aug-89  0705	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	an intriguing screw case for TAYLOR    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 9 Aug 89  07:04:59 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 408566; 9 Aug 89 10:02:59 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 105911; Wed 9-Aug-89 06:56:46 PDT
Date: Wed, 9 Aug 89 06:56 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: an intriguing screw case for TAYLOR
To: macsyma-i@RUSSIAN.SPA.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890809135615.2.RWG@TSUNAMI.SPA.Symbolics.COM>

Suppose you want to expand the lhs of
(C367) (PRODUCT(F(K),K,1,inf)*PRODUCT(1-P↑K/F(K),K,1,inf),EXPAND(PRODCONTRACT(%%)) = %%);

                   inf                  inf         inf
                  /===\                /===\       /===\        K
                   ! !           K      ! !         ! !        P
(D367)             ! !  (F(K) - P ) = ( ! !  F(K))  ! !  (1 - ----)
                   ! !                  ! !         ! !       F(K)
                  K = 1                K = 1       K = 1

at p=0:
(C368) TAYLOR(LHS(%),P,0,3);

TAYLOR unable to expand at a point specified in: ERREXP1
Returned to Macsyma Toplevel.

Yet,
(C369) TAYLOR(RHS(%),P,0,3);

                         inf              inf
                        /===\            /===\
                         ! !              ! !         2
           inf         ( ! !  F(K)) P   ( ! !  F(K)) P
          /===\          ! !              ! !
           ! !          K = 1            K = 1
(D369)/T/  ! !  F(K) - -------------- - ---------------
           ! !              F(1)             F(2)
          K = 1

                                                                 inf
                                                                /===\
                                                                 ! !         3
                                            ((F(2) F(1) - F(3))  ! !  F(K)) P
                                                                 ! !
                                                                K = 1
                                          - ---------------------------------- + . . .
                                                      F(3) F(2) F(1)

works, and probably should be implemented.  But some F(K) can be zero !

(C370) SUM(''(lhs(tayunscru)),J,0,inf);

Time= 106 msecs
                         inf    inf
                         ====  /===\
                         \      ! !      K - J    K + 1
(D370)                    >     ! !  (- Q      - P      + 1)
                         /      ! !
                         ====  K = 0
                         J = 0

can be expanded at p=0 !
(C371) TAYLOR(''(rhs(tayunscru)),P,0,3);

                                           inf
                                          /===\
                                           ! !        K    2
              inf                ((Q - 1)  ! !  (1 - Q )) P
             /===\                         ! !
              ! !        K                K = 1
(D371)/T/ - ( ! !  (1 - Q )) P + ---------------------------
              ! !                             Q
             K = 1

                                                             inf
                                                            /===\
                                        4      3             ! !        K    3
                                   ((2 Q  - 2 Q  + 2 Q - 1)  ! !  (1 - Q )) P
                                                             ! !
                                                            K = 1
                                 - ------------------------------------------- + . . .
                                                      4    3
                                                     Q  - Q

which special-cases the j=k term via
(C372) tayunscru;

        inf                                         inf
       /===\                                       /===\
        ! !      K - J    K + 1             J + 1   ! !        K
(D372)  ! !  (- Q      - P      + 1) = (- P)      ( ! !  (1 - Q ))
        ! !                                         ! !
       K = 0                                       K = 1

                         inf                             J - 1
                        /===\        K - J    K + 1      /===\
                         ! !      - Q      - P      + 1   ! !      K - J    K + 1
                       ( ! !      ---------------------)  ! !  (- Q      - P      + 1)
                         ! !                K - J         ! !
                        K = J + 1      1 - Q             K = 0




∂09-Aug-89  0854	RPG  
 ∂08-Aug-89  2003	JMC 	re: I forgot to mention this  
[In reply to message rcvd 08-Aug-89 13:34-PT.]

This cannot be done immediately or even soon.  I have been exploring
the matter with Vladimir in mind, and they are stalling
Professor (Research) appointments until yet another committee
deliberates and reports.

I had a 1-year time frame in mind. However, a more nearly deterministic 
situation is preferrable.

			-rpg-

∂09-Aug-89  0910	RPG 	re: reply to message
[In reply to message rcvd 09-Aug-89 09:04-PT.]

I publish about 3-5 papers per year. My current CV has about 25 papers
listed.

			-rpg-

∂09-Aug-89  0928	RPG 	re: reply to message
[In reply to message rcvd 09-Aug-89 09:11-PT.]

I can get very strong letters from the following:

Steele, Sussman, Abelson, Friedman, Wise, Hudak, Wand, Winston, Waltz,
David Kuck, Wulf, Fahlman, Bobrow, Deutsch, Balzer, Guttag, Halstead,
Clinger, maybe Minsky, Hillis, and Haynes.

I should note that placess like Utah, MIT, and CMU as me to write tenure
letters for their faculty.

			-rpg-

∂09-Aug-89  1146	CLT 	McCarthy's msg about a problem with permission to hire someone   
To:   levinth@SIERRA.STANFORD.EDU
CC:   JMC@SAIL.Stanford.EDU 

Here are some details.

We attempted to hire Harlan Sexton as a research associate
(part-time) to work on a DARPA project.  The task was contracted
as a low overhead task since the people involved were mainly
Lucid folks and planned to do the work at Lucid.  This was blocked
by the Provost's office on the basis of a `conflict of interest' --
Stanford was unwilling to hire Lucid employees even though
Lucid was agreeable to Sexton spending part of his time on this
project.  (We initiated the appointment papers in February and only
found out in July that it couldn't be done.)
We asked under what conditions we could hire Harlan and were told
(by Mary Hanrahan and Susan Nowlis in the provost's office) that
there was no way we could do that.

This grant times out the end of February 1990 (with the closing
of the umbrella) and there is no possibility of getting a no
cost extension.  Thus we need to go full steam ahead in order
to complete the project.  What we propose is that Dick Gabriel
and Harlan Sexton will take 6-mo leave of absence from Lucid
to work on this project.   We would like to appoint them as full
time research associates for that period.  This needs to be done
by 1-September.   

∂09-Aug-89  1347	VAL 	Gelfond   
Gelfond is visiting with us this week. Would you like to meet with him?
We can invite him to join us for lunch Friday, if you wish.

∂09-Aug-89  1440	MPS  
call John Nefeh - 943-1711

∂09-Aug-89  1518	JONES@vaxb.acs.unt.edu 	RE: re: researchers in logical reasoning 
Received: from vaxb.acs.unt.edu ([192.12.10.4]) by SAIL.Stanford.EDU with TCP; 9 Aug 89  15:18:28 PDT
Date: Wed, 9 Aug 89 17:03 CDT
From: JONES@vaxb.acs.unt.edu
Subject: RE: re: researchers in logical reasoning
To: JMC@SAIL.Stanford.EDU
X-VMS-To: IN%"JMC@SAIL.Stanford.EDU"

 Dear Dr. McCarthy,

	Thanks so much for your prompt response.  I apologize for the
lack of clarity about my research interests, however the briefness of
my message seemed to be the most expedient for both you and me.
	The use of logic to model common sense knowledge and reasoning 
is an appropriate description of what I am interested in.  Your
recommendation of Don Perlis is appreciated, and I will contact him.
I had a hint that he may be a possibility because of his article
in the June issue of Artificial Intelligence.  However, for a novice 
such as myself it is not prudent to assume that an author is also a
respected researcher.  Are there any other researchers/schools you 
would recommend?
	I have noticed that Stanford, more than any other school,
publishes quite a bit of material in this area.  Given the opportunity,
I would dearly love to attend Stanford.  Do you have advice on how a 
good student, but one who may have a somewhat unorthodox background, 
can improve his chances of gaining acceptance into your program?

					Sincerely,
					James Jones
					Jones@vaxb.acs.unt.edu

∂09-Aug-89  1643	TAJNAI@Score.Stanford.EDU 	research interests
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 9 Aug 89  16:43:04 PDT
Date: Wed 9 Aug 89 16:28:56-PDT
From: Carolyn Tajnai <TAJNAI@Score.Stanford.EDU>
Subject: research interests
To: jmc@Sail.Stanford.EDU
Message-ID: <12516846056.20.TAJNAI@Score.Stanford.EDU>


please update, John.


@b{John McCarthy},
@i[Charles M. Pigott Professor of Engineering, Professor of Computer
Science and (by courtesy) Electrical Engineering,  Ph.D. Princeton, 1951].@\
Research Interests@i(:)  Artificial intelligence, computing with
symbolic expressions, time sharing, formalizing common sense, non-monotonic
logic.
One of the founders of artificial intelligence research,  Professor 
McCarthy invented
LISP, the programming language most used in AI research and also
first proposed the general purpose time-sharing mode of using computers.
The emphasis of his AI research has been in identifying the common
sense rules that determine the consequences of actions and other events,
the expression of such rules and other common sense information as
sentences in logical languages in the data bases of artificial intelligent
programs.  @comment[He has also worked on formalizing common sense reasoning,
and his recent work concerns non-monotonic common sense reasoning
whereby people and computers draw conjectural conclusions by assuming that complications are absent from a situation.  ]
A past president of AAAI,
Professor McCarthy received the First IJCAI Research Excellence Award in 1985.
-------

∂10-Aug-89  0200	@RELAY.CS.NET:hibino@ntt-elis.ntt.jp 	Ask your intention of using ELIS
Received: from RELAY.CS.NET ([192.31.103.4]) by SAIL.Stanford.EDU with TCP; 10 Aug 89  02:00:16 PDT
Received: from relay2.cs.net by RELAY.CS.NET id ab09308; 10 Aug 89 5:01 EDT
Received: from ntt.jp by RELAY.CS.NET id ab01778; 10 Aug 89 4:58 EDT
Received: by ntt-sh.ntt.jp (3.2/ntt-sh-03e) with TCP; Thu, 10 Aug 89 17:42:33 JST
Received: by YECL.ntt.jp (1.2/NTTcs02) with TCP; Thu, 10 Aug 89 17:46:52+0900
Received: by ntt-elis.ntt.jp (3.2/NTTcs01b) with TCP; Thu, 10 Aug 89 17:46:33 JST
Date: Thu, 10 Aug 89 17:46:33 JST
Message-Id: <8908100846.AA02513@ntt-elis.ntt.jp>
To: JMC@SAIL.STANFORD.EDU
From: hibino@ntt-elis.ntt.jp
Sender: hibino@ntt-elis.ntt.jp
Subject: Ask your intention of using ELIS
Cc: hibino@ntt-elis.ntt.jp

                                     NIPPON TELEGRAPH AND
                                            TELEPHONE CORPORATION
                                     1-2356 Take Yokosuka-shi
                                     Kanagawa 238-03 Japan
                                     TEL +81 468 59 2330

                                     August 10, 1989

Prof. John  McCarthy
Professor of Computer Science
Stanford University


Dear Professor McCarthy,


I have heard form Mr. Ikuo Takauchi that you have intention to use
our ELIS machine.

As you know, we sent two ELIS machines to KSL of Stanford University.
Those machines were used for the co-operation between KSL and NTT.
However, the co-operation had completed last summer.
So, those machines will be scheduled to be return to Japan in a 
short time.

If your intention of using ELIS is true, we would like to offer you
those machines until next spring. You are displeased with short span
of time, but there are following reasons.

Those machines were imported as temporary imported goods (TIG),
so the period staying in your country cannot exceed 2 years.
Since we sent our machine last spring, the limit will come next
spring.

Fortunately, I have a chance to visit your country the last third of
August to attend IJICAI'89. And I have a plan to stay in San Francisco
form August 30 to 31.
I would like to visit you and ask your intention.
Would you please make a time to meet me?

I am looking forward to hearing from you.

                                     Sincerely Yours,
                                           
                                     Yasushi HIBINO
                                     Research Group Leader
                                     Language Media Laboratory
                                     Human Interface Laboratories



∂10-Aug-89  0727	JONES@vaxb.acs.unt.edu 	RE: re: re: researchers in logical reasoning  
Received: from vaxb.acs.unt.edu (VAXA.ACS.UNT.EDU) by SAIL.Stanford.EDU with TCP; 10 Aug 89  07:27:14 PDT
Date: Thu, 10 Aug 89 09:28 CDT
From: JONES@vaxb.acs.unt.edu
Subject: RE: re: re: researchers in logical reasoning
To: JMC@SAIL.Stanford.EDU
X-VMS-To: IN%"JMC@SAIL.Stanford.EDU"

"UNT" stands for University of North Texas (formerly North Texas State
Unverisity.)  

In response to my question about improving my chances for admission into
Stanford, you recommended:  1) material accepted for publications and
2) strong recommendations.  I do not have such material for publication.
(I am returning to school after having been in the work force for 5 years,
although I do have a master's degree and very strong academic performance.
My research interest represents a change from a traditional MIS background.)

Concerning the strong recommendations, I do have those.  However, the
recommenders are not of national stature.

I am most interested in Stanford, however as a prudent person I intend 
on investigating other programs as well.  In addition to Don Perlis,
are there others you would recommend?

				Sincerely,
				James Jones

∂10-Aug-89  0956	C.COLE@Macbeth.Stanford.EDU 	RE: O.T.A. message to Su-etc   
Received: from Macbeth.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Aug 89  09:56:07 PDT
Date: Thu 10 Aug 89 09:53:55-PDT
From: George Cole <C.COLE@Macbeth.Stanford.EDU>
Subject: RE: O.T.A. message to Su-etc
To: jmc@sail.Stanford.EDU
Message-ID: <12517036290.81.C.COLE@Macbeth.Stanford.EDU>

My public response is going to SU-ETC; I thought I'd add a private note. I've
been asked to help another lawyer with a big plaintiff's case. The plaintiff
is suing many individuals (a complex accident) including a community college 
and an auto manufacturer, claiming products liability. I've got biases against
suing either, generally. When I heard the plaintiff's view of the facts, I was
quite willing to help prosecute the suit; the facts made the difference. I'd
be interested in your reaction to the interaction of the law and technology 
here. 
	4 people are riding in a Toyota Celica, which crashes into a Dodge Van
(which had crashed into a Toyota pickup that was stopped in the fast lane of
280 at night in the rain, because a small front-end collision had knocked out
the entire electrical system and killed the engines and hazard lights). 3 of
the Celica's passengers walked away with cuts and bruises. One was rendered
paraplegic. She was also the only one wearing her seat belt, and had that and
her shoulder harness on, too.
	The two-door Celica has a folding seat, so back-seat passengers can
get in and out. The seat also moves forward. But the release for the seat
back is on the back of the seat, facing the rear-seat passenger, right where
he can trip it if he moves forward -- like he might during a front-end
collision. Not only that, but the seat-back release, if the seat back moves
forward slightly, then triggers the seat-track release, so the whole front
seat moves. 
	There's undoubtedly a "battle of the experts" about to happen. The
(indirect) report from the plaintiff's reconstructionist and biomedical people
is simple: the rear seat passenger went forward and tripped the release. (He
did complain of a bruised ankle.) The seat back went forward, the seat track
release went next, and then the momentum of the rear-seat passenger, seat-back,
and seat, all focused on the front-seat passenger as she hit the restraining
limit of the seat belt. Crack.
	There are several possible design defects here. (1) is obvious -- don't
put the seat-release directly in front of a rear-seat passenger where moving
forward during a front-end collision he can trigger it. Auto manufacturers 
can foresee cars are going to be in collisions and have to design with that in
mind. (The law is not that cars have to be crashproof, but they do have to be
crash-worthy. There is no federal regulation or other standard as to where the
release has to be; and most cars don't have that design.) (2) is also obvious -
don't combine the seat-back and seat-track release. (3) is a bit harder -- put
an inertial lock on those releases. They're not supposed to work when the car
is stopping suddenly, and inertial locks existed already for seat-belts. (4) is
a bit farther afield -- build the seat backs with better cushioning materials
and sturdier frames & tracks, to protect the front-seat passenger better 
passively from rear-seat passenger's momentum. (5) would be far better, but
isn't yet available, damn it -- airbags. (I know. I was looking for them. But
if you get anything below a large car, they're not yet available.)
	My common sense said something was wrong when I heard (a) the belted
passenger was the one crippled and (b) the seat-release lever faced the rear
passenger. The legal research I've done since (at this point I'm confident I'm
abreast of California products liability law, especially as it pertains to
design defects) supports that common sense. The key to manufacturer's liability
is foreseeability. If they can foresee damage from a product from normal use
or expected misuse, they have to design to eliminate or minimize that damage.
Auto accidents are foreseeable misuse. (Why? Otherwise, manufacturers have an
incentive to design as poorly as possible and trumpet the fact, so consumers
have the lowest possible expectation.) Custom in an industry is not a defense
(otherwise, they keep bad customs when better alternatives are feasible). I'm
trying to explain what "state of the art" means. The test becomes a balancing
act between alternatives -- if the manufacturer can show that no safer design
was feasible at a roughly comparable cost (is $4 to avoid a death sensible?),
then he's acting at "state of the art".
	And then there's the stupidity of designing a car so a small jolt to
the front puts the battery through the fusebox (the pickup, remember).....
	...and this is too long a message. I'll only continue if you're 
interested.
				George
-------

∂10-Aug-89  1005	tutiya@russell.Stanford.EDU 	response to your netmail article    
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Aug 89  10:05:45 PDT
Received: from LOCALHOST by russell.Stanford.EDU (4.0/inc-1.0)
	id AA08915; Thu, 10 Aug 89 10:11:37 PDT
Message-Id: <8908101711.AA08915@russell.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: response to your netmail article
Date: Thu, 10 Aug 89 10:11:17 PDT
From: syun tutiya <tutiya@russell.Stanford.EDU>

John,

I am visiting CSLI from Japan and found your would-be
article interesting.  I forwarded to a friend of mine who is
a main figure in promoting computer network development in
Japan.  This is his response.  Just to talk to him directly
if you need further clarification.  Tokyo Univerisity is in
the 5 minuite distance in the netmail world now.

Thanks.

syun

------- Forwarded Message

Return-Path: jun@xroads.cc.u-tokyo.ac.jp
Received: from mtfuji.gw.u-tokyo.ac.jp ([128.167.64.2]) by russell.Stanford.EDU (4.0/inc-1.0)
	id AA06069; Thu, 10 Aug 89 01:42:17 PDT
Received: from relay.cc.u-tokyo.ac.jp by mtfuji.gw.u-tokyo.ac.jp (5.61/WIDE/JUNET-1.2)
	id AA11687; Thu, 10 Aug 89 04:33:57 -0400
Received: from ccut.cc.u-tokyo.ac.jp (ccutrd.cc.u-tokyo.ac.jp) by relay.cc.u-tokyo.ac.jp.cc.u-tokyo.ac.jp (3.2/6.3J-1.1jp)
	id AA12855; Thu, 10 Aug 89 17:32:50 JST
Received: from xroads.cc.u-tokyo.ac.jp by ccut.cc.u-tokyo.ac.jp (5.61/6.4J.6-ut1.43)
	id AA03389; Thu, 10 Aug 89 17:33:31 +0900
Return-Path: <jun@xroads.cc.u-tokyo.ac.jp>
Received: from localhost by xroads.cc.u-tokyo.ac.jp (4.0/JunetWide0.9FromSMI-4.0)
	id AA21212; Thu, 10 Aug 89 17:32:18 JST
Message-Id: <8908100832.AA21212@xroads.cc.u-tokyo.ac.jp>
To: syun tutiya <tutiya@russell.Stanford.EDU>
Cc: jun@cc.u-tokyo.ac.jp, z30050%ccut.u-tokyo.junet%utokyo-relay@RELAY.CS.NET
Subject: Re: Just FYI 
In-Reply-To: Your message of Wed, 09 Aug 89 21:23:50 PDT.
             <8908100423.AA04319@russell.Stanford.EDU> 
Date: Thu, 10 Aug 89 17:32:15 JST
>From: jun@xroads.cc.u-tokyo.ac.jp

thanks for the mail. ..interesting.

i just read it quick, so i might missed the point:

isdn technology which supports g-iv fax seems to me the something you
want though. ccitt x.400/x.500 and isdn covers both email world and
fax world, as well as possibly the interconnection btwn the both,
right? we got isdn (i got it at my home now) and beta x.400/x.500
software. do you really need new technology more than that from a
point of view of your discussion? Even Internet community in the U.S.
is working on these technology, well at least on x.400/x.500 things.

jun

------- End of Forwarded Message

∂10-Aug-89  1030	MPS  
Good morning,

Stephen Coles, a former student of yours, would like to talk to you aaout
robotics.

He works for JPL and is preparing a master tape of the history of
robotics, including the research from MIT, Stanford, JPL, etc., and
would like access to any film you may have or know the whereabouts on the subject
of robotics.

His number is 818-354-9734.

Pat

∂10-Aug-89  1234	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	an intriguing screw case for TAYLOR    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 10 Aug 89  12:34:15 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 409022; 10 Aug 89 13:09:39 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 106006; Thu 10-Aug-89 04:53:41 PDT
Date: Thu, 10 Aug 89 04:53 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Reply-To: "rwg@yukon.symbolics.com"@RUSSIAN.SPA.Symbolics.COM
Subject: an intriguing screw case for TAYLOR
To: macsyma-i@RUSSIAN.SPA.Symbolics.COM
cc: "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, math-fun@RUSSIAN.SPA.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890809135615.2.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890810115306.6.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Wed, 9 Aug 89 06:56 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
    <stuff about expanding sums of prods>

The expansion at p=0 that I gave for

                             inf    inf
                             ====  /===\
                             \      ! !      k - j    k + 1
    (D370)                    >     ! !  (- q      - p      + 1)
                             /      ! !
                             ====  k = 0
                             j = 0

was slightly wrong, because I had (-p)↑(j+1) instead of -p↑(j+1) in

    (C372) tayunscru;

            inf                                       inf
           /===\                                     /===\
            ! !      k - j    k + 1           j + 1   ! !        k
    (D372)  ! !  (- q      - p      + 1) = - p      ( ! !  (1 - q ))
            ! !                                       ! !
           k = 0                                     k = 1

                             inf                             j - 1
                            /===\        k - j    k + 1      /===\
                             ! !      - q      - p      + 1   ! !      k - j    k + 1
                           ( ! !      ---------------------)  ! !  (- q      - p      + 1)
                             ! !                k - j         ! !
                            k = j + 1      1 - q             k = 0

but it probably doen't matter, because I think the sum is numerically meaningless,
since it contains arbitrarily negative powers of q, which must have abs < 1.

On the other hand, the actual application scaled the product by a sufficiently
positive (quadratic) power of q to restore meaning, and I was able to use a slight
generalization of the above to empirically verify the peculiar bibasic identity

                          j   inf                             inf
                         ( ) /===\                           /===\
                  j + 1   2   ! !        k - j      k         ! !        k
       inf   (- 1)      q     ! !  (1 - q      - a p )   a q  ! !  (1 - q )
       ====                   ! !                             ! !
       \                     k = 0                           k = 1
(d546)  >    ----------------------------------------- = ------------------
       /                    j                                  q - p
       ====               /===\
       j = 0               ! !        k
                           ! !  (1 - q )
                           ! !
                          k = 1

(George, note the Reply-To:).  This is one of the simplest cases of that
"nonlocal derangement" formula I got from the sin(infinite series) formula,
which takes two arbitrary sequences and constructs a summation identity.
In this case the sequences were merely geometrics:  a*p↑n/q↑n, and q↑-n.

Using instead the sequences a/n↑2 and 1/n gives that funny thing I sent
with complex gammas, which simplfies to

(D547) 

inf           j   sqrt((j + 1) (j + a))         j - 1   sqrt((j + 1) (j + a))
====  cos(pi (- - ---------------------)) Binom(----- + ---------------------, j)
\             2             2                     2               2
 >    ---------------------------------------------------------------------------
/                                              2
====                                    (j + 1)
j = 0
       
	 3
       pi  (1 - a)
     = ----------- .
	    24

but this is no less weird.  It converges best for negative a (or negative Re(a)),
whereupon the first -a terms are complex!  E.g., for a = -7, the first 11 terms
simplify to the identically real approximation

               sqrt(33) pi            sqrt(15) pi              sqrt(7) pi
         2 cos(-----------)   59 cosh(-----------)   2209 cosh(----------)
                    2                      2                       2
(D548) - ------------------ + -------------------- + --------------------
              190575                  225                    2205

   11 sqrt(5) sin(sqrt(5) pi)   13 sqrt(3) sin(2 sqrt(3) pi)
 + -------------------------- - ----------------------------
            2268000                      2721600

                                                          3
   277 sqrt(3) sinh(sqrt(3) pi)   5 cosh(pi) sinh(pi)   pi
 - ---------------------------- - ------------------- = ---
              1080                        24             3

(C549) SFLOAT(%);

(D549)                          10.335409 = 10.335426

How does it do that?  When j is even, the cosine turns into a cosh, which
is real, and the numerator factors of the Binom can be paired
(foo)*(foo bar), (bar := complex conjugate), which is also real.  When j
is odd, the cos becomes an i sinh, and binom numerator looks like
(foo) (i baz) (foo bar), which is also pure imaginary.  So every term
is real!

∂10-Aug-89  1308	levinth@sierra.STANFORD.EDU 	McCarthy CPL Research Project  
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 10 Aug 89  13:08:35 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Thu, 10 Aug 89 13:07:22 PDT
Date: Thu, 10 Aug 89 13:07:22 PDT
From: levinth@sierra.STANFORD.EDU (Elliott C. Levinthal)
To: BSCOTT@Score.Stanford.EDU
Cc: HK.PLD@forsythe.Stanford.EDU, Levinthal@Sierra.Stanford.EDU,
        BScott@Score.Stanford.EDU, clt@sail, jmc@sail
In-Reply-To: Betty Scott's message of Thu 10 Aug 89 10:45:05-PDT <12517045606.17.BSCOTT@Score.Stanford.EDU>
Subject: McCarthy CPL Research Project


Betty, Before receiving your message, I talked with Pat Devaney. Based
on that conversation and your message you should proceed with the
paperwork. this will be based on the leave of absence and full-time
work under John's supervision on his project. The issue of whether or
not its a six months appointment or less than six months invovles
staff benefits only and I leave it to you and John to work out with
the appointees. There is no problem having a fixed term appointment
that has no implications or commitments to continues employment. 

Elliott

∂11-Aug-89  0616	boesch@vax.darpa.mil 	ISTO DIRECTOR APPOINTMENT   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 11 Aug 89  06:16:43 PDT
Received: from sun44.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA20159>; Fri, 11 Aug 89 09:08:24 -0400
Posted-Date: Fri, 11 Aug 89 09:09:25 EDT
Message-Id: <8908111309.AA01498@sun44.darpa.mil>
Received: from LOCALHOST by sun44.darpa.mil (4.1/5.51)
	id AA01498; Fri, 11 Aug 89 09:09:27 EDT
To: dpsys-pi@vax.darpa.mil
Subject: ISTO DIRECTOR APPOINTMENT
Date: Fri, 11 Aug 89 09:09:25 EDT
From: boesch@vax.darpa.mil


	We are pleased to announce the appointment of Barry 
	Boehm as ISTO Director effective in November.  Plans 
	for the transition will be announced over the next few 
	weeks.  Most of you know Barry's extensive work on the
	problems of very large-scale software systems development. 
	Barry has recently joined the faculty at UCLA after a 
	long and distinguished career at TRW.
					

Brian

∂11-Aug-89  0821	BSCOTT@Score.Stanford.EDU 	Re: McCarthy CPL Research Project
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Aug 89  08:21:24 PDT
Date: Fri 11 Aug 89 08:20:25-PDT
From: Betty Scott <BSCOTT@Score.Stanford.EDU>
Subject: Re: McCarthy CPL Research Project
To: levinth@Sierra.Stanford.EDU
cc: HK.PLD@forsythe.Stanford.EDU, Levinthal@Sierra.Stanford.EDU,
    clt@Sail.Stanford.EDU, jmc@Sail.Stanford.EDU, BSCOTT@Score.Stanford.EDU
In-Reply-To: Message from "levinth@sierra.STANFORD.EDU (Elliott C. Levinthal)" of Thu 10 Aug 89 13:07:49-PDT
Message-ID: <12517281413.21.BSCOTT@Score.Stanford.EDU>

O.K., we will proceed as soon as we settle the benefits issue.

Betty
-------

∂11-Aug-89  1020	MPS 	phone call
Good morning

Call Jack Cate, 321-1225

∂11-Aug-89  1240	tom@Polya.Stanford.EDU 	[jcm@Polya.Stanford.EDU: was this for me?]    
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Aug 89  12:40:28 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA25507; Fri, 11 Aug 89 12:40:26 -0700
Date: Fri, 11 Aug 89 12:40:26 -0700
From: Tom Dienstbier <tom@Polya.Stanford.EDU>
Message-Id: <8908111940.AA25507@Polya.Stanford.EDU>
To: JMC@sail
Subject: [jcm@Polya.Stanford.EDU: was this for me?]

Cc: gotelli@score, baldwin@score
Subject: new account new machine


John, the new machine to house your Score files will be HUDSON(VAX3600
running ULTRIX). Please send me what you would like your account name
to be and we will open it up for you. We will then get with Carolyn to
deal with the accounting. After the acccount is open the files then
can be moved.

Tom


------- End of Forwarded Message


∂11-Aug-89  1558	MPS  
Harriet Gardner, 408 725-1208 has just returned from the
USSR and has a letter for you from Misha Donskoy.

Pat

∂11-Aug-89  1801	harnad@clarity.Princeton.EDU 	Editor of Current Contents    
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 11 Aug 89  18:01:25 PDT
Received: from lexis.Princeton.EDU by Princeton.EDU (5.58+++/2.20)
	id AA13967; Fri, 11 Aug 89 16:01:18 EDT
Received: by lexis.Princeton.EDU (4.0/1.83)
	id AA14092; Fri, 11 Aug 89 15:35:27 EDT
Date: Fri, 11 Aug 89 15:35:27 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8908111935.AA14092@lexis.Princeton.EDU>
To: harnad@clarity.Princeton.EDU
Subject: Editor of Current Contents

To: BBS Editors and Associates

Eugene Garfield, the Founder of the Institute for Scientific
Information in Philadelphia (Current Contents, Science Citation Index)
is retiring as Editor-In-Chief of Current Contents and looking to
groom a successor. The choice is important not only to ISI but to the
world scientific community, because of the important role CC and SCI
play. He's asked for my advice on a successor, and I in turn ask you:

The successor must be literate across scientific disciplines and able to
prepare (with a team of seven researchers working under him) one
article-length editorial per week in CC analyzing some aspect of
science or scientific communication. He must also direct all the
different versions of CC (life sciences, physical sciences, etc.) and
the interactions with the other divsisions of ISI (especially the
citation indices).

In my view this is a position for an energetic, productive and
innovative person with a strong academic background either directly or
indirectly related to scientific communication.

Please let me know if you have any suggestions.

Stevan Harnad

∂11-Aug-89  1804	jun@xroads.cc.u-tokyo.ac.jp 	Re: reply from jmc   
Received: from mtfuji.gw.u-tokyo.ac.jp ([128.167.64.2]) by SAIL.Stanford.EDU with TCP; 11 Aug 89  18:04:43 PDT
Received: from relay.cc.u-tokyo.ac.jp by mtfuji.gw.u-tokyo.ac.jp (5.61/WIDE/JUNET-1.2)
	id AA18297; Fri, 11 Aug 89 21:06:15 -0400
Received: from ccut.cc.u-tokyo.ac.jp (ccutrd.cc.u-tokyo.ac.jp) by relay.cc.u-tokyo.ac.jp.cc.u-tokyo.ac.jp (3.2/6.3J-1.1jp)
	id AA00978; Sat, 12 Aug 89 10:05:09 JST
Received: from xroads.cc.u-tokyo.ac.jp by ccut.cc.u-tokyo.ac.jp (5.61/6.4J.6-ut1.44)
	id AA11435; Sat, 12 Aug 89 10:05:55 +0900
Return-Path: <jun@xroads.cc.u-tokyo.ac.jp>
Received: from localhost by xroads.cc.u-tokyo.ac.jp (4.0/JunetWide0.9FromSMI-4.0)
	id AA23050; Sat, 12 Aug 89 10:04:36 JST
Message-Id: <8908120104.AA23050@xroads.cc.u-tokyo.ac.jp>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: reply from jmc 
In-Reply-To: Your message of Thu, 10 Aug 89 11:21:50 PDT.
             <8908101821.AA11191@russell.Stanford.EDU> 
Date: Sat, 12 Aug 89 10:04:33 JST
From: jun@xroads.cc.u-tokyo.ac.jp

hello,

> It seems to me that your correspondent did miss my point.  I
> don't understand the references to various communication
> standards, like ccitt x.400/x.500, but I suppose them to be
> standards for the handshaking that occurs when messages are
> transmitted.  However, my point concerns the fact that fax uses
> the existing telephone numbers, and the only business requirement
> is to buy a machine and pay the phone bill.  My objection is to
> the organization of electronic mail into networks with which the
> user must concern himself.  I have no problem with the various
> standards for the technology of the transmission, because I
> assume that the people developing them are competent to do so and
> have heard nothing to the contrary.

i just wanted to say that the ccitt technologies assume that both the
existing phone and a user entity in the existing operating
system would be treated in the same space so that the current email
and the fax could be used simultaneously, both from the normal phones
and the `normal' operating system environment.

jun 

∂13-Aug-89  2205	@RELAY.CS.NET:okuno@ntt-20.ntt.jp 	Inquiry about the ELIS again  
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 13 Aug 89  22:05:06 PDT
Received: from relay2.cs.net by RELAY.CS.NET id aa11858; 14 Aug 89 1:06 EDT
Received: from ntt.jp by RELAY.CS.NET id ac16333; 14 Aug 89 1:04 EDT
Received: by ntt-sh.ntt.jp (3.2/ntt-sh-03e) with TCP; Mon, 14 Aug 89 13:53:09 JST
Received: by mecl.ntt.jp (3.2/NTTcs02c) with TCP; Mon, 14 Aug 89 13:55:00 JST
Date: 14 Aug 89 13:46:39 JST
From: "Hiroshi G. Okuno" <okuno@ntt-20.ntt.jp>
Message-Id: <Zmm=IMAP/ELIS617973@Nickel.NTT.JP>
To: jmc@SAIL.STANFORD.EDU
Subject: Inquiry about the ELIS again 
Cc: tak@nuesun.ntt.jp, okuno@ntt-20.ntt.jp

John,

I'm writing to you again to ask whether you still want to
use the ELIS machines.  I sent you the original inquiry on
May 11th, but it's in August.  Sorry for getting back to you
so late.  Since the ELIS machines should be returned to Japan
in February, it is only for a half year for you to use them.
We would greatly appreciate it if you could use them for the
rest of the Temporary Import Goods period.

Looking forward to hearing you good news.

Regards,

- Gitchang -

∂14-Aug-89  0007	@RELAY.CS.NET:okuno@ntt-20.ntt.jp 	re: Inquiry about the ELIS again   
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 14 Aug 89  00:07:25 PDT
Received: from relay2.cs.net by RELAY.CS.NET id aa12728; 14 Aug 89 3:08 EDT
Received: from ntt.jp by RELAY.CS.NET id ac17031; 14 Aug 89 3:03 EDT
Received: by ntt-sh.ntt.jp (3.2/ntt-sh-03e) with TCP; Mon, 14 Aug 89 15:46:15 JST
Received: by mecl.ntt.jp (3.2/NTTcs02c) with TCP; Mon, 14 Aug 89 15:49:20 JST
Date: 14 Aug 89 15:34:14 JST
From: "Hiroshi G. Okuno" <okuno@ntt-20.ntt.jp>
Message-Id: <Zmm=IMAP/ELIS618188@Nickel.NTT.JP>
To: John McCarthy <JMC@SAIL.STANFORD.EDU>
Subject: re: Inquiry about the ELIS again   
Cc: okuno@ntt-20.ntt.jp
In-Reply-To: John McCarthy's message of <NLcL$@SAIL.Stanford.EDU>

John,

Thanks for your quick reply.  I understand your policy and
forwarded your reply to Mr. Hibino and Tak.

- Gitchang -

P.S.  It is very hot and humid here.  I miss California.
I could not visit Stanford, because Tak has moved to Basic
Research Labs and I was appointed as his successor.

∂14-Aug-89  0654	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Radical, Dude 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 14 Aug 89  06:54:06 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 410136; 14 Aug 89 09:52:23 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 106203; Mon 14-Aug-89 06:46:01 PDT
Date: Mon, 14 Aug 89 06:45 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Reply-To: "rwg@yukon.symbolics.com"@RUSSIAN.SPA.Symbolics.COM
Subject: Radical, Dude
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890814134521.7.RWG@TSUNAMI.SPA.Symbolics.COM>

Continued abuse of "nonlocal derangement" got me a slowly convergent,
ugly mess with these intriguing special cases:

       inf          j                     b
       ====    (- 1)  cos(pi Sqrt(j + f + -) Sqrt(j + f + b d))
       \                                  d
(D922)  >      ------------------------------------------------
       /                                 2
       ====                       (j + f)
      j = -inf
                                                           
                                                             2          2   
                                                            d  + 1     pi    
                       =   (cos(pi b) cot(pi f) - sin(pi b) ------) ---------
                                                             2 d    sin(pi f)
                                                                    

(C923) SUBST(1,1.0d0,DFLOAT(EVAL(SUBST([F = 0.69d0,B = 0.105d0,D = 0.288d0,inf = 22,NOUNSUM = SUM],REMBOX(D922)))));

(D923)             - 14.856260559304385d0 = - 14.939713470759953d0

(C924) SUBST(1,1.0d0,DFLOAT(EVAL(SUBST([F = 0.69d0,B = 0.105d0,D = 0.288d0,inf = 99,NOUNSUM = SUM],REMBOX(D922)))));

(D924)             - 14.920856514039029d0 = - 14.939713470759953d0


                                 2                                  2
       inf                    3 b                                3 b
       ====    (j + f + b) (--------- + j + f + b) sin(pi (b - ---------))
       \                    j + f + b                          j + f + b
(D932)  >      -----------------------------------------------------------
       /                                                    2
       ====                  2              2            3 b
      j = -inf  (j + f - 2 b)  (j + f + 2 b)  sin(pi (--------- + f + b))
                                                      j + f + b

                                       2
                                     pi  (cot(pi (f - 2 b)) - cot(pi (f + 2 b)))
                                =    -------------------------------------------
                                                4 sin(pi (f + 2 b))

(C933) SUBST(1,1.0d0,DFLOAT(EVAL(SUBST([F = 0.69d0,B = 0.105d0,inf = 22,NOUNSUM = SUM],D932))));

(D933)                25.02867303741724d0 = 25.07666441062709d0

(C934) SUBST(1,1.0d0,DFLOAT(EVAL(SUBST([F = 0.69d0,B = 0.105d0,inf = 99,NOUNSUM = SUM],D932))));

(D934)                25.065820070729472d0 = 25.07666441062709d0

Bilateral sums nearly always generalize to the form Sum(F(j+f)), where
f is the "phase parameter".  But the first sum is not of this form:
the (-1)↑j (,which *prevents* the series from oscillating!,) can
be slid into the cos in the form of an additive pi*j.  As opposed to
pi*(j+f).  So it is probably possible to do this sum/.Cos->Sin, and
hence introduce another degree of freedom.  (But it won't be trivial:
it can go complex for real values of f, b, and d.)

I would appreciate hearing from anybody (note Reply-To:) who has ever
seen (or needed!?!) identities for sums of trigs of radicals or
rational fcns.  Otherwise, hubris will catch up to me and I'll imagine
that I have finally found something that would have caused Ramanujan
at least a head scratch.

∂14-Aug-89  0800	JMC  
Get more address change.

∂14-Aug-89  0900	JMC  
see about changing Gibbons time

∂14-Aug-89  1114	Mailer 	re: mathematical maturity in our youth    
To:   JMC@SAIL.Stanford.EDU, gds@SPAM.ISTC.SRI.COM,
      su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

[In reply to message from JMC rcvd 12-Aug-89 13:34-PT.]

Gregory Chaitin did important work in descriptive (Kolmogorov)
complexity theory in his teens.

∂14-Aug-89  1129	cloutier@sierra.STANFORD.EDU 	Appointment    
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 14 Aug 89  11:29:26 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Mon, 14 Aug 89 11:28:10 PDT
From: cloutier@sierra.STANFORD.EDU (Mary Cloutier)
Date: Mon 14 Aug 89 11:28:09-PDT
Subject: Appointment
To: jmc@Sail
Cc: cloutier@SIERRA
Message-Id: <619122489.0.CLOUTIER@SIERRA>
Mail-System-Version: <SUN-MM(219)+TOPSLIB(128)@SIERRA>


Dr. McCarthy:

	I have opened up a Friday, ll:00 a.m. meeting for you.  This is
the best I can do; otherwise it will not be for a while since Dr. Gibbons
will be travelling to Japan the end of the week of the 21st and not
returning until after Labor Day.

						M.C.

-------

∂14-Aug-89  1203	rathmann@eclipse.stanford.edu 	Inconsistency and Multiple Theories (progress report) 
Received: from eclipse.stanford.edu by SAIL.Stanford.EDU with TCP; 14 Aug 89  12:03:12 PDT
Received: by eclipse.stanford.edu (4.0/inc-1.0)
	id AA06602; Mon, 14 Aug 89 12:02:36 PDT
Date: Mon, 14 Aug 89 12:02:36 PDT
From: rathmann@eclipse.stanford.edu (Peter K. Rathmann)
Message-Id: <8908141902.AA06602@eclipse.stanford.edu>
To: jmc@sail.stanford.edu
Cc: rathmann@eclipse.stanford.edu
Subject: Inconsistency and Multiple Theories (progress report)

Dear Professor McCarthy,

A few weeks ago, we discussed the possibility of using nonmonotonic
"glue" to connect possibly inconsistent theories.  This seems to work,
and we can replace inconsistency with abnormality.

The default, zero information, behavior seems to be that if a
conclusion is invalidated, doubt is cast on any fact which forms an
essential part of any inference path leading to the invalidated
conclusion.   This behavior strikes me as reasonable, although it is
exactly the opposite of the desired localization of inconsistency.

Right now, there seem to be two ways to deal with this.  The simplest
is simply to accept it.  If we can assume that most inference chains
are fairly shallow, inconsistency will be localized anyway.  Shallow
inference seems to be the rule for expert systems; I don't know if
this carries over to human reasoning though.

The other alternative is to use priorities to break some of the
symmetries, and thereby localize the inconsistency.   This works, but
begs the question of how we come up with the priorities.   If there
are priorities already in place, i.e., to control the semantics of the
subtheories, the examples I looked at seemed to make use of them in a
reasonable way.  

So the research alternatives open right now seem to be to write up the
above observations carefully, and move on to something else, or else
to embark on a search for a systematic way of defining the priorities.

I am not sure which way to go right now, any input you can give would
be much appreciated.
				-Peter

∂14-Aug-89  1228	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Thesis progress 
Received: from IU.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 14 Aug 89  12:28:33 PDT
Received: from peabody.teleos.com by IU.AI.SRI.COM via SMTP with TCP;
	  Mon, 14 Aug 89 12:26:21-PST
Received: by peabody.teleos.com (3.2/4.16) id AA07074 for
	  jmc@sail.stanford.edu; Mon, 14 Aug 89 12:23:04 PDT
Date: Mon, 14 Aug 89 12:23:04 PDT
From: Leslie Kaelbling <leslie@teleos.com>
Message-Id: <8908141923.AA07074@peabody.teleos.com>
To: jmc@sail.stanford.edu
Subject: Thesis progress
Reply-To: leslie%teleos.com@ai.sri.com


I feel like I'm making pretty good progress on my thesis work.  I have some
theory, some algorithms, and a simple working mobile robot demo.  I still
need more of each, but I think I know what I need.

I've been learning a lot about statistics and about "learning automata"-- a 
field pretty much started by a Soviet named Tsetlin-- and the related study
of "Bandit problems" in the US statistics and OR community.

If you're interested, we could schedule an appointment for me to tell you 
about my progress so far and you to tell me how much more progress needs to 
be made.

- Leslie

∂14-Aug-89  1356	MPS  
Dave Goerz, of C. Itoh, referred to you by Nils Reimer would
like to talk to you about AI as applied to information
processing.

415-391-2510

Pat

∂14-Aug-89  1401	pjd@riacs.edu 	re: Text of signed Viewpoint  
Received: from icarus.riacs.edu by SAIL.Stanford.EDU with TCP; 14 Aug 89  14:01:13 PDT
Received: from hydra.riacs.edu by icarus.riacs.edu (5.59/2.1G)
	   id AA04706; Mon, 14 Aug 89 14:02:32 PDT
Received: by hydra.riacs.edu (4.12/2.0N)
	   id AA08816; Mon, 14 Aug 89 14:02:48 pdt
Message-Id: <8908142102.AA08816@hydra.riacs.edu>
Date: Mon, 14 Aug 89 14:02:48 pdt
From: Peter J. Denning <pjd@riacs.edu>
To: JMC@SAIL.Stanford.EDU
Subject: re: Text of signed Viewpoint

Got it, John.  Will look it over and send comments.
Peter

∂14-Aug-89  1743	cloutier@sierra.STANFORD.EDU 	re: Appointment
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 14 Aug 89  17:43:37 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Mon, 14 Aug 89 17:42:24 PDT
From: cloutier@sierra.STANFORD.EDU (Mary Cloutier)
Date: Mon 14 Aug 89 17:42:23-PDT
Subject: re: Appointment
To: JMC@SAIL.Stanford.EDU
Message-Id: <619144943.0.CLOUTIER@SIERRA>
In-Reply-To: <nMprb@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(219)+TOPSLIB(128)@SIERRA>

Dr. McCarthy:

I will have to call the other appointment tomorrow to see if I can
change back; let's hope they haven't filled in the Friday spot since
I spoke to them.

I'll get back to you tomorrow.

					Mary

-------

∂14-Aug-89  1800	JMC  
Leora

∂15-Aug-89  0300	@RELAY.CS.NET:okuno@ntt-20.ntt.jp 	re: Inquiry about the ELIS again   
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 15 Aug 89  03:00:47 PDT
Received: from relay2.cs.net by RELAY.CS.NET id aa00540; 15 Aug 89 6:01 EDT
Received: from ntt.jp by RELAY.CS.NET id aa03549; 15 Aug 89 5:58 EDT
Received: by ntt-sh.ntt.jp (3.2/ntt-sh-03e) with TCP; Tue, 15 Aug 89 18:10:05 JST
Received: by mecl.ntt.jp (3.2/NTTcs02c) with TCP; Tue, 15 Aug 89 18:13:19 JST
Date: 15 Aug 89 17:49:40 JST
From: "Hiroshi G. Okuno" <okuno@ntt-20.ntt.jp>
Message-Id: <Zmm=IMAP/ELIS621339@Nickel.NTT.JP>
To: John McCarthy <JMC@SAIL.STANFORD.EDU>
Subject: re: Inquiry about the ELIS again   
Cc: okuno@ntt-20.ntt.jp
In-Reply-To: John McCarthy's message of <6M0KP@SAIL.Stanford.EDU>

John,

Thanks for your interest.

The main research theme of Tak's group is context understanding in
natural language processing, but he is still engaged in
object-oriented programming research and design of new Lisp chip. 
Other themes studied at Information Division of Basic Research Labs
are lambda calculus, cryptography, neuro computer, bio computer and
pattern recognition.  Note that Shigeki belongs to Software Labs.

My group is engaged in

(1) developing MacELIS-II, a multiprocessor system of ELIS embedded
in Macintosh-II, and network-based parallel TAO,

(2) developing distributed programming environments by embedding
various network tools (mail reader, news reader, dictionaries,
ftp and so on) into Emacs-like editor,

(3) parallelizing AI systems.

The idea of MacELIS-II was presented at the U.S.-Japan Workshop on
Parallel Lisp.  Tak is responsible for extending the kernel of TAO for
MacELIS-II, although he is moved to the Basic Research Labs.  The last
theme is mine and I'll port the system to MacELIS-II after evaluating
the QLISP implementation.  For the moment, I'm applying the ATMS to
some aspects of learning process of the SOAR developed at CMU.

- Gitchang -

∂15-Aug-89  0718	njacobs@vax.darpa.mil 	Technical Reports
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 15 Aug 89  07:18:19 PDT
Received: from sun41.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA01592>; Tue, 15 Aug 89 10:11:16 -0400
Posted-Date: Tue 15 Aug 89 10:11:23-EDT
Received: by sun41.darpa.mil (4.1/5.51)
	id AA17001; Tue, 15 Aug 89 10:11:25 EDT
Date: Tue 15 Aug 89 10:11:23-EDT
From: Nicole L. Fields <NJACOBS@DARPA.MIL>
Subject: Technical Reports
To: dewitt@cs.wisc.edu, mike@postgres.Berkeley.EDU, jmc@sail.stanford.edu,
        lk@lanai.cs.ucla.edu, yemini@cs.columbia.edu, POPEK@lanai.cs.ucla.edu,
        alfred.spector@pt.cs.cmu.edu, smid@st1.icst.nbs.gov, JUBIN@d.isi.edu,
        glauer@bbn.com, chen@multimax.encore.com, jerry@pluto.arc.nasa.gov,
        shacham@spam.istc.sri.com, steve@tis.com, agrawala@mimsy.umd.edu,
        beyer@tsca.istc.sri.com, ken@gvax.cs.cornell.edu
Cc: njacobs@vax.darpa.mil
Message-Id: <619193483.0.NJACOBS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>


To all DARPA/ISTO Human Computer Interface, ALBM, and SAPE PIs:

	ANNUAL TECHNICAL REPORTS

This is the time of year when we require information from contractors
concerning accomplishments for the current fiscal year (FY) and
objectives for the next FY.  This information is used to assist us in
planning incremental funding for those efforts that are expected to
continue into the next FY (which begins 1 Oct 89).  This information
will also be used to brief the incoming Director of ISTO.  (Note that
this position remains unfilled.  We are actively seeking candidates.
If you have nominations please contact me or, if you wish, Craig
Fields directly.)

WHAT IS REQUIRED.  This year, we expect that only two email reports
will be solicited:
	(1) A single annual technical summary.
	(2) Financial summaries.

NOTE: These email reports are in addition to the reports required as
contractual deliverables, and this informal request does not waive any
contractual requirement.

FINANCIAL SUMMARIES.  The format for the financial summaries will be
uniform throughout the entire ISTO community, in order that we can use
automatic means to insert the data you provide into our financial
database.  Details of the format and deadlines will be provided in a
separate message.  PLEASE SEND FINANCIAL SUMMARIES SEPARATELY.

TECHNICAL SUMMARIES.  The technical summaries provide us with an
up-to-date view of the state of activity in our community.  The
challenge is to be concise yet substantive.  Responses are needed by
the morning of 27 July.

PLEASE ADHERE to the format below.  Specifically, (1) do not include
markup commands from a text processor (except in formulas that are
especially complex), and (2) do not use leading indentation or any
other extra embedded horizontal whitespace.  I suggest grabbing this
text with a text editor and filling in the blanks.

Here is what is required:
================================================================

(1) BASIC DATA.

(1.a) DARPA/ISTO project code, ARPA Order number, agent, contract number.

Example: AA, 1111, SPAWAR, N0037-C-0004
The first of these is the "task code" you were just sent.  (Call
Nicole Jacobs if you haven't received this.)

(1.b) Institution.

The institution contracting with the government.

(1.c) Project title. 

One line project title.

(1.d) Project mailbox.  

An internet address for official project email.  This should be active
even when PIs are travelling.  At worst, we will accept a list of
addresses. 

(1.e) PIs. 

For each: name, phone numbers, email address.

(2) PROJECT SUMMARY.

A short paragraph (approximately four sentences) outlining the
specific need for your effort, the opportunity it represents, and the
results to be delivered.  Your summary should indicate the expected
impact of the project, i.e., how future technologies will be different
because of this investment by DARPA.

(3) APPROACH.

A short paragraph describing the overall approach taken to research.
That is, what actual steps will be taken to achieve research goals?

(4) ACCOMPLISHMENTS. 

Two to four concrete accomplishments for fy89.  Emphasize technical
results with externally recognizable impact.  If you achieved a major
milestone of broad community impact, indicate so and describe it.
(There should be such an event every year or two.)

(5) OBJECTIVES.

Two to four for fy90, each as a short (two to three sentence)
paragraph.  Emphasize both technical results and impact.  Be specific.

(6) TECHNOLOGY TRANSFER progress and plans.  

(6.a) Customers. 

Customers for your results, whether the results be theorems or
software components or software interface definitions.  (Include
potential users outside of your research group with whom you have had
discussions as well as actual current users.  Include both research
groups and development groups.)  This and the other responses in this
section should be only two or three sentences each.

(6.b) Interfaces and consensus.

Opportunities pursued for accelerating community consensus, where
needed and where possible, for systems interface definitions that your
work creates, contributes to, and/or depends on.  (Any producer of a
large system prototype should address this issue.  At the research
prototype stage, efforts involving collaboration among research groups
to agree on component interfaces should be described.)

(6.c) Sources of technology.

Major producers of technology that you rely upon, including technical
results, interface definitions, and systems components.  (Do not
mention common commercial components.  You might mention Mach or the
Boyer-Moore theorem prover, but you wouldn't mention DEC or Sun
workstations.)

(7) OTHER INFORMATION.

(7.a) Major personnel changes.   

New hires, departures, etc.

(7.b) Major recent publications, honors, etc.

Send us copies of important publications once in a while.

================================================================

WHAT TO DO.

One report should be sent for each independently funded effort (i.e.,
each project code).  If you receive multiple copies of this message, it
is probably because you are involved with multiple funded efforts.  If
there is any question about what constitutes an "independently funded
effort," please call Nicole or me.

Send reports by email directly to Nicole Jacobs at DARPA:
	njacobs@vax.darpa.mil		(202)694-5800

If you do not receive an acknowledgement within a week, call Nicole to
ensure she has received your message.

Follow the format guidance for your responses.

MAILING ADDRESSES.  Help us keep our mailing list current.  We often
send official correspondence by email.  The best way to do this is for
you to have a virtual project mailbox at your site that will forward
to PIs and other responsible people.  Again, if there is a question,
call Nicole and check to see what address we have on file.

Deadline: Please respond by the morning of 25 August.  Early responses
are requested.  If you MUST respond late, call us first.  Thanks.

Finally: Many of your contracts require regular reports to be sent to
the contracting agent and to certain parts of DARPA.  Please also send
one copy of the reports directly to me at DARPA.

Thanks,
				Brian
-------


-------

∂15-Aug-89  0846	CLT 	research interests  
I can't do this

    ∂15-Aug-89  0829        TAJNAI@Score.Stanford.EDU       John's research interests   
    Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 15 Aug 89  08:29:12 PDT
    Date: Tue 15 Aug 89 08:27:21-PDT
    From: Carolyn Tajnai <TAJNAI@Score.Stanford.EDU>
    Subject: John's research interests
    To: clt@Sail.Stanford.EDU
    Message-ID: <12518331252.9.TAJNAI@Score.Stanford.EDU>


    Carolyn, I would appreciate your checking John's research interests.


    @b{John McCarthy},
    @i[Charles M. Pigott Professor of Engineering, Professor of Computer
    Science and (by courtesy) Electrical Engineering,  Ph.D. Princeton, 1951].@\
    Research Interests@i(:)  Artificial intelligence, computing with
    symbolic expressions, time sharing, formalizing common sense, non-monotonic
    logic.
    One of the founders of artificial intelligence research,  Professor 
    McCarthy invented
    LISP, the programming language most used in AI research and also
    first proposed the general purpose time-sharing mode of using computers.
    The emphasis of his AI research has been in identifying the common
    sense rules that determine the consequences of actions and other events,
    the expression of such rules and other common sense information as
    sentences in logical languages in the data bases of artificial intelligent
    programs.  @comment[He has also worked on formalizing common sense reasoning,
    and his recent work concerns non-monotonic common sense reasoning
    whereby people and computers draw conjectural conclusions by assuming that complications are absent from a situation.  ]
    A past president of AAAI,
    Professor McCarthy received the First IJCAI Research Excellence Award in 1985.

    @begin[itemize]
    @include[x-ai.lifschitz]
    @include[x-ai.mason]
    @include[x-ai.talcott]
    @include[x-ai.weening]
    @end[itemize]
    -------

∂15-Aug-89  1341	pjd@riacs.edu 	re: Text of signed Viewpoint  
Received: from icarus.riacs.edu by SAIL.Stanford.EDU with TCP; 15 Aug 89  13:39:12 PDT
Received: from hydra.riacs.edu by icarus.riacs.edu (5.59/2.1G)
	   id AA07796; Tue, 15 Aug 89 13:40:25 PDT
Received: by hydra.riacs.edu (4.12/2.0N)
	   id AA20235; Tue, 15 Aug 89 13:40:38 pdt
Message-Id: <8908152040.AA20235@hydra.riacs.edu>
Date: Tue, 15 Aug 89 13:40:38 pdt
From: Peter J. Denning <pjd@riacs.edu>
To: JMC@SAIL.Stanford.EDU
Subject: re: Text of signed Viewpoint
Cc: jimm@acmvm.bitnet

John, I've looked over the text and have these comments/suggestions:

pjd@riacs.edu
your comments
As you will see from below, I don't agree with most of them.  However,
I have made a few changes in the text to try to explain a few things
better, especially the reference to politics.  The next message
will be the revised text.

1.  Your complaint seems to be with the addressing schemes and low
density of ports for email, not with email itself.  Your argument
seems to be this: In the current world, there are relatively few
termini for computer networks as compared with telephones.  And
when you do connect your computer to a terminus the addressing
scheme is difficult to fathom or obtaining the address is
difficult.  The current practices surrounding use of telephones are
well understood and people can more easily begin to use email if it
were based on the phone network.  Correct?

ans: It isn't merely that the procedures are better understood.
I understand electronic mail quite well, having used it for
20 years.  However, I can now reach local contractors and
lawyers and Japanese hotels by fax.

2.  What do you mean by "politics"?  Do you mean the need for
negotiation among human beings and the reaching of agreements?
What's wrong with that?

ans: I prefer leaving the politics to the International
Telecommunications Union.  Eligibility for membership in various
noncommercial nets has been and must be determined by politics,
because there are always marginal cases which require negotiation
and influence.  For example, the inclusion of IBM on Arpanet to
several years to negotiate.  Exactly what academic institutions
were eligible for ARPAnet was problematical.  I bet there are
problems about what institutions are eligible for NSF's network.
This contrasts with fax.  All you have to do is buy a machine.

3.  The title does not describe the content very well.  How about:
"The birth and death of electronic mail"?  That will seduce the
reader.

ans: I prefer the present title.  I am not expecting electronic
mail to die, only to be superseded for most communication that
has to be available to most organizations.

4.  You refer to a mistake made by DARPA around 1970, making the
ARPANET a vehicle for electronic mail.  As I understand their
intentions at the time, they were NOT trying to provide a vehicle
for electronic mail.  Even as late as 1980, NSF was resisting the
idea of selling networks (e.g., CSNET) based on electronic mail
benefiting the community.  Electronic mail snuck in and, I think,
surprised everyone with how much people liked it.

ans: I remember electronic mail as having been important from the
beginning of ARPAnet.  Even if you're right about the history, the
mistake was in not providing a set of telephone protocols for
electronic mail.

5.  The Research Internet is designed to take care of part of the
problem of proliferation of networks.  No matter which network
you are connected to, the same protocols will allow you to interact
with anyone or any machine anywhere in any of the networks making
up the Internet.

ans: That is an improvement, although I don't know if any of my
troubles in reaching people have been protocol incompatibilities.
However, even if the protocols are compatible, the addresses may
still be exotic.  Communication between Internet and non-research
people will be difficult, and communication with people on
commercial networks that require payment is particularly
problematical.

6.  The telephone companies are interested in becoming the providers
of ubiquitous "information infrastructure".  The phone jack in your
house will be the end of a fiber and will serve as a high speed
data/voice/video link, providing myriad information and communication
services.  Some of them are also talking about "universal phone numbers"
which are yours and will allow your calls to reach you no matter where
in the world you are.  These trends will eventually supplant today's
email and fax systems, don't you think?

ans: Very likely, if providers of preent services that will be superseded
aren't successful in preventing it by law.  ``anywhere in the world''
means that an awful lot of interests may have power to prevent or
delay the millenium.  I think the millenium is unlikely to be reached
in this millenium.

Please revise your commentary as you see fit based on these comments
and send the result to me with copy to Jim Maurer.  Thanks.  We'll
publish as early as possible during late fall.

Peter


∂15-Aug-89  1813	CLT 	labrea account 
Since file space is going to be our own, it seems silly
to keep the paying for the labrea account.

∂16-Aug-89  0832	MPS 	dentist   
Good morning

You have an appointment on Monday, August 28th at 11:00

Pat

∂16-Aug-89  1010	CLT 	labrea account 
Did you delete any files from Sail on the theory that
they were available at labrea?  If so we should retrieve them.
Otherwise I will just cancel the account.

∂16-Aug-89  1350	siegman@sierra.STANFORD.EDU 	Electronic Mail on Networks    
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 16 Aug 89  13:50:45 PDT
Received: by sierra.STANFORD.EDU (3.2/4.7); Wed, 16 Aug 89 13:49:29 PDT
Date: Wed, 16 Aug 89 13:49:29 PDT
From: siegman@sierra.STANFORD.EDU (Anthony E. Siegman)
To: jmc@sail
Subject: Electronic Mail on Networks 
Message-Id: <CMM.0.88.619303767.siegman@>

Why don't you send your very interesting recent message concerning
email on networks to comp.dcom.telecom?  Address for sending to that
newsgroup is

     telecom@eecs.nwu.edu

I've also send hard copies to Kumar Patel and Bob Lucky at AT&T Bell
Labs.

∂17-Aug-89  0157	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	angels on pinpoints
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 17 Aug 89  01:57:17 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 411303; 17 Aug 89 04:55:47 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 106489; Thu 17-Aug-89 01:49:34 PDT
Date: Thu, 17 Aug 89 01:48 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: angels on pinpoints
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890817084853.9.RWG@TSUNAMI.SPA.Symbolics.COM>

That you nearly always want 0↑0 = 1 is a special case of 1 being the
empty product:
                      - 1
                     /===\
                      ! !
(D1109)               ! !  Anything(k)  =  1 .
                      ! !
                     k = 0

(Analytically, it's also a manifestation of Limit Zero1(z)↑Zero2(z)
                                            z->0
requiring a much "stronger" Zero1 to get anything but 1.)

An interesting "counterexample" is the n = 0 case of the (old) identity

               n - 1
               /===\
                ! !             k     sin(π n z)
(D1110)         ! !  sin(π (z + -)) = ----------
                ! !             n        n - 1
               k = 0                    2

which seems to say that if you scream loud enough, you'll be
heard, even if there are none of you.

∂17-Aug-89  0909	CLT 	gabriels salary

According to the charts (SRA level with 17 years experience)
we can only pay $68k without further justification.
Could you write a recommendation for higher salary (75-80k)
based in rpgs ``world-class'' stature (to quote rpg).
More quote from rpg 
    Did you know that I have 26 scientific publications since 1982 and
    probably 25 pseudo-scientific ones [encyclopedia articles and
    popularization of Lisp and OOP stuff]? JMC didn't know that. Did you know
    that I am asked to write tenure letters for schools like CMU and MIT [and
    some others]?)

Yvette needs the letter ASAP as she has to get the forms in today.

∂17-Aug-89  0958	C.COLE@MACBETH.STANFORD.EDU 	re: O.T.A. message to Su-etc   
Received: from MACBETH.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 17 Aug 89  09:58:23 PDT
Date: Thu 17 Aug 89 09:55:48-PDT
From: George Cole <C.COLE@MACBETH.STANFORD.EDU>
Subject: re: O.T.A. message to Su-etc  
To: JMC@SAIL.STANFORD.EDU
In-Reply-To: <1zKp8P@SAIL.Stanford.EDU>
Message-ID: <12518871640.83.C.COLE@Macbeth.Stanford.EDU>

As far as products liability goes, the law clearly distinguishes between 
liability and blame. The basic rule of strict liability is the allocation of
the risk to those best able to allocate the cost and control the risk.
	You can bring a products liability action alleging negligence (in 
either design or manufacture) that created the defect that injured the
plaintiff. The inquiry focuses on the behavior of the manufacturer. But in
turn the manufacturer is now allowed to look at the behavior of the plaintiff
and to get a reduction for the percentage of negligence shown by the plaintiff.
(And, because of what Prop. 51 became, the allocation of non-economic damages
is strictly by percentage of fault.)
	Or you can bring a products liability action alleging a defective
product. The focus then is on the product and the feasibility of alternatives.
The law in California has evolved two prongs of testing: whether or not the
product performed as safely as the average consumer expected; and, whether or
not the product's benefit outweighed its cost. The latter can be viewed with
hindsight in that the manufacturer can be charged with knowledge that he did
not in fact have, but could have had.
	It's a bit more restricted with "warning" cases, or with prescription
drug cases. The courts have hinted that if there was no way to find out the
information, you are not responsible for not warning people. This is due to
the bulge of asbestos cases in Alameda county working their way through in
a consolidated mass. And the court has limited the liability for drug cases
specifically because of the public policy interest in getting new drugs.

	Of course as a lawyer serving a client, you initially would expect to
file a suit claiming both liability and blame. You might well ask a jury to
determine which, or both, applies and hope they don't find that neither 
applies. That's because most of the determination depends on factual decisions
which the jury gets to make. So you posit the best theories and find out what
the truth actually was. It's this having to predicate a theory based on truths
that will later be resolved that made understanding Doyle's A.T.M.S. so 
natural for me. When you start adding the rules about how an appellate court
has to look at the facts, it gets even more fun. (Different rules apply 
depending on the procedural nature of the case before the appellate court.)

	The economic damages (medical and lost wages) are pretty substantial.
In fact, they are actually at a range where the expense of a major trial (20 -
30 days) and experts' assessments (assume $200k for all parties) begin to make
sense. It would still be better if a good settlement could be structured, but
I don't think that's going to happen. The barrier of attorney egos and outside
concerns seems to be too high. The attorney for Toyota reportedly answered the
discovery judge's comment that Toyota had substantial exposure with "That's 
o.k.; we can handle that." I wasn't there, I didn't get the inflections, but
the best construction I can put on it is that Toyota has the cash to pay a
large exposure. The worst is that the law firm is betting on its ability to
win but risking the damage to its client.

	The "foreseeability" test, incidentally, only applies when you are
bringing a "negligence" cause of products liability. Due to the wide range
of factual situations it actually makes sense to have several doctrines. What
I'm personally interested in is the way the law might develop to cover software
products, particularly AI and ES programs. There is so MUCH that could be done
that costs so very little but reduces both the risk to the ultimate user and
to the company of facing a major, disastrous suit. Sigh. And theculture
is such that I fear a number of "bombs" are being built in. What happens,
for example, when the century turns over and all those wonderful COBOL programs
that didn't leave enough room for the years digits start calculating interest
from 1900? And trigger the "overdue" notices?
	I'll be happy to keep you informed as developments occur. The month of
September starts deposition-frenzy as we depose Toyota's experts, then they
depose our experts. And the responses to our discovery requests and battles
over other questions should be resolved with more information coming in.

George
-------

∂17-Aug-89  1324	MPS  
Beverly Huff, 324-4456, would like to show your house
today.  She would like you to call her.

Pat

∂17-Aug-89  1428	cvenkat@iag.hp.com 	TeleFax - Suggestions    
Received: from sde.hp.com (sde.sde.hp.com) by SAIL.Stanford.EDU with TCP; 17 Aug 89  14:28:25 PDT
Received: from kitzbuhl.iag.hp.com by hp-sde.sde.hp.com with SMTP
        (15.7/SES42.42) id AA22431; Thu, 17 Aug 89 14:29:02 pdt
Received: by kitzbuhl.iag.hp.com
	(15.7/15.4) id AA04754; Thu, 17 Aug 89 14:28:57 pdt
Received: from Version.6.23.N.CUILIB.3.44.SNAP.NOT.LINKED.hpiagcv.HP9000.360
          via MS.5.5.hpiagcv.hpux;
          Thu, 17 Aug 89 14:28:54 -0700 (PDT)
Message-Id: <8YumsKUD0O2L0GXkc0@iag.hp.com>
Date: Thu, 17 Aug 89 14:28:54 -0700 (PDT)
From: Chandra Venkatraman <cvenkat@iag.hp.com>
To: jmc@sail.stanford.edu
Subject: TeleFax - Suggestions
Cc: cvenkat@hpda.hp.com

Dear Prof. McCarthy,
   I have been working on Fax Servers, (inspired by the elegance of fax as an
excellent point to point human communication mechanism) for a few years now.
After a presentation I made yesterday to our group (Information Architecture
Group, HP) on  'Implementation of Fax Servers and Integration with E-mail',
John McGrory was kind enough to forward me the "guest viewpoint" message you
had posted on the Stanford Network.  I totally concur with your observations.

  My suggestion is, to include the reasons for the telefax popularity in your
article.  The following are some of the reasons why,  I think, the telefax has
become popular.

    0. (most important)  Since the output at the received end is interpreted by
a human being, fax seems to be the most efficient way.  If I were to send data
to an application (e.g. bank money transfer), I will require a more robust,
machine understandable version.  IT is MAIL FROM A PERSON TO A PERSON; not a
machine to machine.

  1. It is easy to use; as simple as to use a telephone.

  2. Universal connectivity.  (For a long time to come, fax may be the best (or
even ONLY) way to reach under-developed and developing countries)

  3. Graphics and images. (Picture is worth more than a thousand words,
especially when communicating between different cultures (both business and
societal))

  4. Cost, both terminal and transmission. (Japanese Industry target to bring
down the cost of fax machines to about US$200.  They anticipate 10% of all
homes to have a fax machine by 1995 and about 50% by 2010).

  5. International Standards  (Grey level and color image fax standards are
being currently worked on by CCITT study groups).

I am looking forward to reading your article in the near future. Good luck.

Chandrasekar Venkatraman
Information Architecture Group
Internet: cvenkat@iag.hp.com
Phone: (408) 447-1396

∂17-Aug-89  1459	VAL 	your letter    
I would appreciate if you could check if and when the receipt of your letter
to Manchester was acknowledged--just to get a better idea of what's going on
there.

∂17-Aug-89  1632	telecom@eecs.nwu.edu 	Receipt: NETWORKS CONSIDERED HARMFUL - FOR ELECTRONIC MAIL
Received: from eecs.nwu.edu (listposter.eecs.nwu.edu) by SAIL.Stanford.EDU with TCP; 17 Aug 89  16:27:51 PDT
Date:     Thu, 17 Aug 89 18:27:41 CDT
From:     telecom@eecs.nwu.edu
Subject:  Receipt: NETWORKS CONSIDERED HARMFUL - FOR ELECTRONIC MAIL
To:       John McCarthy <JMC@sail.stanford.edu>
Message-ID:  <8908171827.aa05671@delta.eecs.nwu.edu>

=> Message sent to: telecom@eecs.nwu.edu
=> Dated: 17 Aug 89  1619 PDT
=> Subject: "NETWORKS CONSIDERED HARMFUL - FOR ELECTRONIC MAIL"
=> Received by Telecom Moderator: Thu Aug 17 18:27:39 CDT 1989
=> Message size: 9326 bytes 

This is an automatic reply from TELECOM Digest. Your recent correspondence
was received here per the information shown above. Because the volume of
mail received is frequently quite heavy, this computer generated response
is used to speed a reply to you. It is not intended to lessen the value
of your correspondence. 

If you submitted an item for publication in TELECOM Digest (comp.dcom.telecom)
it will appear in a day or two, or be returned to you with an explanation why
it cannot be used. 

If you requested an addition/deletion/change of address to the mailing list,
the change will be processed within a few hours. If you do not see the desired
change in place within a day or two, please write again, noting 'second
request' in your letter.

If your letter requires a personal answer, I will respond as soon as possible.
Letters of general interest to Digest readers marked 'not for publication' will
be read and considered, but time constraints make these letters difficult for
me to answer in a timely fashion. *Please reconsider allowing it to be printed
in the Digest*.

Patrick Townson / TELECOM Digest Moderator / Usenet's comp.dcom.telecom group

∂17-Aug-89  1755	rpg@lucid.com 	publications   
Received: from lucid.com by SAIL.Stanford.EDU with TCP; 17 Aug 89  17:55:24 PDT
Received: from challenger ([192.9.200.17]) by heavens-gate id AA02615g; Thu, 17 Aug 89 17:55:52 PDT
Received: by challenger id AA03559g; Thu, 17 Aug 89 17:53:06 PDT
Date: Thu, 17 Aug 89 17:53:06 PDT
From: Richard P. Gabriel <rpg@lucid.com>
Message-Id: <8908180053.AA03559@challenger>
To: jmc@sail.stanford.edu
Subject: publications


\centerline{\bf Publications}
\vskip .5truein

\noindent{\bf Miscellania} Approximately two dozen popular articles on Lisp,
programming languages, and programming environments. 1985--1989.

\noindent{\bf Object-Oriented Programming and the Common Lisp Object System},
AI Expert, February, 1989.

\noindent{\bf Requirements for a Common Prototyping System}, primary
author as Editor of the Common Prototyping Working Group, November,
1988.

\noindent{\bf Technical Issues of Separation in Function Cells and Value
Cells} (with Kent Pitman), in Lisp and Symbolic Computation, Vol.~1
No.~1, 1988.

\noindent{\bf Qlisp: Parallel Processing in Lisp} (with Ron Goldman),
1989 Hawaii International Conference on Systems Sciences, Kailua-Kona,
Hawaii, 1989.

\noindent{\bf Preliminary Results with an Initial Implementation of Qlisp}
(with Ron Goldman), 1988 Conference on Lisp and Functional Programming,
Snowbird, Utah, 1988.

\noindent{\bf Qlisp: Experience and New Directions} (with Ron Goldman),
1988 ACM Symposium on Parallel Programming: Experience with Applications,
Languages, and Systems, New Haven, Connecticut, 1988.

\noindent{\bf Qlisp} (with John McCarthy),
``Parallel Computation and Computers for Artificial Intelligence,''
J. S. Kowalik, editor. 1988.

\noindent{\bf The Common Lisp Object System: An Overview}
(with Linda DeMichiel), Proceedings of the First European Conference
on Object-Oriented Programming, Springer-Verlag, 1987.

\noindent{\bf The Common Lisp Object System Specification} (primary author,
with Daniel Bobrow, Linda G. DeMichiel, Sonya Keene, Gregor Kiczales,
and David Moon), Technical Document 87-002 of X3J13. March 1987.

\noindent{\bf Lisp}, an article in the Encyclopedia of Artificial
Intelligence, Stuart C. Shapiro, Editor-in-Chief. 1987.

\noindent{\bf Used Software}, Proceedings of the Aerospace Applications of
Artificial Intelligence Conference, Dayton, Ohio. October 1986.

\noindent{\bf Design of an Optimizing, Dynamically Retargetable Compiler},
(with Rodney Brooks, David Posner, James McDonald, Jon L. White, and
Eric Benson), Proceedings of the 1986 ACM Conference on Lisp and
Functional Programming, Cambridge, Massachusetts. August 1986.

\noindent{\bf Massively Parallel Machines: The Connection Machine and 
NON-VON}, SCIENCE, Volume 231, pp. 975-978, 28 February, 1986.

\noindent{\bf Common Lisp: Its History, Its Uses, and Its Future},
Proceedings of the AI-85 Conference, Long Beach, California. May 1985.

\noindent{\bf Performance and Evaluation of Lisp Systems}, (book),
The MIT Press, Cambridge, Massachusetts. 1985.

\noindent{\bf Queue-based Multi-processing Lisp} (with John McCarthy),
Proceedings of the 1984 ACM Symposium on Lisp and Functional Programming,
August 1984.

\noindent{\bf A Critique of Common Lisp}
(with Rodney Brooks), Proceedings of the
1984 ACM Symposium on Lisp and Functional Programming, August 1984.

\noindent{\bf Deliberate Writing,} an invited chapter in ``Natural Language
Generation Systems,'' edited by Leonard Bolc, published by Springer-Verlag,
1988

\noindent{\bf A Programming Environment for a Timeshared System} (with
Martin Frost), Proceedings of the 1984 Software Engineering Symposium
on Practical Software Development Environments.

\noindent{\bf Lisp-in-Lisp: High Performance and Portability} (with
Rodney A. Brooks and Guy L. Steele Jr.), Proceedings of the 1983
International Joint Conference on Artificial Intelligence, August
1983.

\noindent{\bf An Optimizing Compiler for Lexically-Scoped Lisp} (with
Rodney A. Brooks and Guy L. Steele Jr.), SIGPLAN 82 Symposium on
Compiler Construction, June 23--25, 1982, in Boston, Massachusetts.
Sponsored by the Association for Computing Machinery.

\noindent{\bf Lisp Evaluation and Timing} (with Larry M. Masinter of
Xerox PARC), 1982 ACM Symposium on Lisp and Functional Programming,
August 15--18 1982, in Pittsburgh, Pennsylvania.  Sponsored by the
Association for Computing Machinery.

\noindent{\bf S-1 Common Lisp Implementation}
(with Rodney A. Brooks and
Guy L. Steele Jr.), 1982 ACM Symposium on Lisp and
Functional Programming, August 15-18 1982, in Pittsburgh, Pennsylvania.
Sponsored by
the Association for Computing Machinery.

\noindent{\bf An Organization for Programs in Fluid Domains},
Dissertation, Computer Science Department, STAN-CS-81-856, AIM-342
Stanford University.

\noindent{\bf Results in Knowledge-based Program Synthesis},
(Green, Gabriel et al), 6th IJCAI, 1979.

\noindent{\bf Computer-Assisted Diagnosis of Orthopedic Gait Disorders}
(with Kathleen B. Tracy et al),
PHYSICAL THERAPY March 1979.

\noindent{\bf Natural Language-based Information Retrieval} (with D.
L.  Waltz), Twelfth Annual Allerton Conference Proceedings, October
1974.

\noindent{\bf Structured Descriptions}, MIT AI Lab Working Paper. August 1973.

\vskip .3truein
\sectfilbreak
\centerline{\bf Invited Papers and Lectures}
\vskip .3truein

\noindent{\bf The Common Lisp Object System: An Overview}
(with Linda G. DeMichiel), an invited paper by the European
Conference of Object-Oriented Programming (Ecoop'87). June 1987.

\noindent{\bf Deliberate Writing,} an invited lecture at the Workshop on
Natural Language Generation, August 15--17 1983 at Burg Stettenfels, West
Germany, sponsored by the University of Stuttgart.

\noindent{\bf Panel Discussion on the Performance of Lisp Systems}.
Led a panel discussion with participants: Larry Masinter, Daniel
Weinreb, Scott Fahlman, Richard Greenblatt, John L. White, and Jock
MacKinlay, at the American Association of Artificial Intelligence
Conference, August 24, 1983, Washington D.C.

\noindent{\bf Workshop Lecture on Parallel Lisp Dialects}. Purdue Workshop
on Supercomputers and Symbolic Processing, May 3--4 1984, Purdue University.

\noindent Other publications include reports on programs and research performed
for ONR at Parke Mathematical Laboratories in the field of underwater acoustics
and are no longer available.


∂17-Aug-89  2301	JMC  
iii bill

∂18-Aug-89  0653	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	angels on pinpoints
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 18 Aug 89  06:52:51 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 411760; 18 Aug 89 09:48:07 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 106613; Fri 18-Aug-89 06:32:38 PDT
Date: Fri, 18 Aug 89 06:31 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: angels on pinpoints
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890817084853.9.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890818133156.6.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Thu, 17 Aug 89 01:48 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

    That you nearly always want 0↑0 = 1 is a special case of 1 being the
    empty product:
                          - 1
                         /===\
                          ! !
    (D1109)               ! !  Anything(k)  =  1 .
                          ! !
                         k = 0

    (Analytically, it's also a manifestation of Limit Zero1(z)↑Zero2(z)
                                                z->0
    requiring a much "stronger" Zero1 to get anything but 1.)

    An interesting "counterexample" is the n = 0 case of the (old) identity

                   n - 1
                   /===\
                    ! !             k     sin(π n z)
    (D1110)         ! !  sin(π (z + -)) = ----------
                    ! !             n        n - 1
                   k = 0                    2

    which seems to say that if you scream loud enough, you'll be
    heard, even if there are none of you.

Neat little exercise:  use this to derive

                               n
                              [-]
                               2
                             /===\
                              ! !      j pi           n
(D1341)                       ! !  sin(----) = Sqrt ------
                              ! !       n            n - 1
                             j = 1                  2
                                               
n>0, []:=Floor.  (Try a few n !)

(D1110) also leads to the peculiar result

         n - 1
        [-----]
           2
         /===\                                             k + 1
          ! !           2 k pi        2 j pi          (- 1)      n
(D1337)   ! !    2 (cos(------) - cos(------)) = -----------------------
          ! !             n             n               k pi       2 k pi
         j = 1                                   4 sin(------) sin(------)
         j # k                                             n         n
                                                       Num(-)
							   2

(where Num := Numerator), which is useful in specializing the partial fractions
identity

                      1
(D1339) ------------------------------ = 
          n
        /===\
         ! !                      2
         ! !  (1 - 2 z cos(a ) + z )
         ! !                k
        k = 1

                     n
                    ====                 sin(a  n) - z sin(a  (n - 1))
                    \                         k             k
                     >    ------------------------------------------------------------
                    /                                      n
                    ====                                 /===\
                    k = 1                     2           ! !
                          (1 - 2 z cos(a ) + z ) sin(a )  ! !  2 (cos(a ) - cos(a ))
                                        k             k   ! !          k         j
                                                         j = 1
                                                         j # k

to the case where the angles form a regular polygon:

          1
(D1340) ------ = 
             n
        1 - z

                       2 f k pi          2 (f - 1) k pi        k pi
                  (sin(--------) - z sin(--------------)) sin(------)
            f             n                    n                  n
           ====                                               Num(-)
           \                                                      2
         4  >     ---------------------------------------------------
           /                    k + 1              2 k pi     2
           ====            (- 1)      (1 - 2 z cos(------) + z )
           k = 1                                     n
         -------------------------------------------------------------
                     n (1 - z) (If Oddp(n) Then 1 Else 1 + z)

where n > 2 and f := Floor((n-1)/2).  This allows indefinite
integrals of the form poly(x)/((1+-x↑a)(1+-x↑b)...) to
be integrated in real terms, although swamped in scads of
terms like sin(69 pi/239).

Note that in D1339, the cos a[k] are arbitrary, and if replaced
by c[k], show up on the rhs in Chevychase polynomials of the
2nd kind.

∂18-Aug-89  0900	JMC  
Anne Diesel, Hoover 3-6594 re contract

∂18-Aug-89  1046	dgc@math.ucla.edu 	Re -- Networks Considered Harmful - For Electronic Mail 
Received: from math.ucla.edu (sunset.math.ucla.edu) by SAIL.Stanford.EDU with TCP; 18 Aug 89  10:46:43 PDT
Return-Path: <dgc@math.ucla.edu>
Received: from sonia.math.ucla.edu by math.ucla.edu (Sendmail 5.59/1.02)
	id AA13980; Fri, 18 Aug 89 10:47:15 PDT -0800
Received: from loopback by sonia.math.ucla.edu (Sendmail 3.2/1.02)
	id AA06872; Fri, 18 Aug 89 10:47:08 PDT
Message-Id: <8908181747.AA06872@sonia.math.ucla.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: Re -- Networks Considered Harmful - For Electronic Mail
Date: Fri, 18 Aug 89 10:47:06 PDT
From: dgc@math.ucla.edu

I read your article on the above subject and throughly agree with most
of what you say.

Some minor points, however:

1. At the moment, internet and bitnet email is FREE, from the point of
   view of users like you and me.  I'm working from home with a local
   call to UCLA over a 9600 baud modem and I can reply to you at no
   charge that I see.  I don't even know who is paying for the various
   networks once outside of my own Department.

   On the other hand, with your system I could shoot this message to you
   in less than one minute and, even at daytime rates, that wouldn't be
   much money.  But it's not the same!

2. What you are proposing is, in many ways, simply a digital version
   of "voice mail" already provided by many telcos, etc.

3. You are VERY parochial when you say:

	It would be good if the ACM were to set up a committee to adopt
	a telephone electronic mail standard . . .

   Why should the ACM be the responsible party?  I think the record
   of the "Computer Science Community" in sharing resources, setting
   standards, etc., is abysmally poor and, in fact, most email users
   (internet, bitnet, etc.) are NOT Computer Science people.  In the
   past, organizaations like the IEEE have set up standards of this sort
   and they are undoubtedly more qualified than CS people.

4. Your comments about the importance of not requiring a "login" are not
   relevent to your argument and probably weaken it.  Indeed, "login"
   is simply a way of setting up communication, identifying the caller,
   etc., and, in general, does NOT require previous arrangement.  All
   communication systems that I have ever encountered use some sort
   of "login".  In the USA, the typical voice telephone login (e.g.,
   when I call you) begins "Hello", "hello", "This is Dave, may I speak
   to John", etc., etc. Ftp, which is supported by many unix systems
   allows for an "anonymous" login.  More than that, the underlying mail
   system for most unix systems doesn't even require a formal login of
   any sort.  Try

   	telnet <your-favorite-host> 25

dgc

David G. Cantor
Department of Mathematics
University of California at Los Angeles
Internet:  dgc@math.ucla.edu
UUCP:      ...!{randvax, sdcrdcf, ucbvax}!ucla-cs!dgc

∂18-Aug-89  1531	root@Gang-of-Four.Stanford.EDU     
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Aug 89  15:30:49 PDT
Received: from sail by Gang-of-Four.Stanford.EDU (5.0/25-eef) id AA08107; Fri, 18 Aug 89 15:11:41 PDT
Date: Fri, 18 Aug 89 15:09:32 PDT
From: Charlie Root <root@Gang-of-Four.Stanford.EDU>
Message-Id: <8908182211.AA08107@Gang-of-Four.Stanford.EDU>
Apparently-To: jmc

stuff and nonsense

∂18-Aug-89  1539	ullman@nimbin.Stanford.EDU 	lunch? 
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Aug 89  15:39:21 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA03295; Fri, 18 Aug 89 15:36:58 PDT
Date: Fri, 18 Aug 89 15:36:58 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908182236.AA03295@nimbin.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: lunch?

Do you want to have lunch one day next week and
talk about "Theory vs. AI"?
				---jdu

∂18-Aug-89  1621	CLT 	research interests  

Did you do the update for Carolyn Tajnai?

∂18-Aug-89  1657	ullman@nimbin.Stanford.EDU 	re: lunch?  
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Aug 89  16:57:19 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA03349; Fri, 18 Aug 89 16:54:52 PDT
Date: Fri, 18 Aug 89 16:54:52 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908182354.AA03349@nimbin.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: re: lunch?

I think any day but Wednesday is OK for me.
I'll tentatively reserve 8/28; let me know if you prefer another.
				---jdu

∂18-Aug-89  1726	ole@csli.Stanford.EDU 	Re: Networks considered hamful - for email
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Aug 89  17:25:58 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA06103; Fri, 18 Aug 89 17:27:07 PDT
Date: Fri 18 Aug 89 17:27:06-PDT
From: Ole J. Jacobsen <OLE@CSLI.Stanford.EDU>
Subject: Re: Networks considered hamful - for email
To: JMC@sail.stanford.edu
Message-Id: <619489626.0.OLE@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Phone: (415) 941-3399 (ACE)  (415) 325-9427 (Home)  (415) 990-9427 (Cellular)


John,
	Enjoyed your posting very much, in fact I'd love to get your
permission to print it in ConneXions, our monthly newsletter on
interoperability topics, TCP/IP, OSI, X-windows, email etc. Perhaps
you'd like to expand it into a slightly larger article? Let me know,
and send me your postal address so that I can send you some samples
of ConneXions if you haven't seen it.

Thanks,
	Ole
-------

∂18-Aug-89  1745	ole@csli.Stanford.EDU 	re: Networks considered hamful - for email     
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Aug 89  17:45:13 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA06579; Fri, 18 Aug 89 17:46:21 PDT
Date: Fri 18 Aug 89 17:46:20-PDT
From: Ole J. Jacobsen <OLE@CSLI.Stanford.EDU>
Subject: re: Networks considered hamful - for email   
To: JMC@SAIL.Stanford.EDU
Cc: OLE@CSLI.Stanford.EDU
Message-Id: <619490780.0.OLE@CSLI.Stanford.EDU>
In-Reply-To: <BOuOl@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Phone: (415) 941-3399 (ACE)  (415) 325-9427 (Home)  (415) 990-9427 (Cellular)

John,
	I will send you some samples. I am confused about CACM though,
the message I was referring to appeared in the Telecom digest on USENET.
Is it also going in CACM??? What am I missing?  And, having said all that
would you perhaps be interestd in doing another version for ConneXions,
when you've seen it of course...?

Ole
-------

∂18-Aug-89  2116	rick@uunet.uu.net 	Re: Networks Considered Harmful - For Electronic Mail   
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 18 Aug 89  21:16:12 PDT
Received: by uunet.uu.net (5.61/1.14) 
	id AB12209; Sat, 19 Aug 89 00:16:52 -0400
Date: Sat, 19 Aug 89 00:16:52 -0400
From: rick@uunet.uu.net (Rick Adams)
Message-Id: <8908190416.AB12209@uunet.uu.net>
To: JMC@sail.stanford.edu
Subject: Re: Networks Considered Harmful - For Electronic Mail
In-Reply-To: your article <telecom-v09i0306m01@vector.dallas.tx.us>
News-Path: wuarchive!texbell!vector!telecom-gateway

I dont know where you got your UUCP information but its totally wrong.

UUCP was an experimental file transfer program. It was not, nor is
it now a mail protocol. Some people happen to use it to distribute mail,
but that is not its function.

UUCP does NOT require a login. That is some people chose to set it up.
Many ATT sites run uucp without logins (with public access, not just
internally)

While you comments are reasonable in general, your specific examples
are quite flawed.

∂18-Aug-89  2136	rick@uunet.uu.net 	re: Networks Considered Harmful - For Electronic Mail   
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 18 Aug 89  21:36:24 PDT
Received: by uunet.uu.net (5.61/1.14) 
	id AA16070; Sat, 19 Aug 89 00:37:05 -0400
Date: Sat, 19 Aug 89 00:37:05 -0400
From: rick@uunet.uu.net (Rick Adams)
Message-Id: <8908190437.AA16070@uunet.uu.net>
To: JMC@SAIL.Stanford.EDU
Subject: re: Networks Considered Harmful - For Electronic Mail

Absolutely. A UNIX site reachable by telephone does not require a login.

Actually, a very common setup is to require a well know login "uucp" or "nuucp"
with no password.

The reason for having the login is to act as a multiplexor rather
than security. Otherwise you need to dedicate a modem/port for
incoming mail. This isnt always economically feasible (witness
the devices they sell to let you have a fax machine and a modem
onthe same line).

Your basic argument is correct, however much of your uucp example
is not somethign required by uucp, but rather something chosen
by the people installing the software. (e.g you could put
an encryption box on your fax to prevent unauthorized users. This is
similar to the login/password requirement of uucp. You could
put one of the fax/modem multiplexing devices in front of your fax.
This would be similar to having a well know login with no password.
Or, you can just start up the protocol when someone calls. This is
also possible, but fairly uncommon. I think its for cost reasons
more than anything else)

∂18-Aug-89  2207	thomas@mvac23.uucp 	Comments on telecom posting   
Received: from udel.edu (louie.udel.edu) by SAIL.Stanford.EDU with TCP; 18 Aug 89  22:07:27 PDT
Received: from louie.udel.edu by louie.udel.edu id aa25926; 19 Aug 89 0:38 EDT
Date: Fri, 18 Aug 89 23:05:20 edt
From: Thomas Lapp <thomas@mvac23.uucp>
Subject: Comments on telecom posting
Message-ID: <8908182305.0.UUL1.3#5131@mvac23.UUCP>
To: JMC@sail.stanford.edu
Reply-To: mvac23!thomas@udel.edu

Here are some comments I have on your recent essay.

>    2. It requires that the message forwarding computer have login 
> privileges on the receiver.  This has resulted in a system of relaying messages

Granted, this is true, but I would say that it also was based on the idea
that you had local neighbors, who had local neighbors, and thus telecom costs
were kept down.  Also, UUCP has the advantage that because a login in
necessary, there is some way of keeping "junk" mail from being received
directly.


>    The solution is to go to a system that resembles fax in that the ``net
> addresses'' are just telephone numbers.  The simple form of the command is just

Well, if you are looking for a standard, that is what X.400 is working
toward: one standard way of addressing electronic mail and coming up with
the software on each machine so that regardless of your destination network
or machine, one address form does it all.  It wouldn't seem illogical to
use <user>@<telephone number> as a valid e-mail address.  (I'd want to put
the country code in the number to make it world-wide unique).

> to the current variety of local area networks (LANs). The most likely way for 
> this to be accomplished is for the makers of fax machines to offer ASCII 
> service as well.  This will obviate the growing practice of some users of fax 
> of printing out their messages in an OCR font, transmitting them by fax,
> whereupon the receiver scans them with an OCR scanner to get them back into 
> computer form.

I like this idea.  It basically is like using the fax machine like a
many optioned modem:  it takes care of moving your digital data (character
stream) over a link and recreating digital on the other end.

>    I must confess that I don't understand the relation between this 
> proposal and the various electronic communication standards that have been 
> adopted like X25 and X400.  I only note that the enormous effort put into these

X.25 is a physical standard: how to get it across the wire.  X.400 is more on
a higher level.  Like deciding what headers go on every message, what is
legal to use, how to deal with acknowledgement of receipt, etc.
                         - tom
==============================================================================
Internet: mvac23!thomas@udel.edu                     | Notice:  System
          or                                         | will be going down
          mvac23%thomas@udel.edu                     | at 4:45pm today for
uucp: {ucbvax,mcvax,psuvax1,uunet}!udel!mvac23!thomas| a disk crash.
Location: Newark, DE, USA                            |
==============================================================================

∂19-Aug-89  0744	ficc!peter@uunet.UU.NET 	Re: Networks Considered Harmful - For Electronic Mail  
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 19 Aug 89  07:44:25 PDT
Received: from ficc.UUCP by uunet.uu.net (5.61/1.14) with UUCP 
	id AA19324; Sat, 19 Aug 89 10:45:07 -0400
Date: Sat, 19 Aug 89 10:45:07 -0400
From: ficc!peter@uunet.uu.net
Message-Id: <8908191445.AA19324@uunet.uu.net>
To: uunet!sail.stanford.edu!JMC@uunet.UU.NET
Subject: Re: Networks Considered Harmful - For Electronic Mail
In-Reply-To: your article <telecom-v09i0306m01@vector.dallas.tx.us>
News-Path: uunet!wuarchive!texbell!vector!telecom-gateway

You missed one other advantage FAX has... you don't have to know how to
type to use the stupid thing. I think this outweighs all the other technical
and political differences by far.
---
Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation.
Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-'
"Optimization is not some mystical state of grace, it is an intricate act   U
   of human labor which carries real costs and real risks." -- Tom Neff

∂19-Aug-89  0900	JMC  
ua to change ticket

∂19-Aug-89  0904	rick@uunet.uu.net 	re: Networks Considered Harmful - For Electronic Mail   
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 19 Aug 89  09:04:41 PDT
Received: by uunet.uu.net (5.61/1.14) 
	id AA29408; Sat, 19 Aug 89 12:05:20 -0400
Date: Sat, 19 Aug 89 12:05:20 -0400
From: rick@uunet.uu.net (Rick Adams)
Message-Id: <8908191605.AA29408@uunet.uu.net>
To: JMC@SAIL.Stanford.EDU
Subject: re: Networks Considered Harmful - For Electronic Mail


If you call a unix site and it says login, then you need to know what
they want you to say for login (usually uucp).

However, some site dont prompt you with login. They  answer the
phone and expect the protocol to start up. It up to the site
how they do it.

∂19-Aug-89  1100	JMC  
old house

∂19-Aug-89  1113	lotus!bobf@uunet.UU.NET 	Re: Networks Considered Harmful - For Electronic Mail  
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 19 Aug 89  11:13:21 PDT
Received: from lotus.UUCP by uunet.uu.net (5.61/1.14) with UUCP 
	id AA04056; Sat, 19 Aug 89 14:14:04 -0400
Received:  by lotus.com (5.52/25-eef)
	id AA22594; Sat, 19 Aug 89 14:12:02 EDT
From: Bob Frankston (BFrankston) <lotus!bobf@uunet.UU.NET>
Message-Id: <8908191812.AA22594@lotus.com>
To: JMC@sail.stanford.edu, telecom@eecs.nwu.edu
Subject: Re: Networks Considered Harmful - For Electronic Mail
Date: Sat, 19 Aug 89 14:11:59 EDT

Though I too am frustrated by the slow adoption of electronic mail,
the reason that FAX is winning at the moment is that it is simply
much much easier to use.  You plunk down $1k or less, plug it into
the wall, and stick a piece of paper in.  That is about the same as
pressing play on a VCR.  (I know that the VCR has all those other
buttons, but I'm pleased that people have learned they can ignore
them instead of having to learn them).  The situation is the same for
voice mail -- you just leave a message.  Never mind that it is a pain
for the listener; most people don't know how much better it should be
in terms of message management.

Electronic mail needs similar ease of use.  That is why I implemented
Lotus Express so that email would be part of my PC in the background
always available.

A second requirement is ubiquity.  Creating a transport and the
subsequent development of SMTP, primitive though it is, was a
necessary step in this direction.  It used existing equipment --
requiring new equipment in 1970 just wasn't feasible.  Even now, the
current phone network is barely up to the task of supporting FAX.
Many machines do not have their own addresses (nee phone number).

The good news is that the telcos are becoming more aware of this.
X.400 provides an email transport and ISDN is a step towards a next
generation phone network that treats voice as one form of data with
X.400 providing a multimedia store and forward capability and X.500
automating directory assistance.

The process will be instantaneous by telco standards (less than 40
years).  Fax can be coopted by treating the FAX machine as an X.400
user agent and treating a FAX phone number as someone's email
address.  The FAX message would arrive as an X.400 body part.  There
is already a service that will take faxes sent to you, OCR them and
deliver them as MCI mail.

In summary, it is taking a lot longer than I'd like but there is an
inevitability of email.  Remember CB?  It still exists but cellular
phones are much more effective for messaging.


Full name: Bob Frankston

∂19-Aug-89  1925	doug@research.att.com    
Received: from arpa.att.com by SAIL.Stanford.EDU with TCP; 19 Aug 89  19:25:22 PDT
From: doug@research.att.com
Date: Sat, 19 Aug 89 22:15:18 EDT
To: JMC@sail.stanford.edu

This went to Dave Farber and the executive committee of CSNET.

Thanks, Dave, for bringing John McCarthy's email manifesto to our
attention.

Regardless of how helpful or indispensable we think our networks are,
John is right.  Life would be better if the specialized networks
weren't necessary for everyday applications that have no special
needs.  Mail is a universal, not a special, need.  The most important
attribute of a universal communication system is universality itself -
a thesis goes back at least to Theodore Vail, who took it as his
guiding motto in building AT&T.  Before Vail, the fact that someone had
a phone was no assurance you could reach that person by phone.  You had
to be on the right network.

So it is with email today, with the added burden of the inordinate
cost of joining a network - in dollars, in software, and in human
resources.  McCarthy tells a painful truth that there is basically
nothing to email beyond auto-dial, auto-answer, and a simple protocol.
The other services one buys by joining the Internet (aliases, name
servers, remote login, file transfer, store-and-forward, gatewaying,
etc.)  should properly be regarded as value-added services.  They have
little to do with the basic task of getting mail through.  In fact they
detract from it, by making the service and the addressing so arcane
that nobody can install, maintain, or use it without a guru on hand.

The Internet turns McCarthy's model on its head.  Far from being a
basic service, email becomes a value-added service on an inherently
much more difficult network designed to promote computing at a
distance.  Joining the Internet to get email is like installing a guest
room so you could entertain when all you really wanted was a mail slot
so you could communicate.

Yet email is the main reason for most members to connect to the
Internet.  The easy value-added service provides the economic support
for the difficult specialized powers of the network.  This is fine for
the network, but it is bad for the country.  Only rich folks can play.
And even their wealth doesn't bring the power they would wish:  they
cannot through the Internet obtain the attention of anybody outside
the same cocoon.

McCarthy's declaration deserves wide circulation.  Perhaps an alert
entrepreneur will notice there the outline of some simple,
indispensable software, the demand for which could far exceed that for
Nintendo.  When that happens, email will begin to work, outside the
fraternity of grant swingers, for all the people, to all the people,
and among all the people.

Doug McIlroy

∂20-Aug-89  0901	JMC  
Susie about things at old house and about Sierra letter, sarah

∂20-Aug-89  1042	eaf@sumex-aim.stanford.edu 	[The Mailer Daemon <Mailer@score.stanford.edu> : Message of   
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 20 Aug 89  10:42:12 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA10066; Sun, 20 Aug 89 10:44:38 PDT
Date: Sun, 20 Aug 1989 10:44:38 PDT
From: Edward A. Feigenbaum <eaf@sumex-aim.stanford.edu>
To: jmc@sail.stanford.edu
Subject: [The Mailer Daemon <Mailer@score.stanford.edu> : Message of
        15-Aug-89 14:58:48 ]
Message-Id: <CMM.0.88.619638278.eaf@sumex-aim.stanford.edu>

JOHN, HERE IS A REPEAT OF THE PREVIOUS MESSAGE ABOUT THE PROFESSOR(RESEARCH).

ED


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

Return-Path: <Mailer@score.stanford.edu>
Received: from Score.Stanford.EDU by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA20917; Fri, 18 Aug 89 15:09:03 PDT
Message-Id: <8908182209.AA20917@sumex-aim.stanford.edu>
Date: Fri 18 Aug 89 15:05:20-PDT
From: The Mailer Daemon <Mailer@score.stanford.edu>
To: eaf@sumex-aim.stanford.edu
Subject: Message of 15-Aug-89 14:58:48

Message undeliverable and dequeued after 3 days:
Churd@Score.Stanford.EDU.#Internet: Disk quota exceeded
	    ------------
Received: from sumex-aim.stanford.edu by SCORE.STANFORD.EDU with TCP; Tue 15 Aug 89 14:58:49-PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA17683; Tue, 15 Aug 89 15:01:55 PDT
Date: Tue, 15 Aug 1989 15:01:54 PDT
From: Edward A. Feigenbaum <eaf@sumex-aim.stanford.edu>
To: faculty@score.stanford.edu
Subject: Search for Research Professor
Message-Id: <CMM.0.88.619221714.eaf@sumex-aim.stanford.edu>

Dear colleagues,

Bruce Buchanan has resigned his Professor (Research) position in favor of a
professorship at the University of Pittsburgh. Bruce added tremendous
strength to our work in the experimental side of AI research, to
knowledge-based systems research and research on expert systems.  The
Department has been authorized to conduct a search to fill the position
vacated by Bruce.

As chairman of the search committee, I would like to solicit your inputs and
suggestions concerning the people most appropriate for our search committee
to consider.  Our aspirations for the strength of the areas covered and for
the candidates are both "best in the world." Let's not settle for less.

Please bring the enclosed advertisement (to appear in AI Magazine and CACM)
to the attention of such candidates; and please send me your suggestions
for personal followup by me.

With thanks,

Ed Feigenbaum

15-Aug-89 18:59:02-GMT,2658;000000000001
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA13216; Tue, 15 Aug 89 11:59:01 PDT
Resent-Message-Id: <8908151859.AA13216@sumex-aim.stanford.edu>
Return-Path: <irvine@sumex-aim.stanford.edu> 
Received: by sumex-aim.stanford.edu (4.0/inc-1.0) id AA13155; Tue, 15 Aug 89
        11:56:50 PDT 
Date: Tue, 15 Aug 1989 11:56:48 PDT 
From: Sue Irvine <irvine@sumex-aim.stanford.edu>
To: irvine@sumex-aim.stanford.edu
Subject: BGB AD 
Message-Id: <CMM.0.88.619210608.irvine@sumex-aim.stanford.edu> 
Resent-To: eaf@sumex-aim.stanford.edu
Resent-Date: Tue, 15 Aug 1989 11:59:00 PDT
Resent-From: Sue Irvine <irvine@sumex-aim.stanford.edu>

PROFESSOR OR ASSOCIATE PROFESSOR (RESEARCH)
EXPERIMENTAL ARTIFICIAL INTELLIGENCE
STANFORD UNIVERSITY DEPARTMENT OF COMPUTER SCIENCE

Applications are invited for the positions of Professor (Research) or
Associate Professor (Research) of Computer Science.  The search focuses on
computer scientists who work in basic and applied artificial intelligence
research, with a project and publication history in knowledge-based
interdisciplinary systems.  The search will focus on established scientists
in mid-career.  These positions carry full privileges of independent
Principal Investigator.  No classroom teaching is required.

Desireable attributes include:

a.  Experimental, system-building orientation.

b.  Relevant areas:  knowledge representation; model-based and case-based
reasoning methods; issues of large-scale knowledge bases; knowledge
acquisition and machine learning; expert systems.

c.  Orientation toward experimental task areas in engineering and physical
science, e.g. representing real physical devices/systems and reasoning about
their structure and function.  Experience at collaborating with engineers or

scientists on such problems.

d.  Experience with proposing and sustaining substantial research funding and
managing large research projects.

e.  Ability to thrive in collaborative environments, such as the department's
Knowledge Systems or Robotics Laboratories, the Center for Integrated
Systems, or the Stanford Institute for Manufacturing Automation.

f.  Ph.D degree in Computer Science (or equivalent) preferred.

Stanford University is an equal-opportunity employer, and applications from
female and minority applicants are encouraged.

Applicants should include complete curriculum vitae, bibliography of
scientific and technical publications, and a few most important publications.
 These should be sent to:

Edward A. Feigenbaum, Chairman, Search Committee
Department of Computer Science
Margaret Jacks Hall
Stanford University
Stanford, Ca 94305 USA

by November 6, 1989.







-------

∂20-Aug-89  1300	bzs@CS.BU.EDU 	Networks Considered Harmful - For Electronic Mail 
Received: from CS.BU.EDU (BU-CS.BU.EDU) by SAIL.Stanford.EDU with TCP; 20 Aug 89  13:00:18 PDT
Received: by CS.BU.EDU (5.58/4.7)
	id AA29443; Sun, 20 Aug 89 16:03:05 EDT
Date: Sun, 20 Aug 89 16:03:05 EDT
From: bzs@CS.BU.EDU (Barry Shein)
Message-Id: <8908202003.AA29443@CS.BU.EDU>
To: JMC@sail.stanford.edu (John McCarthy)
In-Reply-To: JMC@sail.stanford.edu's message of 17 Aug 89 23:19:00 GM
Subject: Networks Considered Harmful - For Electronic Mail


About five years ago (perhaps a little less), when the domain naming
scheme was being discussed on the TCP-IP mailing list, I made a
"modest proposal" suggesting that we scrap the entire (not yet
implemented) domain naming scheme and instead use telephone numbers to
identify computer systems. I'm sure someone remembers the howling it
caused.

The biggest advantage was that now one could use the telephone
companies as a naming authority. I could even have my computer's phone
number listed in the white pages and to get to me you could call 411
or 555-1212. We'd just need a standard way to list them so you could
ask the operator. "Shein, Barry Email" comes to mind, the phone
company is pretty easy about how you would like your name listed.

The biggest objection was that now we would have these user unfriendly
strings of digits to remember and no one would accept that. It boggles
the mind to figure out exactly how AT&T has gotten away with this
crime against humanity for so long! Or worse, to assess the obvious
failure of AT&T due to this unforgiveable human engineering flaw.

I did suggest that mapping (more "user friendly") strings to these
numbers when interpreting typed in mail addresses seemed within the
ken of the state of the art in computing, I even posted a lisp program
to accomplish this apparently brain-stunning feat (as I remember it
involved ASSOC and not much else.)

At any rate, I agree. There is far too much silliness and
not-invented-here going on in this whole network biz. The Tower of
Babel is the correct parable.

	-Barry Shein

Software Tool & Die, Purveyors to the Trade
1330 Beacon Street, Brookline, MA 02146, (617) 739-0202
Internet: bzs@skuld.std.com
UUCP:     encore!xylogics!skuld!bzs or uunet!skuld!bzs

∂20-Aug-89  1549	thomas@mvac23.uucp 	re: Comments on telecom posting    
Received: from udel.edu (louie.udel.edu) by SAIL.Stanford.EDU with TCP; 20 Aug 89  15:49:46 PDT
Received: from louie.udel.edu by louie.udel.edu id aa19365; 20 Aug 89 18:35 EDT
Date: Sun, 20 Aug 89 15:44:50 edt
From: Thomas Lapp <thomas@mvac23.uucp>
Subject: re: Comments on telecom posting    
Message-ID: <8908201544.0.UUL1.3#5131@mvac23.UUCP>
To: JMC@sail.stanford.edu
In-Reply-To: Your message of 18 Aug 89  2217 PDT
Reply-To: mvac23!thomas@udel.edu

> 2. I had imagined that X400 was complete.  Do you know in what sense it
> is working towards allowing telephone numbers as adresses?  I

Sorry John, I don't know if it is working in that direction.  Sorry if I
gave that impression earlier.  Point is, I am not sure what (if anything)
X.400 people have done for standardizing the the format of the addressee.
I got a headful of this stuff, but it was last year, and memory is failing :-(.
Seems to me like I heard that within small domains the idea was to use the
person's full name as an address (ie. To: Thomas.L.Lapp).  I guess it
wouldn't be much harder to add the phone number as an address domain
(ie. Thomas.L.Lapp@3022929798).

If you want, I'll see if I still have the notes I took from that meeting.
                         - tom

Internet: mvac23!thomas@udel.edu  
uucp    : {ucbvax,mcvax,psuvax1,uunet}!udel!mvac23!thomas
Location: Newark, DE, USA                          
Quote   : NOTICE: System will have a scheduled disk crash at 4:45pm today.

∂20-Aug-89  1831	ficc!peter@uunet.UU.NET 	re: Networks Considered Harmful - For Electronic Mail  
Received: from uunet.uu.net by SAIL.Stanford.EDU with TCP; 20 Aug 89  18:30:57 PDT
Received: from ficc.UUCP by uunet.uu.net (5.61/1.14) with UUCP 
	id AA10677; Sun, 20 Aug 89 21:31:32 -0400
From: ficc!peter@uunet.uu.net
Message-Id: <8908210131.AA10677@uunet.uu.net>
To: uunet!SAIL.Stanford.EDU!JMC@uunet.UU.NET
Subject: re: Networks Considered Harmful - For Electronic Mail
Date: Sun Aug 20 09:06:19 1989

Really? They always seem to be graphical or at least have written notes on them.

∂20-Aug-89  2143	lars@salt.acc.com 	Re: Networks Considered Harmful - For Electronic Mail   
Received: from salt.acc.com by SAIL.Stanford.EDU with TCP; 20 Aug 89  21:43:30 PDT
Received: by salt.acc.com (5.61/1.34)
	id AA09432; Sun, 20 Aug 89 21:44:18 -0700
Date: Sun, 20 Aug 89 21:44:18 -0700
From: lars@salt.acc.com (Lars J Poulsen)
Message-Id: <8908210444.AA09432@salt.acc.com>
To: JMC@sail.stanford.edu
Subject: Re: Networks Considered Harmful - For Electronic Mail
Newsgroups: comp.dcom.telecom
In-Reply-To: <telecom-v09i0306m01@vector.dallas.tx.us>
Organization: Advanced Computer Communications, Santa Barbara, California
Cc: 

I saw your message on comp.dcom.telecom, which appears to be an
inappropriate place for this discussion. When I find it in a more
suitable place (I assume you've crossposted it widely, it is an
invitation to a flamefest) I will post a response.

/ Lars
-- 
/ Lars Poulsen <lars@salt.acc.com>   (800) 222-7308  or (805) 963-9431 ext 358
  ACC Customer Service              Affiliation stated for identification only
                My employer probably would not agree if he knew what I said !!

∂21-Aug-89  0514	eli@chipcom.com 	comments on McCarthy's fax/email essay
Received: from ursa-major.SPDCC.COM by SAIL.Stanford.EDU with TCP; 21 Aug 89  05:13:53 PDT
Received: by ursa-major.SPDCC.COM with sendmail-5.59/4.7 
	id <AA25544@ursa-major.SPDCC.COM>; Mon, 21 Aug 89 08:11:58 EDT
From: eli@chipcom.com
Received: from localhost by com (4.0/SMI-4.0)
	id AA03395; Mon, 21 Aug 89 07:38:24 EDT
Message-Id: <8908211138.AA03395@com>
To: spdcc!eecs.nwu.edu!telecom
Cc: spdcc!sail.stanford.edu!JMC
Subject: comments on McCarthy's fax/email essay
Date: Mon, 21 Aug 89 07:38:22 -0400


a few points.

1 -- there capability to integrate uucp mail and fax is already with us
     a number of demonstrations have been made, such that users can send
     mail to address!phone_num, and have the node at "address" convert their
     email message to group 3 fax and fax it to any fax machine.

2 -- fax does not necessarily lose the ability to edit documents.  PC-fax
     cards allow one to modify received faxes as well as generate them from
     scratch (or printer output).

3 -- the only thing stopping me from setting up a permanent email->fax
     gateway is problems running UUPC with a hayes clone modem.  also, 
     with DID inbound fax, the reverse gateway could be provided for $1
     per month per user.  each user would have his own DID fax number..


-- Steve Elias
-- eli@spdcc.com, eli@chipcom.com 
-- voice mail: 617 239 9406
-- work phone: 617 890 6844          

∂21-Aug-89  0616	greg%cs.heriot-watt.ac.uk@NSFnet-Relay.AC.UK 	Functional language workshop 
Received: from NSFnet-Relay.AC.UK by SAIL.Stanford.EDU with TCP; 21 Aug 89  06:15:51 PDT
Received: from cs.heriot-watt.ac.uk by NSFnet-Relay.AC.UK   via Janet with NIFTP
           id aa01436; 21 Aug 89 12:29 BST
From: Greg Michaelson <greg%cs.heriot-watt.ac.uk@NSFnet-Relay.AC.UK>
Date: Mon, 21 Aug 89 09:54:15 BST
Message-Id: <8744.8908210854@brahma.cs.hw.ac.uk>
To: jmc@sail.stanford.edu
Subject: Functional language workshop

Dear John

I hope this finds you well.

I am organising a workshop on the development of functional programming
languages under the auspices of the British Computer Society Formal
Aspects of Computing Group. If you're going to be in Europe next spring
then would you like to come and talk about the development of LISP for
about an hour?

I'm afraid that we can't afford to pay your fare & expenses from the USA
but we could pay your fare from say the South of England, and overnight
accomodation costs.

Best wishes!

Greg Michaelson

∂21-Aug-89  0857	arch_ems@gsbacd.uchicago.edu 	RE: Networks Considered Harmful... 
Received: from gsbacd.uchicago.edu by SAIL.Stanford.EDU with TCP; 21 Aug 89  08:57:37 PDT
Date: 21 Aug 89 10:55:00 CDT
From: arch_ems@gsbacd.uchicago.edu
Subject: RE: Networks Considered Harmful...
To: "jmc" <jmc@sail.stanford.edu>

Dear John McCarthy;

I read with interest your comments regarding EMAIL vs. FAX.  our office, as
a pseudo-commercial entity within a major university relies on both ARPANET
electronic mail and faxes.  I agree with a basic tenet underlying your
article -- that the facsimile machine is an unnecessary piece of technology
and merely points to the confusion which currently exists in the multiplying
computer hardware communication packages and protocals.  Of course, there is
another reason for the proliferation of facsimile machines.  You don't have
to be a rocket scientist to run one.  This may seem trivial to a well-connected
(sorry) member of a large university but the average business person still
doesn't know the difference between a SCSI port and an ethernet gateway. With
a fax you feed paper in one end and it comes out the other end.  The other
"barrier to entry" as business folks like to say, is that the philosophy
behind the PC or workstation is that it sit on an individual's desk and is
used by that individual.  A piece of paper can come into a centralized point,
be distributed easily, and be responded to by phone, handwriting, secretary
or typing pool, phone, etc.  By contrast, email means sitting in front 
of a computer screen.  The Wall Street Journal tells today (Monday Aug 20) of
computer use in large corporations---

COMPANY              # of COMPUTERS        EMPLOYEES:COMPUTERS ratio
General Electric     42,000                8.54:1
General Electric     45,000                8.54:1
General Electric     45,000                6.71:1
Ford Motor Co.       42,000                8.54:1
Du Pont Co.          40,000                2.75:1
GTE Corp.            32,000                3.12:1
General Motors       31,500                24.28:1
General Dynamics     30,000                3.16:1
Westinghouse Elec    30,000                3.33:1
Boeing Co.           28,000                5.46
Eastman Kodak, Co    25,000                3.00
Merrill Lynch        25,000                1.92

Those are just the biggies.  If you think about your average middle to small
size business I suspect the average # of computers to people hovers around
5:1 (people:computers).  Faxes will still be more efficient for awhile (espec.
since many computers, even in business, are still not on LANs!)  Now some
interesting things are taking place.  It is making an increasing amount of
sense for business to set up email links to businesses which they communicate
with alot.  We use CE Software's QUICKMAIL to communicate with another office,
our lawyers, and a half-dozen small companies which we started (we are a
technology management/venture capital company).  In addition we use ARPANET
to communicate with University of Chicago researchers and Argonne researchers
(we manager intellectual property coming out of those two institutions).
Increasingly electronic mail vendors are coming out with "bridges" to other
people's products.  For QuickMail there are bridges to:
  MCI MAIL
  COMPUSERVE
  VAXMAIL
  UNIX MM
  and to remote quickMail sites
  and to single users calling in.
Built into the product is a facility for writing your own bridges so, in theory
anyway, the product can hook up with anything.

In short, I think three things:  Your right.  It won't happen yet.  Market
forces will solve this problem.

Nothing would stop you, however, from coming up with a good solution to this
problem to sell to the market.  I have two ideas:

1)  A subscription service which has leased t-1 lines connecting major
metropolitan areas and will take input from fax, electronic mail, or 
digitizer (high end imaging applications), transmit to the site closest
to receiver and output as fax, electronic mail, high-end imaging, or regular
hard copy through the mail.  This would provide an information link to
every competing system from every competing system, cut companies long
distance phone bills (you would save by batching transmissions, compressing\
and getting cut rates for volume), and provide a transition from fax to
electonic mail.

2)  Invent a very cheap reception data device with cheap screen technology
which would plug into a regular phone connection.  You could have this
screen fax on your desk, ready to take a 'fax' call instead of a regular
voice call.  It could display ASCII characters from an electronic mail
message of graphics from a fax and send either.  It would have three ports:
phone line, keyboard (optional) and network (to connect to a computer or
a scanner).  Make it about binder size so that you could carry it to a
central printer or scanner and dump/load memory to print/store fax msgs.
Of course you would probably have to sell them for about $250 to compete
against the $1500 fax machine.  But if you invent one I'd invest in the company
to market the things...

Forgive the 'business' flavor of this msg.  I just think that, at this point
the solution to this communications dillema is going to have to come from
the market, not the committee.

yours,

--Ted

Edward Shelton, Project Manager
ARCH Development Corporation
arch_ems@gsbacd.uchicago.edu


∂21-Aug-89  1431	dgc@math.ucla.edu 	Re: Re -- Networks Considered Harmful - For Electronic Mail  
Received: from math.ucla.edu (sunset.math.ucla.edu) by SAIL.Stanford.EDU with TCP; 21 Aug 89  14:31:05 PDT
Return-Path: <dgc@math.ucla.edu>
Received: from joshua.math.ucla.edu by math.ucla.edu (Sendmail 5.59/1.02)
	id AA01887; Mon, 21 Aug 89 14:31:39 PDT -0800
Received: from loopback by joshua.math.ucla.edu (Sendmail 3.2/1.02)
	id AA02713; Mon, 21 Aug 89 14:31:16 PDT
Message-Id: <8908212131.AA02713@joshua.math.ucla.edu>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: Re -- Networks Considered Harmful - For Electronic Mail 
In-Reply-To: Your message of 18 Aug 89 15:46:00 -0700.
Date: Mon, 21 Aug 89 14:31:15 PDT
From: dgc@math.ucla.edu

I hadn't realized that your article was intended for CACM.  If I had,
I wouldn't have commented on the ACM vs the IEEE.

I know of no direct telephone access to "telnet" any more.  There used
to be "tips" which were, in effect, modems directly connected to the
arpanet, but I think that they are all gone.

One other comment in support of your proposal:  There's no reason that
the user interface couldn't be made to look essentially the same as
the present one.  My local system could accept the mail addressed, for
example, as this one is, to "JMC@SAIL.Stanford.EDU" and
(automatically) translate it to John_McCarthy@408-xxx-xxxx.  
This means that your system could be introduced gradually and without
trauma to the user.  Once some relatively simple software was widely
distributed, then Stanford CS, could, for example, say that it will
no longer accept mail via internet or bitnet--instead use this
telphone number, . . .  However, I don't think this will happen
until it's recharged its fair share of the internet costs!

dgc

David G. Cantor
Department of Mathematics
University of California at Los Angeles
Internet:  dgc@math.ucla.edu
UUCP:      ...!{randvax, sdcrdcf, ucbvax}!ucla-cs!dgc

∂22-Aug-89  0725	JIMM%ACMVM.BITNET@Forsythe.Stanford.EDU 	Re: Signed Viewpoint, final texless version 
Received: from Forsythe.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Aug 89  07:25:40 PDT
Received: by Forsythe.Stanford.EDU; Tue, 22 Aug 89 07:24:11 PDT
Received: by ACMVM (Mailer R2.03B) id 6612; Tue, 22 Aug 89 10:23:31 EST
Date:         Tue, 22 Aug 89 10:22:54 EST
From:         "J. Maurer (Ex. Editor)" <JIMM%ACMVM.BITNET@Forsythe.Stanford.EDU>
Subject:      Re: Signed Viewpoint, final texless version
To:           John McCarthy <JMC@SAIL.Stanford.EDU>
In-Reply-To:  Your message of 17 Aug 89 1803 PDT

     Thanks.  I'll let you know the schedule as soon as we nail it down.
Jim Maurer

∂22-Aug-89  1131	alex@jessica.Stanford.EDU 	thesis, requests for change...   
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Aug 89  11:31:32 PDT
Received: by jessica.Stanford.EDU; Tue, 22 Aug 89 11:31:15 PDT
Date: Tue, 22 Aug 1989 11:31:14 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: jmc@sail.Stanford.EDU
Subject: thesis, requests for change...
Message-Id: <CMM.0.88.619813874.alex@Jessica.Stanford.EDU>

Prof. McCarthy,

	I'm not sure how this phase of the PhD usually proceeds (this is
only my first) but if you have comments, or requests for changes, additions,
etc... before you'll sign my dissertation, please let me know.

	Thanks,
				Alex

∂22-Aug-89  1528	arc!chet@apple.com 	Networks Considered Harmful - For Electronic Mail 
Received: from apple.com by SAIL.Stanford.EDU with TCP; 22 Aug 89  15:28:34 PDT
Received: by apple.com (5.59/25-eef)
	id AA02774; Tue, 22 Aug 89 15:29:19 PDT
	for JMC@sail.stanford.edu
Received: by arc.UUCP (4.0/SMI-4.0)
	id AA05360; Tue, 22 Aug 89 15:16:42 PDT
Date: Tue, 22 Aug 89 15:16:42 PDT
From: arc!chet@apple.com (Chet Wood)
Message-Id: <8908222216.AA05360@arc.UUCP>
To: JMC@sail.stanford.edu
In-Reply-To: John McCarthy's message of 17 Aug 89 23:19:00 GMT
Subject: Networks Considered Harmful - For Electronic Mail

Hi, John.

	I like your idea! I do have a couple of comments, however...



JM>     2. [uucp]  requires that the message forwarding computer have login
JM> privileges on the receiver.

	The uucp protocol in itself does not require this, therefore
it could be adapted to serve in your proposed system. Login privileges
are required by security-conscious _systems_. It would be difficult to
modify these systems to accept 

                     MAIL <use>@$<telephone number>

	connections while satisfying security concerns. But it would
be easy to modify unix uucp programs to _originate_ this type of
message.

Tapping into the existing uucp network would give the system a good
head start.


JM>     Perhaps the organizationally simplest solution would be to get
JM> one or more of the various UNIX consortia to add a direct mail
JM> telephone protocol to UUCP.  Such a protocol would allow mail to
JM> be addressed to a user-id at a telephone number. It mustn't require
JM> pre- arrangement between the sending and receiving computers, and
JM> therefore cannot involve any kind of login. Non-UNIX systems would
JM> then imitate the protocol.

Oh. You already thought of that. 

JM>  The computer would require a dialer and a modem with whatever
JM> characteristics were taken as standard and it would be well to use
JM> the same standards as have been adopted for telefax.

Since computers are mostly hooked up to modems already ( Even
fax-modem cards have standard modem included) I think it would be
better to use this standard. Computers are better suited to ascii
email since the software can be more easily retrofitted. It would be
possible to design a fax machine that speaks both protocols (at least
for receiving), but it would have to be a new design anyway. But the
fax software would have to be modified to speak the new protocol
anyway (convert chars to bits), so why not modify its modem to be
compatible with a standard modem?


JM>     Fortunately, there is free enterprise. Therefore, the most
JM> likely way of getting direct electronic mail is for some company
JM> to offer a piece of hardware as an electronic mail terminal
JM> including the facilities for connecting to the current variety of
JM> local area networks (LANs). The most likely way for this to be
JM> accomplished is for the makers of fax machines to offer ASCII
JM> service as well.  This will obviate the growing practice of some
JM> users of fax of printing out their messages in an OCR font,
JM> transmitting them by fax, whereupon the receiver scans them with
JM> an OCR scanner to get them back into computer form.

I can think of several products which could work with the system.

 * Faxmodem cards which are smart enough to tell which kind of modem is
calling them, and communicate this information to higher-level
software.

 * "Daemon" software to run on Macs and PCs to work with the faxmodem
cards.

 * If the current generation of home computers won't support daemons
(not multitasking), smarter faxmodem cards with buffering could be
made available. 

 * Standalone FAX machines w/ email facility. Probably a display and
maybe a keyboard would have to be built in, because more
sophisticated software is needed for reading (and sending) mail.
For example, the problem of junk mail will be worse w/ ascii email
since it will be cheaper to send it. In order to save trees, it
would be desirable for a user to be able to preview the incoming
messages and choose which ones to print.

 * LAN-connected mail bridge you describe above. This would isolate
incoming mail calls from security-conscious systems, so they wouldn't
have to ask for a login and password.

By the way, junk mail is another reason why Unix requires login
protocols. There should be a provision in the new proposal for logins and
passwords if the receiving station wants. Actually, the current uucp
system is pretty close to this. After all, a unix system administrator
_could_ set up a process to answer the phone with "uucico" instead of
"getty." 



JM>     More generally, suppose the same need can be met either by
JM> buying a product or subscribing to a service.  If the costs are at
JM> all close, the people who sell the product win out over those
JM> selling the service.  Why this is so I leave to psychologists, and
JM> experts in marketing, but I suppose it has to do with the fact
JM> that selling services requires continual selling to keep the
JM> customers, and this keeps the prices high.

Have you had any response from people that might like to make
"products?" I'd like to get in on the ground floor of this. Sounds
like an idea whose time is here.

Regards,

Chet.

Chet Wood                       ~                         (408)727-3357
     arc!chet@apple.COM         .  Advansoft Research Corporation
          chet@arc.UUCP         .       4301 Great America Parkway
               apple!arc!chet   .            Santa Clara, CA 95054, USA

∂22-Aug-89  1540	pjd@riacs.edu 	Re:  Signed Viewpoint, final texless version 
Received: from icarus.riacs.edu by SAIL.Stanford.EDU with TCP; 22 Aug 89  15:39:05 PDT
Received: from hydra.riacs.edu by icarus.riacs.edu (5.59/2.1G)
	   id AA01802; Tue, 22 Aug 89 15:40:21 PDT
Received: by hydra.riacs.edu (4.12/2.0N)
	   id AA14720; Tue, 22 Aug 89 15:40:24 pdt
Message-Id: <8908222240.AA14720@hydra.riacs.edu>
Date: Tue, 22 Aug 89 15:40:24 pdt
From: Peter J. Denning <pjd@riacs.edu>
To: JMC@SAIL.Stanford.EDU
Subject: Re:  Signed Viewpoint, final texless version
Cc: jimm@acmvm.bitnet

John, got your revision.  Thanks for adding clarifications to the
text.  I'll forward to Jim Maurer for processing.

Peter

∂22-Aug-89  1741	qphysics-owner@neat.cs.toronto.edu 	test.    
Received: from neat.cs.toronto.edu by SAIL.Stanford.EDU with TCP; 22 Aug 89  17:41:37 PDT
Received: by neat.cs.toronto.edu id 10688; Tue, 22 Aug 89 20:37:11 EDT
Received: by neat.cs.toronto.edu id 10686; Tue, 22 Aug 89 20:36:56 EDT
From:	Jean-Francois Lamy <lamy@ai.utoronto.ca>
To:	qphysics@cs.toronto.edu
Subject: test.
Message-Id: <89Aug22.203656edt.10686@neat.cs.toronto.edu>
Date:	Tue, 22 Aug 89 20:36:54 EDT
Resent-From: qphysics-owner@cs.toronto.edu
Resent-To: qphysics-distribution@cs.toronto.edu
Resent-Reply-To: Jean-Francois Lamy <lamy@ai.utoronto.ca>
Resent-Message-Id: <89Aug22.203711edt.10688@neat.cs.toronto.edu>
Resent-Date: Tue, 22 Aug 89 20:37:05 EDT

A test of the updated mailing list, to weed out obsolete/broken addresses.
I will let you know when the qualitative physics mailing list has stabilized.

∂23-Aug-89  1400	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
Received: from IU.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 23 Aug 89  13:59:58 PDT
Received: from peabody.teleos.com by IU.AI.SRI.COM via SMTP with TCP;
	  Wed, 23 Aug 89 13:58:31-PST
Received: by peabody.teleos.com (3.2/4.16) id AA00542 for
	  nilsson@score.stanford.edu; Wed, 23 Aug 89 13:57:24 PDT
Date: Wed, 23 Aug 89 13:57:24 PDT
From: Leslie Kaelbling <leslie@teleos.com>
Message-Id: <8908232057.AA00542@peabody.teleos.com>
To: nilsson@score.stanford.edu, jmc@sail.stanford.edu, stan@teleos.com
Subject: Convening the committee
Reply-To: leslie%teleos.com@ai.sri.com


According to current rules and regulations, I'm supposed to convene my thesis
committee once a year and (I guess) explain what I've been up to.  Since it's
been a year since I filed my G81, it seems to be time.

Maybe we can aim for sometime in September.  I have few constraints, but I know
each of you has many.   Let me know what days (and times?) in September would
be appropriate and I'll try to find an intersection.

- Leslie

∂24-Aug-89  2105	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:STICKEL@warbucks.ai.sri.com 	CADE-10 Call for Papers  
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Aug 89  21:05:16 PDT
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU.#Internet) by SCORE.STANFORD.EDU with TCP; Thu 24 Aug 89 20:59:25-PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa09615;
          24 Aug 89 23:53 EDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 24 Aug 89 23:51:23 EDT
Received: from lcs.mit.edu (CHAOS 15044) by MC.LCS.MIT.EDU 24 Aug 89 23:50:11 EDT
Received: from Warbucks.AI.SRI.COM by mintaka.lcs.mit.edu id aa09433;
          24 Aug 89 23:28 EDT
Date: Thu 24 Aug 89 18:47:31-PDT
From:     Mark Stickel <STICKEL@warbucks.ai.sri.com>
Subject: CADE-10 Call for Papers
To:       theorem-provers@mc.lcs.mit.edu
cc:       STICKEL@warbucks.ai.sri.com
Message-ID: <620012851.0.STICKEL@AI.SRI.COM>
Mail-System-Version: <VAX-MM(229)+TOPSLIB(126)@AI.SRI.COM>
Reply-To: STICKEL@warbucks.ai.sri.com

                               CADE-10

        10th International Conference on Automated Deduction

                    Kaiserslautern, West Germany
                          July 23-27, 1990

                          Call for Papers

CADE is the major forum at which research on all aspects of automated
deduction can be presented.  Papers on automated deduction (for classical
and nonclassical logics) in the following and related fields are invited:

 Theorem Proving          Decision Procedures      Logic Programming
 Unification       Program Verification/Synthesis  Inference Systems
 Term Rewriting           Deductive Databases      Applications


                          Program Committee

 Peter Andrews            Claude Kirchner          William Pase
 Wolfgang Bibel           Jean-Louis Lassez        Lawrence Paulson
 W.W. Bledsoe             Donald Loveland          Fernando Pereira
 Alan Bundy               Ewing Lusk               David Plaisted
 Robert Constable         Michael McRobbie         Joerg Siekmann
 Jean-Pierre Jouannaud    Dale Miller              Mark Stickel, Chairman
 Deepak Kapur             Hans Juergen Ohlbach     Richard Waldinger
 Matt Kaufmann            Ross Overbeek            Christoph Walther


Original research papers (up to 5,000 words; 15 proceedings pages, 6 X 9
inches, 12 point type, will be allotted) are solicited.  Also solicited
are system summaries that describe working reasoning systems (2
proceedings pages) and problem sets that provide realistic, interesting
challenges for automated reasoning systems (5 proceedings pages).  The
title page of the submission should include author's name, address,
phone number, and E-mail address.  Papers must be unpublished and not
submitted for publication elsewhere.  Late papers and papers that
require major revision, including submissions that are too long, will be
rejected.

    Submission receipt deadline:      November 27, 1989
     Author notification date:        February 15, 1990
 Camera-ready copy receipt deadline:  April 2, 1990
 
Six paper copies should be sent to arrive by November 27, 1989 to

                          Mark E. Stickel
                   Artificial Intelligence Center
                         SRI International
                       333 Ravenswood Avenue
                    Menlo Park, CA 94025  U.S.A.

Inquiries about CADE can also be sent by electronic mail to
Stickel@AI.SRI.COM

CADE-10 is supported by International Joint Conferences
on Artificial Intelligence, Inc.

*************************** PLEASE POST ***************************
-------


∂24-Aug-89  2154	cheriton@Pescadero.Stanford.EDU 	Gorbis 
Received: from Pescadero.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Aug 89  21:54:31 PDT
Received:  by Pescadero.Stanford.EDU (5.59/25-eef) id AA10547; Thu, 24 Aug 89 21:55:17 PDT
Date: Thu, 24 Aug 89 21:55:17 PDT
From: David Cheriton <cheriton@Pescadero.Stanford.EDU>
Message-Id: <8908250455.AA10547@Pescadero.Stanford.EDU>
To: jmc@su-ai
Subject:  Gorbis

John,
  Aloex Gorbis came to me, asking about an RAship and expressing interest
in switching to the computer communication area.  Would you be willing to
give me some indication of what you think his strengths and weaknesses
are as a potential researcher, etc.  I dont generally take on grad students
as RAs unless they have strong potential for doing a thesis with me.
Any input would be appreciated.
David Cheriton

∂25-Aug-89  0819	harnad@clarity.Princeton.EDU 	On What Is and Isn't Symbolic 
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 25 Aug 89  08:18:53 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA12108; Fri, 25 Aug 89 10:56:57 EDT
Received: by psycho.Princeton.EDU (4.0/1.84)
	id AA10442; Fri, 25 Aug 89 10:47:23 EDT
Date: Fri, 25 Aug 89 10:47:23 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8908251447.AA10442@psycho.Princeton.EDU>
To: harnad@clarity.Princeton.EDU
Subject: On What Is and Isn't Symbolic

              ON WHAT IS AND ISN'T SYMBOLIC

To: Oded Maler, Department of Applied Mathematics and Computer Science,
Weizmann Institute of Science, Rehovot 76100, Israel (oded@wisdom.bitnet)

CC: Recipients of Symbol Grounding preprint

Oded, you wrote:

> You [i.e., SH] say that a connectionist system (or any other
> dynamical system) is not symbolic because
> 
>   "The patterns of interconnections do not decompose, combine and
>   recombine according to a formal syntax that can be given a systematic
>   semantic interpretation..."
> 
> I'm not so sure about that. There might be some complex rules for the
> decomposition of dynamical systems. They might not be as simple as
> rules for combining symbolic tokens, but this is a quantitative issue.
> Since you agree that being symbolic is a systemic property, maybe the
> (structural) combination of two networks into a larger one whose
> pattern of interaction (behavior) is some kind of composition of the
> original patterns, can be viewed by a symbolic-minded observer as a
> (syntactic) composition of symbolic entities with their corresponding
> (semantic) composition.

There seem to be two possibilities, one more straightforward than the other:

(1) The first, straightforward possibility is that a dynamical system
might implement a symbol system, say, a programmable computer. This
possibility is not interesting, because the functions of interest would
then be occurring at the symbolic level, which, as always, is
independent of its implementation (and subject to the Symbol Grounding
Problem). The dynamical system would be providing only one of the many
possible hardwares for doing symbol manipulation. (A slightly more
interesting variant of this first possibility is one in which the
endstate is the same -- a net providing the hardware for doing pure
symbol manipulation -- but this state is arrived at by "training"
rather than hard-wired design. The fact that a net can be trained to
support symbol manipulation is interesting, but if in the end all the
action is at the symbolic level, then the history does not much
matter.)

(2) The second possibility is a rather vague one, and a little too
close to some of the extremely general senses of "turing equivalence,"
according to which everything -- dynamic or static -- is doing
computation, in that it is equivalent to a certain turing machine in a
certain state. This over-general notion is tantamount to saying that
everything is a symbol system (hence, trivially, so is a net, and so
are we).

In the Symbol Grounding paper I listed eight much more constrained
criteria for what many theorists seem to take to be a symbol system
(although even those were probably too vague); the general sense of
turing equivalence would not be sufficient to satisfy those criteria.
In fact, the notion of turing equivalence probably capitalizes illictly
on the symbol grounding problem itself in allowing our interpretations
to serve as the mapping between the phenomenon we are interpreting and
the hypothetical turing machine to which it is equivalent. That is a
classical example of the parasitism (on our "symbolic-mindedness," as
you put it) that lies at the heart of the Symbol Grounding Problem,
except that here it is not only extended to the meanings of the symbols
in a symbol system that meets the eight criteria for being a symbol
system; here the criteria themselves -- especially systematicity -- are
met only in the mind of the beholder: Even a static system with no
capacity for ruleful composition -- e.g., an inert chair -- is given a
full-blown semantic interpretation (thanks to the full power of the
[grounded] semantics in the mind of the interpreter).

But perhaps this is not what you had in mind. Perhaps you meant that
it might turn out that a net -- working as a net rather than just as
an architecture for running a Lisp program -- might turn out to be
literally and explicitly fulfilling the eight criteria for being a
symbol system that I listed. I'm not sure about this. First it would
have to be shown to be the case. But it's not clear whether your
conjecture is sufficiently critical. You seem to be willing to allow
it to be something more "complex" than the explicit manipulation of
symbol tokens. But explicit symbol manipulation is the essence of the
eight criteria I listed; abandon that and you're no longer talking
about symbol systems in the quasi-formal sense analyzed in the paper
(and implicitly used by the proponents of the theory that cognition is
just symbol manipulation).

To put it another way: According to the sense of "symbol system"
considered in the paper, in a symbol system the semantics may be a
matter of interpretation but the SYNTAX may not. The latter must be
literally there. It's not clear that in the "decomposition" you have in
mind, the syntax (the symbols and the rules for manipulating them) is
literally present and operative in the system, or the the system can
merely be INTERPRETED as having and operating according to that
syntax.

One test might be programmability: One of the criterial features of
symbol systems is that their symbols must be explicitly manipulable.
Hence if you have a net which, according to some decomposition, is
really a symbol system doing symbol manipulation, then it should be
possible to intervene directly at this symbolic level to show that the
decomposition has independent reality.

Let me mention one caveat, though: What I referred to in the paper as a
"dedicated symbol system" -- one whose operations are in any way
constrained by hard causal connections between its symbol tokens and
the nonsymbolic objects they stand for -- does NOT meet the criteria
for being a pure symbol system, which must be subject to syntactic
constraints only, i.e., formal rules for manipulating meaningless
formal or physical symbol tokens purely on the basis of their
(arbitrary) shapes. (In fact, with its interpretations "fixed" by the
hard-wiring, a dedicated system would not have a symbol grounding
problem either.) This may be analogous to the hypothetical
decomposition of a dynamical system you have in mind here, and its
failure to meet the criteria for being a symbol system -- especially
explicit representation and manipulabilty -- may be for similar reasons.

> Maybe systems become more and more symbolic as their behavior can be
> understood with less and less energy considerations.

There's no continuum. Symbolicity is not a matter of degree. Symbolic
function must always be independent of energy considerations (except
that, unless it's just on paper, the formal system must be implemented
dynamically SOMEHOW). A symbol system is hardware-independent -- it can
be implemented energetically in many different ways, but all the ways
must be equivalent in their systematic properties, including explicit
representation and explicit manipulabilty. If in one "implementation" a
"rule" is hard-wired instead of being explicitly represented
symbolically, the implementation may no longer be a symbol system at
all.

> You also claim that symbol grounding is a problem for cognitive
> modeling and not necessarily for AI (making machines), although you
> acknowledge the potential contribution of its solution to the design
> of robots. Finally you conclude that there is only one route from
> senses to symbols, from the ground up. I need a clarification of this
> point: does it apply only to cognitive modeling (and if so what is the
> semantics of the existence of a unique route), and as a design
> methodology for intelligent artifacts, should they be built only from
> the bottom-up? Are all the efforts and research on (symbolic)
> knowledge representation and reasoning, common-sense theories etc.
> premature?  My feeling is that this (bottom-up) is the correct
> strategy but I fail to find a convincing arguments to show why the
> lower-level senso-motoric modules will not plug into the independently
> developed (top-down) symbolic components.

I will send you by regular mail the other papers you request in which
this problem is discussed more explicitly, but the gist of the answer
is this: At the heart of it is a convergence argument and what I've
called the "Total Turing Test." There are many ways to generate a toy
performance capacity, and, because of the Symbol Grounding Problem, the
purely symbolic solution to a toy problem is open to many (ungrounded)
interpretations. As the toy problem grows ("scales up," or what have
you), the number of alternative ways to solve it shrinks. When we get
to lifesize problems -- in particular, that of producing a candidate
that passes the (linguistic version of the) Turing Test, i.e., one that
has all of our cognitive capacity that is demonstrable symbolically --
the number of alternatives has presumably shrunk considerably, yet the
candidate system is still susceptible to the the multiple semantic
interpretations symptomatic of the Symbol Grounding Problem.
Successfully scale up to the Total Turing Test, however, which calls
for not only symbolic I/O but full robotic interactions with the world,
and you will simultaneously have solved four problems:

(1) You will have rendered the empirical demands you make of the robot
candidate EXACTLY the same as those you make of your fellow man
(thereby confronting the "other-minds" problem as directly as it can be
confronted; there is of course no guarantee that the model has captured
the true laws of cognition, or really has a mind; but at least the TTT
is as close as you can hope to come).

(2) You will have made the level of "underdetermination" of your model
EXACTLY the same as what it is in the rest of natural science and
engineering (a Total Turing model of the universe is all that physics
can hope to provide; it cannot guarantee that another, radically
different, empirically equivalent, equiparametric Total model cannot
exist -- or that either of them is necessarily the "true" model of the
laws of nature): One cannot prove that there is a "unique" route; only
a convergent methodology.

(3) You will have (presumably, if the model does pass the TTT) solved
the Symbol Grounding Problem by fixing the interpretations of whatever
internal symbols are involved through their connections to the objects
they pick out -- as demonstrated robotically.

(4) You will have protected your candidate from Searle's Chinese Room
Argument, because transducer function (and any other nonsymbolic
function involved in passing the TTT) is immune to his objection, which
works only against pure symbolic AI models of the mind.

I have no proof that the relation (between (i) the nonsymbolic
"transducer/effector" component and (ii) whatever symbol manipulation will
be needed in this TTT model) will have to be non-modular, but I
strongly suspect it will. First, it seems that to successfully
ground the entire system the transducer component will have to be doing
so much of the work itself that the portion left over for the pure
symbolic module (if such it is) would be radically smaller and more
circumscribed than what seems to be envisioned by top-down theorists.
Second, if it turns out that I am right about HOW the symbolic
component needs to be grounded -- i.e., by being a dedicated, hybrid
symbolic/nonsymbolic system, with its symbol manipulations constrained
not only by syntactic rules operating on the arbitrary shapes of
symbol tokens, as in a pure symbol system, but by the causal
connections between the elementary symbol tokens and the NONarbitrary
shapes of the icons and invariant icons to which they are connected --
then there will be no independent symbolic module at all!

How relevant this all is to robotics outside the context of cognitive
modeling I cannot say. It all depends on how ambitious roboticists are,
and how closely their goals are related to TTT capacity. But yes, I do
suspect that "all the efforts and research on (symbolic) knowledge
representation and reasoning, common-sense theories etc. [are]
premature." Not only premature, perhaps altogether wrong-headed.

Stevan Harnad

∂26-Aug-89  1633	Mailer 	failed mail returned  
The following message was undeliverable to the address(es) below
because the destination Host Name(s) are Unknown:
newell@cmu-10a

------- Begin undelivered message: -------
 26-Aug-89  1632	JMC 	copies of slides    
To:   newell@cmu-10a   
If you have them in computer mailable form, I would
be grateful for your list of human capabilities and
the one labelled "flexibility for autonomy".
Congratulations on an illuminating talk.

------- End undelivered message -------

∂26-Aug-89  1832	@ri.cmu.edu:Gripe@FAC.CS.CMU.EDU 	Mail address warning 
Received: from RI.CMU.EDU by SAIL.Stanford.EDU with TCP; 26 Aug 89  18:32:52 PDT
Date: Sat, 26 Aug 89 21:32:27 EDT
From: MAILER-DAEMON@RI.CMU.EDU
Reply-To: Gripe@FAC.CS.CMU.EDU
To: JMC@SAIL.STANFORD.EDU
Subject: Mail address warning

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

Mail addressed to:

    Newell@ri.cmu.edu

was ambiguous, but is being delivered to:

    Allen Newell    <an@natasha.mach.cs.cmu.edu>

Other possible matches for "Newell" were:

    Paul Newell

In the future, specify this address as:n
     Allen.Newell@ri.cmu.edu

in order to avoid this warning message.

------- Message contents appear below ------

Received: from SAIL.STANFORD.EDU by RI.CMU.EDU; 26 Aug 89 21:32:15 EDT
Message-ID: <fSvAz@SAIL.Stanford.EDU>
Date: 26 Aug 89  1831 PDT
From: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: copies of two slides
To:   Newell@RI.CMU.EDU

If you have them in computer mailable form, I would
be grateful for your list of human capabilities and
the one labelled "flexibility for autonomy".
Congratulations on an illuminating talk.

∂26-Aug-89  2011	Allen.Newell@NATASHA.MACH.CS.CMU.EDU 	Re: copies of two slides   
Received: from NATASHA.MACH.CS.CMU.EDU by SAIL.Stanford.EDU with TCP; 26 Aug 89  20:11:34 PDT
Received: from natasha.mach.cs.cmu.edu by NATASHA.MACH.CS.CMU.EDU id aa22837;
          26 Aug 89 23:12:03 EDT
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: copies of two slides 
In-reply-to: Your message of 26 Aug 89 18:31:00 -0700.
             <fSvAz@SAIL.Stanford.EDU> 
Date: Sat, 26 Aug 89 23:12:01 EDT
Message-ID: <22835.620190721@NATASHA.MACH.CS.CMU.EDU>
From: Allen.Newell@NATASHA.MACH.CS.CMU.EDU


John: Thanks for the good word.  Here are the text form of the transparencies.
I also include an earlier text form of the GPS - Soar comparison.  I added the
last item (reflection) on this new graphic version, because Soar's taking of
impasses is an reflection, though it is not done deliberately (as in some of
the other reflective systems) but done when it runs into difficulty.

AN

---------------------------------------------------------------------------
             FUNCTIONAL REQUIREMENTS 
       FOR A HUMAN-SCALE INTELLIGENT SYSTEM

1. Behave flexibly as a function of the environment

2. Exhibit adaptive (rational, goal-oriented) behavior

3. Operate in real time

4. Operate in a rich, complex, detailed environment

   4.1. Perceive immense amounts of changing detail

   4.2. Use vast amounts of knowledge

   4.3. Control a motor system of many degrees of freedom

5. Use symbols and abstractions

6. Use language, both natural and artificial

7. Learn from the environment and from experience

8. Acquire capabilities through development

9. Operate autonomously, but within a social community

10. Be self aware and have a sense of self
---------------------------------------------------------------------------
           ON FLEXIBILITY FOR AUTONOMY

What can change after the programmer departs

Immense progress in obtaining flexibility

   Compare GPS (1959) and Soar (1989)


             None Fixed Prog Select Ajar Open

Formulation
  Problem spaces        GPS              Soar

Focus
  Goal variety                GPS        Soar
                               3
  Difficulty                        GPS  Soar

Search control 
  Methods                     GPS        Soar
                               2
  Alternatives    GPS                    Soar

  Knowledge             GPS   Soar       Soar
                              all

Communication     GPS                    Soar


Acquisition  GPS  Soar                   Soar?


Reflection   GPS                         Soar

---------------------------------------------------------------------------
     COMPARISON OF SOAR (1989) AND GPS (1959)

1. Problem spaces
   GPS had a single problem space
   Soar has multiple problem spaces, can generate new ones

2. Goals
   GPS had a fixed set (3) of deliberate subgoal types 
   Soar generates its own subgoals of unlimited diversity

3. Difficulty driven
   GPS had a fixed set of state differences + operator nonapply
   Soar detects any lack of knowledge (by impasses)

4. Methods
   GPS had means-ends analysis and operator subgoaling
   Soar has all the weak methods, creates strong methods 

5. What decision to make
   GPS had fixed context-dependent occasions for decisions
   Soar permits any decision to be made on any occasion

6. Decision making
   GPS had fixed schemes for selecting methods and subgoals
   Soar has open-ended retrieval of knowledge for deliberation
   Soar can convert any decision into general problem solving

7. Communication of results and context
   GPS had prepackaged data communication
   Soar operates in a single open context

8. Learning
   GPS had no learning
   Soar learns continuously from experience on all aspects

∂27-Aug-89  1909	underdog@Portia.stanford.edu 	re: Columbia (vis a vis the drug cartel) (actually lunch)   
Received: from Portia.stanford.edu by SAIL.Stanford.EDU with TCP; 27 Aug 89  19:09:51 PDT
Received:  by Portia.stanford.edu (5.61/25-eef) id AA27417; Sun, 27 Aug 89 19:07:41 -0700
Date: Sun, 27 Aug 89 19:07:41 -0700
From: Dwight Joe <underdog@Portia.stanford.edu>
Message-Id: <8908280207.AA27417@Portia.stanford.edu>
To: JMC@SAIL.Stanford.EDU
Subject: re: Columbia (vis a vis the drug cartel) (actually lunch)

Sure.

How about Tuesday, August 29?

---Dwight

∂27-Aug-89  2218	underdog@Portia.stanford.edu 	re: Columbia (vis a vis the drug cartel) (actually lunch)   
Received: from Portia.stanford.edu by SAIL.Stanford.EDU with TCP; 27 Aug 89  22:18:46 PDT
Received:  by Portia.stanford.edu (5.61/25-eef) id AA29760; Sun, 27 Aug 89 22:16:37 -0700
Date: Sun, 27 Aug 89 22:16:37 -0700
From: Dwight Joe <underdog@Portia.stanford.edu>
Message-Id: <8908280516.AA29760@Portia.stanford.edu>
To: JMC@SAIL.Stanford.EDU
Subject: re: Columbia (vis a vis the drug cartel) (actually lunch)

I've got black hair and no beard.
hmmm.....

Physically, I look Japanese.  8↑)

---Dwight

∂28-Aug-89  0800	JMC  
Furno

∂28-Aug-89  0907	Mailer 	re: Quote for August 28, 1989   
Received: from rml2.sri.com by SAIL.Stanford.EDU with TCP; 28 Aug 89  09:07:25 PDT
Date: Mon 28 Aug 89 09:06:47-PST
From: Ric Steinberger <RIC@RML2.SRI.COM>
Subject: re: Quote for August 28, 1989   
To: JMC@SAIL.STANFORD.EDU
Cc: RIC@RML2.SRI.COM, comments@KL.SRI.COM, su-etc@SAIL.STANFORD.EDU
Message-ID: <620323607.930000.RIC@RML2.SRI.COM>
In-Reply-To: <6T7S4@SAIL.Stanford.EDU>
Mail-System-Version: <VAX-MM(246)+TOPSLIB(136)@RML2.SRI.COM>

I'm not sure what The New Yorker has done to deserve all the condemnation
that JMC has unleashed.  Does the fact that it often takes liberal viewpoints
(but has condemned some Soviet international actions as strongly as some
conservatives) make it unacceptable?  

    The cartoons in The new Yorker are some of the best around, and the
writing is excellent.  Elizabeth Drew, The New Yorker's Washington policical
corespondent, is one of the most perceptive observers of Washington's
power struggles.  The stories about the Tienanmen Square massacre by
Fred Sharparo include some of the best coverage of those events.  The 
short stories by authors such as John Updike and John Irving are superb,
even if the magazine doesn't tell you who either of them are.  (There's
a bit od snobbery here - you're supposed to know who all the authors are.)
This is a magazine NOT FOR NEW YORKERS only!  Go out and buy it!  

-ric steinberger

JMC: If you have a point of contention with the Elizabeth Drew quotation,
why not respond specifically to that?


-------

∂28-Aug-89  0937	MPS 	Knowledge Systems   

Good morning

Dr White's office (U. of Md.) called asking if you had the publishable
notes ready for them, along with commonsense logic and elephant 2000.
They want to put together the handbook for certification.  I have
the secretary's name and number.  Dr. White would like to have an
answer as soon as possible.

Pat

∂28-Aug-89  0959	ullman@nimbin.Stanford.EDU 	lunch  
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Aug 89  09:59:20 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA00118; Mon, 28 Aug 89 09:57:04 PDT
Date: Mon, 28 Aug 89 09:57:04 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908281657.AA00118@nimbin.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: lunch

I made a reservation at the faculty club for 12:30.
See you there?
				---jdu

∂28-Aug-89  1345	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
Received: from IU.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 28 Aug 89  13:45:27 PDT
Received: from peabody.teleos.com by IU.AI.SRI.COM via SMTP with TCP;
	  Mon, 28 Aug 89 13:44:30-PST
Received: by peabody.teleos.com (3.2/4.16) id AA04649 for
	  nilsson@tenaya.stanford.edu; Mon, 28 Aug 89 13:42:32 PDT
Date: Mon, 28 Aug 89 13:42:32 PDT
From: Leslie Kaelbling <leslie@teleos.com>
Message-Id: <8908282042.AA04649@peabody.teleos.com>
To: nilsson@tenaya.stanford.edu, jmc@sail.stanford.edu, stan@teleos.com
Subject: Convening the committee
Reply-To: leslie%teleos.com@ai.sri.com

It looks like September's out:  the intersection of constraints is null.  So,
what about major constraints in October?

- Leslie

∂28-Aug-89  1447	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	Re:  party and dinner   
Received: from zaphod.es.llnl.gov ([128.115.4.30]) by SAIL.Stanford.EDU with TCP; 28 Aug 89  14:47:13 PDT
Received: by zaphod.es.llnl.gov (4.0/1.26)
	id AA03699; Mon, 28 Aug 89 14:47:52 PDT
Date: Mon, 28 Aug 89 14:47:52 PDT
From: clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov (Cynthia Mason  422-8911)
Message-Id: <8908282147.AA03699@zaphod.es.llnl.gov>
To: JMC@sail.stanford.edu
Subject: Re:  party and dinner

Bonjour!

So nice to hear from you again!  I was wondering how your
party went!  I felt a real loss at not being able to go,
and yet, that night, we discovered "MOTOWN"!!  You know
it was right down the street the whole time!

Yes, I would CERTAINLY like to have dinner.  Especially
in San Francisco...one of my favorite places!  I wondered
if I would hear from you...that was one of the most memorable
evenings of my life...you have WONDERFUL stories to tell, I
could listen forever!!

Well my schedule is fairly flexible, as school has not strated
yet (oops - should be "started")  When do classes start for you??

Cheers.

Cindy

∂28-Aug-89  1504	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
Received: from IU.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 28 Aug 89  15:04:05 PDT
Received: from peabody.teleos.com by IU.AI.SRI.COM via SMTP with TCP;
	  Mon, 28 Aug 89 14:59:13-PST
Received: by peabody.teleos.com (3.2/4.16) id AA04771 for
	  JMC@SAIL.Stanford.EDU; Mon, 28 Aug 89 14:57:12 PDT
Date: Mon, 28 Aug 89 14:57:12 PDT
From: Leslie Kaelbling <leslie@teleos.com>
Message-Id: <8908282157.AA04771@peabody.teleos.com>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 28 Aug 89  1424 PDT
	     <1FTrZp@SAIL.Stanford.EDU>
Subject: Convening the committee   
Reply-To: leslie%teleos.com@ai.sri.com

How about Weds. October 11 at 11?  It shouldn't take more than an hour.

Do you know what is supposed to go on at this meeting?

I met someone at the machine learning workshop (because he reviewed my paper)
named Rich Sutton.  Of anyone I've talked to, he seems to have the most
interest and experience in the area I'm working in.  Would it be reasonable
to invite him to be on my reading committee?  I think someone has to make
special provisions for such an external reader, but I think it would be good
to have someone like him around to check my proofs.

- Leslie

∂28-Aug-89  1528	@IU.AI.SRI.COM,@peabody.teleos.com:leslie@teleos.com 	Convening the committee   
Received: from IU.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 28 Aug 89  15:28:06 PDT
Received: from peabody.teleos.com by IU.AI.SRI.COM via SMTP with TCP;
	  Mon, 28 Aug 89 15:27:23-PST
Received: by peabody.teleos.com (3.2/4.16) id AA04819 for
	  JMC@SAIL.Stanford.EDU; Mon, 28 Aug 89 15:25:20 PDT
Date: Mon, 28 Aug 89 15:25:20 PDT
From: Leslie Kaelbling <leslie@teleos.com>
Message-Id: <8908282225.AA04819@peabody.teleos.com>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 28 Aug 89  1515 PDT
	     <bTsJJ@SAIL.Stanford.EDU>
Subject: Convening the committee   
Reply-To: leslie%teleos.com@ai.sri.com

How about Tuesday the 10th at 9?  

Rich Sutton did his PhD at UMass under Barto.  He now works at GTE, I think in 
the same group as Oliver Selfridge.  His thesis work was a set of
"connectionist" (he doesn't call it that, but it has that flavor) algorithms 
that learn to solve the 1-D pole-balancing problem, in which the system only
knew it had made a mistake when the pole had fallen all the way over or the
cart had run off the track.  He has a good background in statistics and such,
so his algorithms come with convergence proofs instead of much of the usual
connectionist handwaving.  (There are still some constants to tune, however...)

My approaches are not all that similar to his, but they address similar
problems and share a reliance on statistics.  I've been reading lots of 
statistics lately, but don't necessarily trust myself to have gotten it
right.

I could send you papers of his or get a copy of his CV if you are interested.

- L

ullman@nimbin
stable models vs. well founded semantics
Vladimir is unrepentant.  He said that stable models have found
more uses than well founded semantics and offers the following
citations.  He says that proving the desirable properties of
stable models is also much easier than proving the properties
of well founded semantics.  I dunno about generality.
Probably you and he should talk.

∂28-Aug-89  1705	VAL 	Three uses of stable models   

K. Eshghi and R. A. Kowalski (Imperial College), "Abduction Compared
with Negation by Failure".

	"Horn clause logic programming can be extended to include
	abduction with integrity constraints... The declarative
	semantics of the resulting abductive framework is equivalent
	to the stable model semantics of the original logic program."

D. Sacca and C. Zaniolo (MCC), "Stable models and Non-determinism in
Logic Programs with Negation".

	"Previous researchers have proposed generalizations of Horn
	clause logic to support negation and non-determinism as two
	separate extensions. In this paper we show that the stable
	model semantics for logic programs provides a unified basis
	for the treatment of both concepts. Thus, we show that stable
	models subsume the use of the non-deterministic choice
	construct of LDL, which in turn subsumes the use of the cut
	in Prolog. Since stable models also subsume previously
	proposed deterministic two-valued semantics... stable models
	provide a theoretical basis for a unified semantics of logic
	programs. The main perceived drawback of the stable model
	semantics is that no procedure is currently known to construct
	these models. Thus we introduce a new procedure, called
	backtracking fixpoint, that non-deterministically constructs
	a stable model if one exists, or, otherwise, proves the non-
	existence of such models."

	"Well-founded models represent a very important notion...
	However, since there only exists one well-founded (possibly
	partial) model for each program, they do not allow us to
	capture the semantics of non-determinism."

S. G. Pimentel and J. L. Cuarado (Institute for Defense Analyses),
"A Truth Maintenance System Based on Stable Models".

	"In this paper, we present a new system for truth maintenance
	which combines the strengths and corrects the deficiencies of
	of several previous systems... Unlike previous systems, ours
	is grounded in a formal semantics, using the stable models of
	[Gelfond and Lifschitz 88]. As a result, the behavior of our
	TMS can be immediately related to the use of negation as
	failure in logic programs and to other nonmonotonic formalisms
	proposed in the literature."

∂28-Aug-89  1810	ullman@nimbin.Stanford.EDU 	Re:  stable models vs. well founded semantics  
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Aug 89  18:10:45 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA01113; Mon, 28 Aug 89 18:08:30 PDT
Date: Mon, 28 Aug 89 18:08:30 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908290108.AA01113@nimbin.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: Re:  stable models vs. well founded semantics
Cc: val@sail.stanford.edu

The point that others are looking at stable models cannot be refuted.
However, my understanding is that well-founded semantics is
strictly more general, and that the Stable-Model paper points this out
(I don't have it, so I can't say for sure).

Perhaps expressivity is not the only issue, however.
If one is looking for a class of logic programs that is tractable
in some sense, then one might want a less expressive class.
For example, Ken Ross is looking at applying "magic sets" and
fixed on the weak locally stratified semantics, which, as I understand it,
is strictly less expressive than stable-model semantics, because that
condition was what he needed to make his rule rewriting valid.

My position at lunch was, and still is, that when a new model is proposed,
the reason why it is a win over all other proposals needs to be a
prominent part of the paper, and that the field is best served by
program committees and referees who demand this sort of comparative
analysis.
				---jdu

∂28-Aug-89  2000	BEDIT@Score.Stanford.EDU 	Summary of July computer charges. 
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Aug 89  20:00:12 PDT
Date: Mon 28 Aug 89 19:38:11-PDT
From: Billing Editor <BEDIT@Score.Stanford.EDU>
Subject: Summary of July computer charges.
To: MCCARTHY@Score.Stanford.EDU
Message-ID: <12521861245.9.BEDIT@Score.Stanford.EDU>

Dear Mr. McCarthy,

Following is a summary of your computer charges for July.

Account     System   Billed    Pct      Cpu    Job   Disk  Print   Adj   Total

JMC         SAIL     2-DMA816  100   591.86  29.64 ***.**  54.22  5.00 2937.05
MCCARTHY    SCORE    2-DMA816  100      .00    .00  30.47    .00  5.00   35.47
jmc         LABREA   2-DMA816  100      .00    .00 105.90    .00  5.00  110.90

Total:                               591.86  29.64 ***.**  54.22 15.00 3083.42


University budget accounts billed above include the following. 

Account     Princip Inv      Title                      Comment             

2-DMA816    McCarthy         N00039-84-C-0211           Task 23              


The preceding statement is a condensed version of the detailed summary sheet 
sent monthly to your department. 

Please verify each month that the proper university budget accounts are paying 
for your computer usage.  Please also check the list of account numbers below 
the numeric totals.  If the organizations/people associated with that account 
number should NOT be paying for your computer time, send mail to BEDIT@SCORE. 

Please direct questions/comments to BEDIT@SCORE. 
-------

∂28-Aug-89  2131	ullman@nimbin.Stanford.EDU 	re:  stable models vs. well founded semantics  
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Aug 89  21:31:40 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA02641; Mon, 28 Aug 89 21:29:25 PDT
Date: Mon, 28 Aug 89 21:29:25 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908290429.AA02641@nimbin.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: re:  stable models vs. well founded semantics

OK, sure there are a variety of dimensions in which to
measure things, and there can be a number of different
"winners," e.g., a most expressive semantics, one that
is most expressive with a known way to test membership in
the chosen model efficiently, etc.

But it's more than just referencing the literature.
I see an important contribution of the "theoretical" way
of looking at things being the "competitive"approach.
It's not just saying "the problem was also looked at by a and b and ..."
If you can also say, "and my way is better than each of these
for the following reason..." then your paper deserves publication.
If not, not.

By belief is that encouraging such justification helps
strengthen the field, and probably leads to better research
in the long run.
				---jdu

∂29-Aug-89  0547	boyer@CLI.COM 	[postmaster@CLI.COM: Returned mail: User unknown] 
Received: from CLI.COM by SAIL.Stanford.EDU with TCP; 29 Aug 89  05:46:52 PDT
Received: by CLI.COM (4.0/1); Tue, 29 Aug 89 07:43:01 CDT
Date: Tue, 29 Aug 89 07:43:01 CDT
From: Robert S. Boyer <boyer@CLI.COM>
Message-Id: <8908291243.AA11034@CLI.COM>
To: jmc@sail.stanford.edu
Subject: [postmaster@CLI.COM: Returned mail: User unknown]
Reply-To: boyer@cli.com

The following is from an editorial by Rich Gabriel and Guy Steele in
the journal ``Lisp and Symbolic Computation'', Vol. 2, Number 2, p.
103.

   Physicists once won a war for the US.  Since then, physicists have
   set up the system to favor themselves.  The largest single computer
   science project we know about was the S1 project at Lawrence
   Livermore National Laboratory, which was to build a parallel
   supercomputer along with all the needed software.  The head of the
   project was a prominent physicist, and congressional testimony in
   favor of the project was provided by Edward Teller -- the father of
   the hydrogen bomb.

   Physicists and chemists have official lobbying bodies.  The ACM
   explicitly refuses to support lobbying.  Everyone believes
   physicists and chemist are scientists, but few believe computer
   scientists are.

   Wouldn't we regard a $4.4 billion dollar computer science project
   by the US government a gift from heaven?  Don't physicists regard
   the $4.4 billion superconducting supercollider as their just
   desserts?



∂29-Aug-89  0824	ARK 	smithsonian    
To:   JMC, CLT, ME
 ∂29-Aug-89  0628	BH 	sail  
Can't we donate it to the Smithsonian, on condition that they
keep running it forever?

∂29-Aug-89  0911	MPS 	Thesis    
It will be mailed (hardcopy) today from Princeton.

Pat

∂29-Aug-89  0954	MPS 	meeting   
MAD meeting - Thursday, 31st at 9:30

Don't forget to bring the signed consents.

Pat

∂29-Aug-89  1020	ullman@nimbin.Stanford.EDU 	re:  stable models vs. well founded semantics  
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Aug 89  10:20:07 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA02860; Tue, 29 Aug 89 10:17:47 PDT
Date: Tue, 29 Aug 89 10:17:47 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908291717.AA02860@nimbin.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: re:  stable models vs. well founded semantics
Cc: ullman@nimbin.Stanford.EDU, val@sail.stanford.edu

OK; you're forcing me to formalize my ideal policy regarding
the writing and publication of papers.  How about this:
In order to be published, a paper needs to address the issue of
in what way(s) it is superior to all previously known approaches to
its problem.  Program committees and referees should not hesitate
to reject papers with no such justification, or only minor improvements
in directions that do not look promising.

Benefits:
1. It focuses researchers on what they should be doing, i.e.,
improving things rather than lengthening their publication list
or finding some excuse to go to a conference.

2. It significantly cuts down on the number of published papers,
saving trees and letting the good ideas become more widely known.

3. It helps trace the origins of ideas and the contributions made by each
paper.  That's useful for "peripheral" issues like tenure decisions.

Downside:
It can surpress some good ideas by people who don't play by the "rules."
The only significant instance I can think of was when I was chairman
of the first POPL conference, and we rejected a paper that was an
independent discovery of the idea of "continuation semantics"
because no one on the committee could understand the paper,
and the author did not clearly explain what the achievement was.

As to how it applies to your three examples:

1. I don't think that the evaluation of programming languages
should or does take place by people writing papers proving the
virtues of one language or another.

2. The introduction of formalisms is an interesting test case.
Surely, one justification for a paper is that NO previous work has
addressed the question.  However, the paper that proposes a different
way to look at a problem, without any evidence of superiority, clearly
does violate my proposed policy.  I can only suggest that reviewers
need to exercise a little judgement.  If there are not even informal
suggestions as to why the proposal might be a win, perhaps the work
should NOT be published.

The [GL] paper on stable models is an interesting case.  I gather
that the citations VAL provided are each cases where the logic
programs with stable models were EXACTLY what was needed, no more nor less,
just as in Ross' case, the programs with weak locally stratified semantics
are exactly what his method of evaluation can handle.
My policy would have required [GL] to mention at least one of
these applications where well-founded semantics was "too much."
That might have been an injustice, since [GL] may not have been
thinking along any of these lines, yet by publishing their
paper they were able to stimulate others who were so thinking.
OK.  This is a flaw in my policy.  I can only suggest that the
benefits outweigh the disadvantages.  We cannot be clairvoyant
about when a proposal will turn out to be "just right" in some
situation, so no policy will admit all those papers that turn out
to have some unforseen benefit, yet not admit all papers.

3. As for algorithms, it is always legitimate to say "my algorithm
is best for n <= 10↑1000, but worse thereafter.  Asymptotic analysis
is not the only criterion for evaluation.
I can see you again arguing that if we didn't admit the paper
with an algorithm that was best for NO n, how would we have obtained
the improvement that was better for "small" n?
Again, my policy would only accept the paper with the improvement,
and perhaps a chain of development from one algorithm to the improved
would have been lost.  But again, I see no way to avoid accepting
EVERY possible paper on the grounds that someone might later improve
it to the point where it was "best" in some way.

Summary:  I still think that toughness of standards is an important
characteristic of a field; it is something that the theory community
has, and I think that AI would benefit from adopting a similar stance.
I'd like very much to see more joint activity of the PODS community
in particular and the people interested in nonmonotonic or
commonsense reasoning, knowledge representation, etc.
				---jdu

∂29-Aug-89  1209	MPS  
Sarah will be around this evening and said if you
wanted to stop by the house and go through things
that would be fine with her.

Pat

∂29-Aug-89  1251	hoffman.pa@Xerox.COM 	debate  
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 29 Aug 89  12:51:02 PDT
Received: from Semillon.ms by ArpaGateway.ms ; 29 AUG 89 12:51:17 PDT
Date: 29 Aug 89 12:46 PDT
From: Reid Hoffman <hoffman.pa@Xerox.COM>
Subject: debate
To: jmc@sail.stanford.edu
cc: hoffman.pa@Xerox.COM
Message-ID: <890829-125117-3559@Xerox>

Professor McCarthy,

I have just sent mail to Searle asking him about what time would be good
and what format he would like.  I recall that you suggested that each of
you present a short list of subjects for the debate, but I  don't recall
what sort of formal format you would like (if any).  So what timing and
what format would you like?

thanks
reid


----- rh

∂29-Aug-89  1302	bhr@sumex-aim.stanford.edu 	CS300 Speakers   
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 29 Aug 89  13:02:38 PDT
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA03872; Tue, 29 Aug 89 13:04:58 PDT
Date: Tue, 29 Aug 1989 13:04:57 PDT
From: Barbara Hayes-Roth <bhr@sumex-aim.stanford.edu>
To: CAB@sail.stanford.edu, binford@coyote.stanford.edu,
        cheriton@pescadero.stanford.edu, Gail@sol-margret.stanford.edu,
        dill@amadeus.stanford.edu, feigenbaum@SUMEX-AIM.Stanford.EDU,
        rwf@sail.stanford.edu, RPG@sail.stanford.edu,
        Genesereth@score.stanford.edu, Ginsberg@polya.stanford.edu,
        goldberg@polya.stanford.edu, golub@patience.stanford.edu,
        ag@pepper.stanford.edu, halpern@ibm.com,
        hayes-roth@SUMEX-AIM.Stanford.EDU, jlh@amadeus.stanford.edu,
        Herriot@score.stanford.edu, Iwasaki@SUMEX-AIM.Stanford.EDU,
        ARK@sail.stanford.edu, OK@coyote.stanford.edu, DEK@sail.stanford.edu,
        lam@mojave.stanford.edu, latombe@coyote.stanford.edu,
        val@sail.stanford.edu, linton@amadeus.stanford.edu,
        zm@sail.stanford.edu, mayr@polya.stanford.edu, jmc@sail.stanford.edu,
        ejm@sierra.stanford.edu, miller@kl.sri.com, jcm@polya.stanford.edu,
        bmoore@ai.sri.com, nii@SUMEX-AIM.Stanford.EDU,
        nilsson@tenaya.stanford.edu, oliger@pride.stanford.edu,
        pratt@jeeves.stanford.edu, ALS@sail.stanford.edu,
        shoham@score.stanford.edu, singh@score.stanford.edu,
        shortliffe@SUMEX-AIM.Stanford.EDU, tobagi@sierra.stanford.edu,
        ullman@score.stanford.edu, waldinger@ai.sri.com,
        Wiederhold@SUMEX-AIM.Stanford.EDU, wilf@score.stanford.edu,
        winograd@csli.stanford.edu
Cc: bhr@sumex-aim.stanford.edu
Subject: CS300 Speakers
Message-Id: <CMM.0.88.620424297.bhr@sumex-aim.stanford.edu>


I will be organizing CS300, the Department Lecture Series for new Ph.D.
students, again this Fall quarter. We will meet on Thursdays, at 4:15-5:30,
in Room 420 050 in the Psychology Building. 

As most of you know, the purpose of CS300 is to give students an overview
of current research in the department and to acquaint them with particular
projects and faculty with whom they might wish to work. If you would like
to make a presentation, please let me know as soon as possible and suggest
2-3 Thursday dates during the quarter that would be convenient for you.

Thanks very much,
Barbara Hayes-Roth

∂29-Aug-89  1424	lars@salt.acc.com 	re: Networks Considered Harmful - For Electronic Mail   
Received: from salt.acc.com by SAIL.Stanford.EDU with TCP; 29 Aug 89  14:24:17 PDT
Received: by salt.acc.com (5.61/1.34)
	id AA00988; Tue, 29 Aug 89 14:24:58 -0700
Date: Tue, 29 Aug 89 14:24:58 -0700
From: lars@salt.acc.com (Lars J Poulsen)
Message-Id: <8908292124.AA00988@salt.acc.com>
To: JMC@SAIL.Stanford.EDU
Subject: re: Networks Considered Harmful - For Electronic Mail

	Date: 26 Aug 89  1807 PDT
	From: John McCarthy <JMC@SAIL.Stanford.EDU>
	Subject: re: Networks Considered Harmful - For Electronic Mail  
	To: lars@SALT.ACC.COM
	
	[In reply to message sent Sun, 20 Aug 89 21:44:18 -0700.]
	
	I haven't posted it widely.  Besides at Stanford, comp.dcom.telecom
	is the only place.  If you find a more suitable place, let
	me know.  It will appear in CACM late this year as a
	"Guest Viewpoint".  comp.dcom.telecom posting resulted
	in considerable response, mostly agreeing and the rest
	pointing out what was considered minor inaccuracies.

I would have expected this type of discussion in comp.misc - but I may
have been alone in that expectation. Certainly, there has been lively
discussion of it in .telecom.

In general, I think there are two things to say:
(1) We already have too many standards, rather than not enough.
    Among these are: UUCP, X.25/TCP-IP/SMTP, TeleTex, etc.
(2) FAX is now ubiquitous; whereas "the right thing" would be TeleTex,
    the best option at this point would probably be to add a text mode
    to FAX, so that consenting faxes could avoid formatting/redigitizing
    messages, and save both transmission time and storage costs.

Meanwhile, UUCP works just fine among consenting adults.

But the time is right for discussing how to crank the dial up a bit
more.

    add a text mode to

∂29-Aug-89  1436	korf@CS.UCLA.EDU 	Re: Advertisement     
Received: from lanai.cs.ucla.edu by SAIL.Stanford.EDU with TCP; 29 Aug 89  14:36:14 PDT
Return-Path: <korf@CS.UCLA.EDU>
Received: from Denali.CS.UCLA.EDU by lanai.cs.ucla.edu (Sendmail 5.61.ucla-34/2.19)
	id AA12170; Tue, 29 Aug 89 14:37:00 -0700
Message-Id: <8908292137.AA12170@lanai.cs.ucla.edu>
Received: by denali; Tue, 29 Aug 89 14:46:03 pdt
Date: Tue, 29 Aug 89 14:46:03 pdt
From: Richard E Korf <korf@CS.UCLA.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 26 Aug 89  1628 PDT <jStNN@SAIL.Stanford.EDU>
Subject: Re: Advertisement  

Dear John,
  Thanks for sending along the posting. Unfortunately, I'm tenured here, so I
wouldn't be inclined to resign my position for an untenured slot.  However, I'm
due for a sabbatical in the near future and may be interested in a visiting
research position at some point.
  In any case, I enjoyed our chat in Detroit, and would be interested in coming
up to continue it, and possibly give a talk on some of my recent work if there
is sufficient interest. We don't start school until the 28th of September, so
next month would be good for me if you're back in session. Except for the week
of September 11, when I'll be at the Distributed AI Workshop near Seattle. I
could also come up later, but then I'll be constrained by classes on Tuesday and
Thursdays.
                          -rich

                          


 

∂29-Aug-89  2357	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	The Saganic Verses 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 29 Aug 89  23:57:46 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 415208; 30 Aug 89 02:54:58 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107332; Tue 29-Aug-89 23:49:03 PDT
Date: Tue, 29 Aug 89 23:48 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: The Saganic Verses
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890830064805.9.RWG@TSUNAMI.SPA.Symbolics.COM>

Reply-to:  rwg@yukon.scrc.symbolics.com

The Chudnovskys have had a gigadigit of π for a couple
of weeks, but IBM newsmanagers have only now okayed the
announcement.  (Apparently, it had become a matter of
national prestige, but by the time the only agency with
enough computrons resolved some awkward matters of
national origin, the IBM cruncher muddled to victory.)

David and Gregory want to stress that Jenks's ScratchPad
project was essential in deriving and partially autocoding
the hairy congruences by which they datacompressed
intermediate matrix products.

And if this wan't enough, hardcore Life hacker Dean Hickerson
has just sent me a c/3 orthogonal spaceship, which was
discovered by a program!

∂30-Aug-89  0134	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu,@albanycs.albany.edu:kapur@sutra.albany.edu 	post-doctoral fellowship
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Aug 89  01:34:01 PDT
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU.#Internet) by SCORE.STANFORD.EDU with TCP; Wed 30 Aug 89 01:32:36-PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa13402;
          30 Aug 89 4:16 EDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 30 Aug 89 04:16:16 EDT
Received: from lcs.mit.edu (CHAOS 15044) by MC.LCS.MIT.EDU 30 Aug 89 04:14:19 EDT
Received: from ALBANYCS.ALBANY.EDU by mintaka.lcs.mit.edu id aa09812;
          29 Aug 89 22:55 EDT
Received: by albanycs.albany.edu (5.54/4.8)
	id AA06519; Tue, 29 Aug 89 22:49:51 EDT
Received: from sutra.albany.edu by cssun.albany.edu (3.2/SMI-3.2)
	id AA09443; Tue, 29 Aug 89 22:48:13 EDT
Received: by sutra.albany.edu (3.2/SMI-3.2)
	id AA22423; Tue, 29 Aug 89 22:48:11 EDT
Date: Tue, 29 Aug 89 22:48:11 EDT
From: kapur@sutra.albany.edu
Message-Id: <8908300248.AA22423@sutra.albany.edu>
To: rewriting@crin.crin.fr, narrow@a.cs.uiuc.edu, 
    theorem-provers@mc.lcs.mit.edu, clp@cs.cmu.edu
Subject: post-doctoral fellowship


	PLEASE CIRCULATE OR POST. Thanks. - Deepak Kapur


-------


	POST-DOCTORAL FELLOWSHIP AVAILABLE

Applications are invited for a post doctoral fellowship
established under the Graduate Research Initiative at University
at Albany, State University of New York starting from October 1,
1989, or on a mutually agreed date.  The fellowship will be in an
institute devoted to programming, logics and systems being
planned jointly with RPI, Troy.  Appointment may be possible for
up to three years. The applicant will have the primary
responsibility to design and build a specification and
verification system for distributed and parallel computation.
There will also be ample opportunity to work on others projects
related to theorem proving and its applications to specification
and verification of software and hardware.

Candidates should possess a PhD or have equivalent research or
industrial experience. Candidates with a strong research record
in the use of formal methods in software and hardware,
distributed computation, and theorem proving will be given
preference.

Applicants should send a CV and the names of two references to:

 Deepak Kapur
 Department of Computer Science
 State University of New York at Albany
 Albany, NY 12222

The University at Albany, is an Equal Opportunity Affirmative
Action Employer.  Applications from women, minority persons,
handicapped persons, and or special disabled or Vietnam era
veterans are especially welcome.

Further details may be obtained from Deepak Kapur (at the above
address or e-mail to kapur@albanycs.albany.edu).











∂30-Aug-89  0452	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	daily trivium 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 30 Aug 89  04:52:01 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 415223; 30 Aug 89 07:41:28 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107347; Wed 30-Aug-89 04:35:34 PDT
Date: Wed, 30 Aug 89 04:34 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: daily trivium
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890830113439.5.RWG@TSUNAMI.SPA.Symbolics.COM>

1 - 2 sin(pi/14) = sec(pi/7)/2

∂30-Aug-89  0555	cross@vax.darpa.mil 	Visit on 19 September 1989   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 30 Aug 89  05:55:24 PDT
Received: from sun46.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA01658>; Wed, 30 Aug 89 08:52:39 -0400
Posted-Date: Wed 30 Aug 89 08:57:00-EDT
Received: by sun46.darpa.mil (4.1/5.51)
	id AA01243; Wed, 30 Aug 89 08:57:02 EDT
Date: Wed 30 Aug 89 08:57:00-EDT
From: Steve Cross <CROSS@DARPA.MIL>
Subject: Visit on 19 September 1989
To: jmc@sail.stanford.edu
Cc: cross@vax.darpa.mil, randw@vax.darpa.mil
Message-Id: <620485020.0.CROSS@SUN46.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@SUN46.DARPA.MIL>

John, Vadimir, and Yoav:  Rand Waltzman and myself would like to visit 
Stanford on the 19th.  We will be spending the morning with the people we 
support under Ed Feigenbaum.  Can we set up a meeting to review your 
DARPA funded projects in the afternoon, say beginning at 2:00 PM?  Thanks. 
Steve Cross
-------

∂30-Aug-89  1004	VAL 	re:  stable models vs. well founded semantics
To:   ullman@NIMBIN.STANFORD.EDU, JMC@SAIL.Stanford.EDU   
[In reply to message from ullman@nimbin.Stanford.EDU sent Mon, 28 Aug 89 18:08:30 PDT.]

> However, my understanding is that well-founded semantics is
> strictly more general, and that the Stable-Model paper points this out
> (I don't have it, so I can't say for sure).

It's the other way around. "We show here that when a program has a
well-founded model, that is also the unique stable model. However,
we discuss programs that have no well-founded model and do have a
unique stable model." (Van Gelder/Ross/Schlipf).

--Vladimir

∂30-Aug-89  1103	ullman@nimbin.Stanford.EDU 	re:  stable models vs. well founded semantics  
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Aug 89  11:03:07 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA03883; Wed, 30 Aug 89 11:00:51 PDT
Date: Wed, 30 Aug 89 11:00:51 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908301800.AA03883@nimbin.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU, VAL@SAIL.Stanford.EDU, ullman@NIMBIN.STANFORD.EDU
Subject: re:  stable models vs. well founded semantics

Then I must apologize for my assumption that the [GL] paper lacked
justification.  I don't think that was clear from the extended abstract
that you showed me.  If it was, how could the paper have been rejected
from PODS?

Do you really mean that when a program has a WF model with no
"unknown" ground atoms, then it has a stable model?
I thought that the "win" program
	w(X) :- move(X,Y) & not w(Y)
(in the case where move has cycles, was an example of a program with
a WF model but no stable model, or a stable model that didn't give the
"right" answer.

Can you give an example of a case where SM gives the "right" answer,
but WF gives the "wrong" answer (it is really impossible for WF
to give no answer at all, although in cases like p :- not q; q :- not p
it assigns "unknown to both p and q.)
				---jdu

∂30-Aug-89  1107	ullman@nimbin.Stanford.EDU 	reply to VAL
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Aug 89  11:07:13 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA03889; Wed, 30 Aug 89 11:04:58 PDT
Date: Wed, 30 Aug 89 11:04:58 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908301804.AA03889@nimbin.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: reply to VAL

I thought I was copying you on my reply to Vladimir's claim that
SM >= WF.  Apparently not, so would you try to see his reply.
				---jdu

∂30-Aug-89  1108	ullman@nimbin.Stanford.EDU 	forget it   
Received: from nimbin.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Aug 89  11:07:56 PDT
Received: by nimbin.Stanford.EDU (5.51/inc-1.01)
	id AA03896; Wed, 30 Aug 89 11:05:43 PDT
Date: Wed, 30 Aug 89 11:05:43 PDT
From: Jeffrey D. Ullman <ullman@nimbin.Stanford.EDU>
Message-Id: <8908301805.AA03896@nimbin.Stanford.EDU>
To: jmc@sail.stanford.edu
Subject: forget it

Apparently the cc did take effect.
If not, I have a copy.
				---jdu

∂30-Aug-89  1737	betsy@russell.Stanford.EDU 	Blinov's TINLunch
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Aug 89  17:36:57 PDT
Received: by russell.Stanford.EDU (4.0/inc-1.0)
	id AA11242; Wed, 30 Aug 89 17:43:13 PDT
Date: Wed 30 Aug 89 17:43:13-PDT
From: Betsy Macken <BETSY@CSLI.Stanford.EDU>
Subject: Blinov's TINLunch
To: jmc@sail.stanford.edu
Message-Id: <620527393.0.BETSY@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

So far I have scheduled Blinov for October 12 as you suggested, but
you wanted me to check back with you.  I need to confirm the
first few weeks in October soon -- can you confirm Oct 12 for Blinov
yet?
THanks,
Betys
-------

∂31-Aug-89  0332	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Hickerson's Life breakthrough
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 31 Aug 89  03:32:45 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 415625; 31 Aug 89 06:31:32 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107431; Thu 31-Aug-89 03:25:46 PDT
Date: Thu, 31 Aug 89 03:24 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Hickerson's Life breakthrough
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890831102449.8.RWG@TSUNAMI.SPA.Symbolics.COM>

In response to overwhelming requests (one), here are the
goods from Penn State.
    . . .
       Anyway, I wrote a program to search for spaceships and oscillators and
    have found some new orthogonal spaceships with speeds c/2 and c/3 and some
    oscillators with periods 3, 4, and 6.  I'll mail them to you tomorrow;
    meanwhile here's the smallest of the c/3 spaceships:

	    . . . . . . . . . . . . . o o .
	    . o o . . . o . . . o . . . . o
	    o . . . . . . o o . . o o o o .
	    . o o o . o o . o . o o . . . .
	    . . . . . o . o o . . . . . . .

    Here are (I think) the smallest known oscillators with periods 3 and 4:
	    . o o . . . . .        . . o . o .
	    . . . o . . . .        . . . . . o
	    o . . . . . . .        . o . . o .
	    o . . . o . . .        o . o . o .
	    o . . . o o o o        o . . o . .
	    . . o . . . . .        . o o . . .

    I've named these "the caterer" and "mold", repectively.
				    DRH

This spaceship seems very exotic--unnatural (hard to build from
gliders), improbable (p(soup) = 10↑-69), and implausible (unlikely to
be invented by an unaided human life hacker).

The last one ("mold" is highly appropriate) is made of familiar
components in an unlikely relationship, but it seems to me the
Canadians should have found it.

I hope Dean's letter reveals whether his program was restricted
to two-step diagonal flips, or was looking for more general
period 4s.

I'm still trying to figure out the name "caterer".  Here's a long
shot:  it serves the baker's dozen!

   . . o
   . o o
   . . .
   . o .
   o . o
   o . . o . . o o
   . . . . o . o .
   . . o o . . . .
   . . . . . . . .
   . . . . . . . .
   . . . . . . . . o
   . . . . . . . o . o . . . o
   . . . . o . . o . . o . o o
   . . . o o . . . o o . . . .
   . . . . . . . . . . . . . .
   . . . . . . . . . . . . . .
   . . . . . . . . . . . . . .
   . . . . . . . . . . o o o .
   . . . . . . . . o . . . . .
   . . . . . . . . o . . . . o
   . . . . . . . . . o . . . .
   . . . . . . . . . . . o o .
   . . . . . . . . . . . . o .
   . . . . . . . . . . . . o .
   . . . . . . . . . . . . o .

This is a period 12 loaf tumbling between the central blocks,
but needs sparks at a submultiple of 12 to stay alive.  The
caterer may be the first period 3 to achieve this.  I can't
do it with the mold.  At the top, I did it with a little-
known Wainwright period 6.

∂31-Aug-89  1501	NIZ@Forsythe.Stanford.EDU 	AP newsfeed  
Received: from Forsythe.Stanford.EDU by SAIL.Stanford.EDU with TCP; 31 Aug 89  15:01:35 PDT
Date:      Thu, 31 Aug 89 15:00:04 PDT
To:        jmc@sail
From:      "Jim Nisbet" <NIZ@Forsythe.Stanford.EDU>
Subject: AP newsfeed

John,

Re:  AP Newsfeed

Your name came up a couple of times as the person who knows how
the AP newsfeed is connected to SAIL... is this true?  Is it also
true that this service will be going away soon?

If I can find out the details, then I believe I can connect it to
one of our Unix computers in the Data Center.  Are there
restrictions if we want to be able to redistribute it to other
computers on campus?

If AIR is already going to do this, then I'll wait for that.  I was
just thinking it wouldn't be too difficult to set up; so even if
this is an interim solution then it's fine with me.  I'd appreciate
any guidance you can offer.

Thanks,
/j

----

FORWARDED MESSAGE 08/30/89 22:57 FROM LES@SAIL.STANFORD.EDU "Les Earnest": re:
Access to online AP Wire/Dow Jones?

Received: by Forsythe.Stanford.EDU; Wed, 30 Aug 89 22:56:59 PDT
Message-ID: <bUDr7@SAIL.Stanford.EDU>
Date: 30 Aug 89  2258 PDT
From: Les Earnest <LES@SAIL.Stanford.EDU>
Subject: re: Access to online AP Wire/Dow Jones?
To:   TNAVARRETE@FORSYTHE.Stanford.EDU

[In reply to message sent Wed, 30 Aug 89 22:28:47 PDT.]

Tony,

Thanks for the information.  I have been communicating with Cathy Smith
about their possible involvement.  If all else fails, I suspect that
John McCarthy will find a way to keep it alive as a Computer Science
service, but I would rather see it available to the larger community.

        -Les Earnest


∂31-Aug-89  2301	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Baker's dozen correction
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 31 Aug 89  23:01:32 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 416018; 1 Sep 89 02:00:37 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107497; Thu 31-Aug-89 22:47:23 PDT
Date: Thu, 31 Aug 89 22:46 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Baker's dozen correction
To: "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: The message of 31 Aug 89 09:09 PDT from Dean Hickerson <HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu>
Message-ID: <19890901054623.2.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date:    Thu, 31 Aug 89 12:09 EDT
    From:    "Dean Hickerson" <HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu>

       In the picture you sent of the baker's dozen, the unix at the top needs to
    be moved left one unit.

Ick!  Right.  (Left!)  I guess I'd better implement a :PRINT-DOTS message to
macrocells.  Why's it called a unix?

∂01-Sep-89  0032	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Hickerson strikes again 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 1 Sep 89  00:32:05 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 416033; 1 Sep 89 03:30:46 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107517; Fri 1-Sep-89 00:22:40 PDT
Date: Fri, 1 Sep 89 00:21 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Hickerson strikes again
To: "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: The message of 31 Aug 89 04:15 PDT from Dean Hickerson <HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu>
Message-ID: <19890901072143.3.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date:    Thu, 31 Aug 89 07:15 EDT
    From:    "Dean Hickerson" <HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu>

    Bill:
       When I found mold, I was just looking for 2 generation diagonal flips.  I
    figured that since mazing

      . . o
      . . o
      o o o
 =    o o .          ?
      o o . . o o o
      . . o o o . .
      . . o o o . .
			      worked that way, it was worth looking for others.
    Here's another, which is kind of pretty:

	    o o . . . o o
	    . o . o . o .
	    . o o . o o .
	    . o . o . o .
	    o o . . . o o

Kind of!  Call it monogram, and mail it to Conway (IAS Princeton)
as a 20th anniversary present!  Gardner, too.

    I've also done some searching for more general period 4s, but haven't found
    anything smaller than mold.  But the program slows down quite a bit when
    the congruence period (# of generations until the object reappears, possibly
    in a different orientation) exceeds 3, so there could be something I've
    missed.  I don't even guarantee that mold is the smallest (in terms of
    population) diagonal flipper, just that it's the only one that stays in a
    6 by 6 square.
       The name "caterer" is derived from the fact that it provides a bit (at left
    in generation 1) which can be used in other oscillators.  Wainwright and I
    have found several such "catered" oscillators with period 3.  And now you've
    given one with period 12.

That thing lay around for years waiting for a "caterer".  I think
Wainwright was first to make the marriage.  As for catering theory,
note that I was unable to use the proffered spark, but succeeded with
the adjoining blinkopod.  Period 3 free sparks are generally too
parenchymal to withstand their own excitements.  I'd like to see
something taste that hors d'oeuvre without wasting the caterer.

I'd clean forgotten this:  ~ Oct 78 Wainwright noticed that two big
beacons (period 8) can play catch with a T4 (=
o o .
o o o
o o . )

Have you seen Buckingham's eaters?  He has a p13 based on eater-2.
I don't know what he did with eater-3, which is a truly amazing
example of two-stage self-repair.  Any idea where he is?

It would be nice to collect the smallest example of each known
period and linear family of periods.  (Although I don't feel much
like shaking out the p900 gun3!)

It probably won't matter, but math-fun spent some unsuccessful kilobytes
trying to agree on what was an oscillator (say, p6), and what was two
oscillators (say, p2 and p3).  I want to insist on a p6 bit, and that
you can't erase any of it without affecting the future of the rest
(to outlaw, e.g., lcming crossed, noninteracting glider streams).

Is that too stringent, too lax, or both?

What is the smallest period of which you are unaware?

Professor Igarashi tells me that in Japan, gliders are called
scooters.  And Marco Aiello tells me they're bullets in Italy.

∂01-Sep-89  1308	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	private  
Received: from zaphod.es.llnl.gov by SAIL.Stanford.EDU with TCP; 1 Sep 89  13:08:17 PDT
Received: by zaphod.es.llnl.gov (4.0/1.26)
	id AA06724; Fri, 1 Sep 89 13:09:10 PDT
Date: Fri, 1 Sep 89 13:09:10 PDT
From: clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov (Cynthia Mason  422-8911)
Message-Id: <8909012009.AA06724@zaphod.es.llnl.gov>
To: mccarthy@sail.stanford.edu
Subject: private

HI John,

Gotcher phone msg...Sorry I haven't gotten back to you sooner...Life
here has been a zoo, and this office is a MADHOUSE at the moment!
I haven't even had time to do my laundry since I returned from
Detroit.  Right now, am trying to move al my books out of the shelves,
so they can expand this office while I'm gone (DAI workshop), and
pay the police so they won't arrest me and take my truck away,
meet with my advisor, write, read, eat...somewhere in there I might
manage a shower!

Well, I just want to thank you for a very nice time on Wednesday.
The dinner was good.  I never thought I would eat tongue!  But you 
know I really don't think Carol Doda's a place you should take a 
girl to.  If you don't understand I will try to explain.  But I'm 
sure if RIchard knew he would not approve, nor would Carolyn!

Richard will be driving up from LA tonight.  

Cindy

∂01-Sep-89  1352	VAL  
Ullman postponed my talk until the beginning of the fall quarter.

∂01-Sep-89  1354	alex@jessica.Stanford.EDU 	["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to say 
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  13:54:41 PDT
Received: by jessica.Stanford.EDU; Fri, 1 Sep 89 13:54:57 PDT
Date: Fri, 1 Sep 1989 13:54:56 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: jmc@sail.Stanford.EDU
Subject: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to say
        the least, but may be right on ]
Message-Id: <CMM.0.88.620686496.alex@Jessica.Stanford.EDU>

Prof. McCarthy,

	I don't know if you received this or not; if you didn't, it strikes
at the core of the failures you were pointing at in the short paper you
gave me a month ago.  It's also a beautiful lesson of applied economics...

				Alex
                ---------------

Received: from forsythe.Stanford.EDU by jessica.Stanford.EDU with TCP; Fri, 1 Sep 89 06:49:12 PDT
Date:      Fri,  1 Sep 89 06:47:23 PDT
To: nethax@jessica
From: "Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU>
Subject: controversial to say the least, but may be right on

FYI...THIS OUGHT TO GENERATE A FEW FLAMES....

To:  NETHAX@JESSICA, BARRNET@UCBVAX.BERKELEY.EDU

FORWARDED MESSAGE 09/01/89 05:43 FROM TT@SH.CS.NET: controversial to say the
least, but may be right on

Received: by BITNIC (Mailer R2.03B) id 1480; Fri, 01 Sep 89 08:38:33 EST
Date:         Fri, 1 Sep 89 08:33:11 -0400
Reply-To:     TT@SH.CS.NET
Sender:       ONEnet Transition Team <TT-L@BITNIC.BITNET>
From:         farber@LINC.CIS.UPENN.EDU
Subject:      controversial to say the least, but may be right on
Comments: To: tt@SH.CS.NET
To:           Bill Yundt <GD.WHY@STANFORD>

------- Forwarded Message

Date:    Thu, 31 Aug 89 20:12:08 -0700
From:    crucible@fernwood.mpk.ca.us
To:      Distribution:;@venera.isi.edu
Subject: THE INTERNET CRUCIBLE - Volume 1, Issue 1

THE CRUCIBLE                                                  INTERNET EDITION
an eleemosynary publication of the                                August, 1989
Anterior Technology IN MODERATION NETWORK(tm)               Volume 1 : Issue 1

			     Geoff Goodfellow
				 Moderator

   In this issue:
	A Critical Analysis of the Internet Management Situation

   THE CRUCIBLE is an irregularly published, refereed periodical on the
   Internet.  The charter of the Anterior Technology IN MODERATION NETWORK
   is to provide the Internet and Usenet community with useful, instructive
   and entertaining information which satisfies commonly accepted standards
   of good taste and principled discourse.  All contributions and editorial
   comments to THE CRUCIBLE are reviewed and published without attribution.
   Cogent, cohesive, objective, frank, polemic submissions are welcomed.

Mail contributions/editorial comments to:	crucible@fernwood.mpk.ca.us

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

	 A Critical Analysis of the Internet Management Situation:
		       The Internet Lacks Governance


				  ABSTRACT

At its July 1989 meeting, the Internet Activities Board made some
modifications in the management structure for the Internet.  An outline of
the new IAB structure was distributed to the Internet engineering community
by Dr. Robert Braden, Executive Director.  In part, the open letter stated:

   "These changes resulted from an appreciation of our successes, especially
    as reflected in the growth and vigor of the IETF, and in rueful
    acknowledgment of our failures (which I will not enumerate).  Many on
    these lists are concerned with making the Internet architecture work in
    the real world."

In this first issue of THE INTERNET CRUCIBLE we will focus on the failures
and shortcomings in the Internet.  Failures contain the lessons one often
needs to achieve success.  Success rarely leads to a search for new
solutions.  Recommendations are made for short and long term improvements
to the Internet.


			A Brief History of Networking

The Internet grew out of the early pioneering work on the ARPANET.  This
influence was more than technological, the Internet has also been
significantly influenced by the economic basis of the ARPANET.

The network resources of the ARPANET (and now Internet) are "free".  There
are no charges based on usage (unless your Internet connection is via an
X.25 Public Data Network (PDN) in which case you're well endowed, or better
be).  Whether a site's Internet connection transfers 1 packet/day or a 1M
packets/day, the "cost" is the same.  Obviously, someone pays for the
leased lines, router hardware, and the like, but this "someone" is, by and
large, not the same "someone" who is sending the packets.

In the context of the Research ARPANET, the "free use" paradigm was an
appropriate strategy, and it has paid handsome dividends in the form of
developing leading edge packet switching technologies.  Unfortunately,
there is a significant side-effect with both the management and technical
ramifications of the current Internet paradigm: there is no
accountability, in the formal sense of the word.

In terms of management, it is difficult to determine who exactly is
responsible for a particular component of the Internet.  From a technical
side, responsible engineering and efficiency has been replaced by the
purchase of T1 links.

Without an economic basis, further development of short-term Internet
technology is has been skewed.  The most interesting innovations in
Internet engineering over the last five years have occurred in resource
poor, not resource rich, environments.

Some of the best known examples of innovative Internet efficiency
engineering are John Nagle's tiny-gram avoidance and ICMP source-quench
mechanisms documented in RFC896, Van Jacobsen's slow-start algorithms and
Phil Karn's retransmission timer method.

In the Nagle, Jacobsen and Karn environments, it was not possible or cost
effective to solve the performance and resource problems by simply adding
more bandwidth -- some innovative engineering had to be done.
Interestingly enough, their engineering had a dramatic impact on our
understanding of core Internet technology.

It should be noted that highly efficient networks are important when
dealing with technologies such as radio where there is a finite amount of
bandwidth/spectrum to be had.  As in the Nagle, Jacobsen and Karn cases,
there are many environments where adding another T1 link can not be used
to solve the problem.  Unless innovation continues in Internet
technology, our less than optimal protocols will perform poorly in
bandwidth or resource constrained environments.

Developing at roughly the same time as Internet technology have been the
"cost-sensitive" technologies and services, such as the various X.25-based
PDNs, the UUCP and CSNET dial-up networks.  These technologies are all
based on the notion that bandwidth costs money and the subscriber pays for
the resources used.  This has the notable effect of focusing innovation to
control costs and maximize efficiency of available resources and bandwidth.
Higher efficiency is achieved by concentrating on sending the most amount
of information through the pipe in the most efficient manner thereby making
the best use of available bandwidth/cost ratio.

For example, bandwidth conservation in the UUCP dial-up network has
multiplied by leaps and bounds in the modem market with the innovation of
Paul Baran's (the grandfather of packet switching technology) company,
Telebit, which manufactures a 19.2KB dial-up modem optimized especially for
UUCP and other well known protocol transfers.  For another example,
although strictly line-at-a-time terminal sessions are less "user friendly"
than character-oriented sessions, they make for highly efficient use of
X.25 PDN network resources with echoing and editing performed locally on
the PAD.

While few would argue the superiority of X.25 and dial-up CSNET and UUCP,
these technologies have proved themselves both to spur innovation and to be
accountable.  The subscribers to such services appreciate the cost of the
services they use, and such often costs form a well-known "line item" in
the subscriber's annual budget.

Nevertheless, the Internet suite of protocols are eminently successful,
based solely on the sheer size and rate of growth of both the Internet and
the numerous private internets, both domestically and internationally.  You
can purchase internet technology with a major credit card from a mail order
catalog.  Internet technology has achieved the promise of Open Systems,
probably a decade before OSI will be able to do so.


			   Failures of the Internet

The evolution and growth of Internet technology have provided the basis for
several failures.  We think it is important to examine failures in detail,
so as to learn from them.  History often tends to repeat itself.


Failure 1:- Network Nonmanagement

The question of responsibility in todays proliferated Internet is completely
open.  For the last three years, the Internet has been suffering from
non-management.  While few would argue that a centralized czar is necessary
(or possible) for the Internet, the fact remains there is little to be done
today besides finger-pointing when a problem arises.

In the NSFNET, MERIT is incharge of the backbone and each regional network
provider is responsible for its respective area.  However, trying to debug
a networking problem across lines of responsibility, such as intermittent
connectivity, is problematic at best.  Consider three all too true refrains
actually heard from NOC personal at the helm:

	"You can't ftp from x to y?  Try again tomorrow, it will
	 probably work then."

	"If you are not satisfied with the level of [network]
         service you are receiving you may have it disconnected."

	"The routers for network x are out of table space for routes,
	 which is why hosts on that network can't reach your new
	 (three-month old) network.  We don't know when the routers will
	 be upgraded, but it probably won't be for another year."

One might argue that the recent restructuring of the IAB may work towards
bringing the Internet under control and Dr. Vinton G. Cerf's recent
involvement is a step in the right direction.  Unfortunately, from a
historical perspective, the new IAB structure is not likely to be
successful in achieving a solution.  Now the IAB has two task forces, the
Internet Research Task Force (IRTF) and the Internet Engineering Task Force
(IETF).  The IRTF, responsible for long-term Internet research, is largely
composed of the various task forces which used to sit at the IAB level.
The IETF, responsible for the solution of short-term Internet problems, has
retained its composition.

The IETF is a voluntary organization and its members participate out of
self interest only.  The IETF has had past difficulties in solving some of
the Internet's problems (i.e., it has taken the IETF well over a year to
not yet produce RFCs for either a Point-To-Point Serial Line IP or Network
Management enhancements).  It is unlikely that the IETF has the resources
to mount a concerted attack against the problems of today's ever expanding
Internet.  As one IETF old-timer put it: "No one's paid to go do these
things, I don't see why they (the IETF management) think they can tell us
what to do" and "No one is paying me, why should I be thinking about the
these things?"

Even if the IETF had the technical resources, many of the Internet's
problems are also due to lack of "hands on" management.  The IETF

	o  Bites off more than it can chew;
	o  Sometimes fails to understand a problem before making a solution;
	o  Attempts to solve political/marketing problems with technical
	     solutions;
	o  Has very little actual power.

The IETF has repeatedly demonstrated the lack of focus necessary to
complete engineering tasks in a timely fashion.  Further, the IRTF is
chartered to look at problems on the five-year horizon, so they are out of
the line of responsibility.  Finally, the IAB, per se, is not situated to
resolve these problems as they are inherent to the current structure of
nonaccountability.

During this crisis of non-management, the Internet has evolved into a
patch quilt of interconnected networks that depend on lots of
seat-of-the-pants flying to keep interoperating.  It is not an unusual
occurrence for an entire partition of the Internet to remain disconnected
for a week because the person responsible for a key connection went on
vacation and no one else knew how to fix it.  This situation is but one
example of an endemic problem of the global Internet.


Failure 2:- Network Management

The current fury over network management protocols for TCP/IP is but a
microcosm of the greater Internet vs. OSI debate going on in the
marketplace.  While everyone in the market says they want OSI, anyone
planning on getting any work done today buys Internet technology.  So it
is with network management, the old IAB made the CMOT an Internet
standard despite the lack of a single implementation, while the only
non-proprietary network management protocol in use in the Internet is the
SNMP.  The dual network management standardization blessings will no
doubt have the effect of confusing end-users of Internet
technology--making it appear there are two choices for network
management, although only one choice, the SNMP has been implemented.  The
CMOT choice isn't implemented, doesn't work, or isn't interoperable.

To compound matters, after spending a year trying to achieve consensus on
the successor to the current Internet standard SMI/MIB, the MIB working
group was disbanded without ever producing anything: the political climate
prevented them from resolving the matter.  (Many congratulatory notes were
sent to the chair of the group thanking him for his time.  This is an
interesting new trend for the Internet--congratulating ourselves on our
failures.)

Since a common SMI/MIB could not be advanced, an attempt was made to
de-couple the SNMP and the CMOT (RFC1109).  The likely result of RFC1109
will be that the SNMP camp will continue to refine their experience
towards workable network management systems, whilst the CMOT camp will
continue the never-ending journey of tracking OSI while producing demo
systems for trade shows exhibitions.  Unfortunately the end-user will
remain ever confused because of the IAB's controversial (and technically
questionable) decision to elevate the CMOT prior to implementation.

While the network management problem is probably too large for the SNMP
camp to solve by themselves they seem to be the only people who are
making any forward progress.


Failure 3:- Bandwidth Waste

Both the national and regional backbone providers are fascinated with T1
(and now T3) as the solution towards resource problems.  T1/T3 seems to
have become the Internet panacea of the late 80's.  You never hear anything
from the backbone providers about work being done to get hosts to implement
the latest performance/congestion refinements to IP, TCP, or above.
Instead, you hear about additional T1 links and plans for T3 links.  While
T1 links certainly have more "sex and sizzle" than efficient technology
developments like slow-start, tiny gram avoidance and line mode telnet, the
majority of users on the Internet will probably get much more benefit from
properly behaving hosts running over a stable backbone than the current
situation of misbehaving and semi-behaved hosts over an intermittent catenet.


Failure 4:- Routing

The biggest problem with routing today is that we are still using phase I
(ARPANET) technology, namely EGP.  The EGP is playing the role of routing
glue in providing the coupling between the regional IGP and the backbone
routing information.  It was designed to only accommodate a single point of
attachment to the catenet (which was all DCA could afford with the PSNs).
However with lower line costs, one can build a reasonably inexpensive
network using redundant links.  However the EGP does not provide enough
information nor does the model it is based upon support multiple
connections between autonomous systems.  Work is progressing in the
Interconnectivity WG of the IETF to replace EGP.  They are in the process
of redefining the model to solve some of the current needs.  BGP or the
Border Gateway Protocol (RFC1105) is an attempt to codify some of the ideas
the group is working on.

Other problems with routing are caused by regionals wanting a backdoor
connection to another regional directly.  These connections require some
sort of interface between the two routing systems.  These interfaces are
built by hand to avoid routing loops.  Loops can be caused when information
sent into one regional network is sent back towards the source.  If the
source doesn't recognize the information as its own, packets can flow until
their time to live field expires.

Routing problems are caused by the interior routing protocol or IGP.  This
is the routing protocol which is used by the regionals to pass information
to and from its users.  The users themselves can use a different IGP than
the regional.  Depending on the number of connections a user has to the
regional network, routing loops can be an issue.  Some regionals pass
around information about all known networks in the entire catenet to their
users.  This information deluge is a problem with some IGPs.  Newer IGPs
such as the new OSPF from the IETF and IGRP from cisco attempt to provide
some information hiding by adding hierarchy.  OSPF is the internets first
attempt at using a Dykstra type algorithm as an IGP.  BBN uses it to route
between their packet switch nodes below the 1822 or X.25 layer.

Unstable routing is caused by hardware or hosts software.  Older BSD
software sets the TTL field in the IP header to a small number.  The
Internet today is growing and its diameter has exceed the software's
ability to reach the other side.  This problem is easily fixed by
knowledgeable systems people, but one must be aware of the problem before
they can fix it.

Routing problems are also perceived when in fact a serial line problem or
hardware problem is the real cause.  If a serial line is intermittent or
quickly cycles from the up state into the down state and back again,
routing information will not be supplied in a uniform or smooth manner.
Most current IGPs are Bellman-Ford based and employ some stabilizing
techniques to stem the flow of routing oscillations due to "flapping"
lines.  Often when a route to a network disappears, it may take several
seconds for it to reappear.  This can occur at the source router who waits
for the route to "decay" from the system.  This pause should be short
enough so that active connections persist but long enough that all routers
in the routing system "forget" about routes to that network.  Older host
software with over-active TCP retransmission timers will time out
connections instead of persevering in the face of this problem.  Also
routers, according to RFC1009, must be able to send ICMP unreachables when
a packet is sent to a route which is not present in its routing database.
Some host products on the market close down connections when a single ICMP
reachable is received.  This bug flies in the face of the Internet parable
"be generous in what you accept and rigorous in what you send".

Many of the perceived routing problems are really complex multiple
interactions of differing products.


			    Causes of the Failures

The Internet failures and shortcomings can be traced to several sources:

First and foremost, there is little or no incentive for efficiency and/or
economy in the current Internet.  As a direct result, the resources of the
Internet and its components are limited by factors other than economics.
When resources wear thin, congestion and poor performance result.  There is
little to no incentive to make things better, if 1 packet out of 10 gets
through things "sort of work".  It would appear that Internet technology
has found a loophole in the "Tragedy of The Commons" allegory--things get
progressively worse and worse, but eventually something does get through.

The research community is interested in technology and not economics,
efficiency or free-markets.  While this tack has produced the Internet
suite of protocols, the de facto International Standard for Open Systems,
it has also created an atmosphere of intense in-breeding which is overly
sensitive to criticism and quite hardened against outside influence.
Meanwhile, the outside world goes on about developing economically viable
and efficient networking technology without the benefit of direct
participation on the part of the Internet.

The research community also appears to be spending a lot of its time
trying to hang onto the diminishing number of research dollars available
to it (one problem of being a successful researcher is eventually your
sponsors want you to be successful in other things).  Despite this, the
research community actively shuns foreign technology (e.g., OSI), but,
inexplicably has not recently produced much innovation in new Internet
technology.  There is also a dearth of new and nifty innovative
applications on the Internet.  Business as usual on the Internet is mostly
FTP, SMTP and Telnet or Rlogin as it has been for many years.  The most
interesting example of a distributed application on the Internet today is
the Domain Name System, which is essentially an administrative facility,
not an end-user service.

The engineering community must receive equal blame in these matters.
While there have been some successes on the part of the engineering
community, such as those by Nagel, Jacobsen and Karn mentioned above, the
output of the IETF, namely RFCs and corresponding implementations, has
been surprisingly low over its lifetime.

Finally, the Internet has become increasingly dependent on vendors for
providing implementations of Internet technology.  While this is no doubt
beneficial in the long-term, the vendor community, rather than investing
"real" resources when building these products, do little more than
shrink-wrap code written primarily by research assistants at universities.
This has lead to cataclysmic consequences (e.g., the Internet worm
incident, where Sendmail with "debug" command and all was packaged and
delivered to customers without proper consideration).  Of course, when
problems are found and fixed (either by the vendor's customers or software
sources), the time to market with these fixes is commonly a year or longer.
Thus, while vendors are vital to the long-term success of Internet
technology, they certainly don't receive high marks in the short-term.


			       Recommendations

Short-term solutions (should happen by year's end):

In terms of hardware, the vendor community has advanced to the point where
the existing special-purpose technologies (Butterfly, NSSs) can be replaced
by off-the-shelf routers at far less cost and with superior throughput and
reliability.  Obvious candidates for upgrade are both the NSFNET and
ARPANET backbones.  Given the extended unreliability of the mailbridges,
the ARPA core is an immediate candidate (even though the days of net 10 are
numbered).

In terms of software, ALL devices in the Internet must be network
manageable.  This is becoming ever more critical when problems must be
resolved.  Since SNMP is the only open network management protocol
functioning in the Internet, all devices must support SNMP and the Internet
standard SMI and MIB.

Host implementations must be made to support the not-so-recent TCP
enhancements (e.g., those by Nagle, Jacobsen and Karn) and the more
recent linemode TELNET.

The national and regional providers must coordinate to share network
management information and tools so that user problems can be dealt with in
a predictable and timely fashion.  Network management tools are are a big
help, but without the proper personnel support above this, the benefits can
not be fully leveraged.

The Internet needs leadership and hands-on guidance.  No one is seemingly
in charge today, and the people who actually care about the net are pressed
into continually fighting the small, immediate problems.

Long-term solutions:

To promote network efficiency and a free-market system for the delivery of
Internet services, it is proposed to switch the method by which the network
itself is supported.  Rather than a top-down approach where the money goes
from funding agencies to the national backbone or regional providers, it is
suggested the money go directly to end-users (campuses) who can then select
from among the network service providers which among them best satisfies
their needs and costs.

This is a strict economic model: by playing with the full set of the laws of
economics, a lot of the second-order problems of the Internet, both present
and on the horizon, can be brought to heel.  The Internet is no longer a
research vehicle, it is a vibrant production facility.  It is time to
acknowledge this by using a realistic economic model in the delivery of
Internet services to the community (member base).

When Internet sites can vote with their pocketbooks, some new regionals
will be formed; some, those which are non-performant or uncompetitive, will
go away; and, the existing successful ones will grow.  The existing
regionals will then be able to use their economic power, as any consumer
would, to ensure that the service providers (e.g., the national backbone
providers) offer responsive service at reasonable prices.  "The Market" is
a powerful forcing function: it will be in the best interests of the
national and regional providers to innovate, so as to be more competitive.
Further, such a scheme would also allow the traditional telecommunications
providers a means for becoming more involved in the Internet, thus allowing
cross-leverage of technologies and experience.

The transition from top-down to economic model must be handled carefully,
but this is exactly the kind of statesmanship that the Internet should
expect from its leadership.
- -------

------- End of Forwarded Message


∂01-Sep-89  1404	MPS 	JMC @ NCSU
 ∂01-Sep-89  1247	drb@cscfac.ncsu.edu 	JMC @ NCSU    
Received: from ncsuvx (ncsuvx.ncsu.edu) by SAIL.Stanford.EDU with TCP; 1 Sep 89  12:47:30 PDT
Received: by ncsuvx (5.57/2 4/27/87)
	id AA29973; Fri, 1 Sep 89 15:46:51 EDT
Posted-Date: Fri, 1 Sep 89 15:48:40 edt
Received: by cscfac.ncsu.edu (1.2/Ultrix2.0-B)
	id AA00697; Fri, 1 Sep 89 15:48:40 edt
Date: Fri, 1 Sep 89 15:48:40 edt
From: drb@cscfac.ncsu.edu (Dennis R. Bahler)
Message-Id: <8909011948.AA00697@cscfac.ncsu.edu>
To: mps@sail.stanford.edu
Subject: JMC @ NCSU

Pat:

A while ago I got you to ask Prof. McCarthy if during his visit to NC State
he could give a technical talk for CS faculty/gradstudents on Monday as well
as his public lecture Tuesday evening.  (This is our video CS seminar series.)
You indicated he said OK.  We have Sept. 25 reserved.

Can we get a title/abstract of this talk for the publicity?

The brochures/posters have gone out for the cognitive science series.
and everyone is excited.
JMC is now first (of four) as Chomsky was laid up and had to reschedule.

Thx again for everything,

Dennis Bahler
Dept. of Computer Science Box 8206 INTERNET - drb@cscadm.ncsu.edu 
North Carolina State University    CSNET    - drb%cscadm.ncsu.edu@relay.cs.net
Raleigh, NC   27695-8206           UUCP     - ...!decvax!mcnc!ncsu!cscadm!drb

∂01-Sep-89  1427	alex@jessica.Stanford.EDU 	re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  14:27:37 PDT
Received: by jessica.Stanford.EDU; Fri, 1 Sep 89 14:27:53 PDT
Date: Fri, 1 Sep 1989 14:27:51 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to
        say 
In-Reply-To: Your message of 01 Sep 89 1413 PDT 
Message-Id: <CMM.0.88.620688471.alex@Jessica.Stanford.EDU>

You're welcome.  By the way, does that mean that you are on NETHAX and
therefore I shouldn't forward stuff from there to you?

				Alex

∂01-Sep-89  1446	alex@jessica.Stanford.EDU 	re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  14:46:05 PDT
Received: by jessica.Stanford.EDU; Fri, 1 Sep 89 14:46:19 PDT
Date: Fri, 1 Sep 1989 14:46:18 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
In-Reply-To: Your message of 01 Sep 89 1434 PDT 
Message-Id: <CMM.0.88.620689578.alex@Jessica.Stanford.EDU>

Well, actually, NETHAX is really just that: just a bunch of local hackers
trying to keep the local network operational.  What I would suggest is that
you send your memo to the editor/creator of this (new) Crucible electronic
journal.  From what I know of him, he is intelligent and entrepreneurial,
and that way your paper would have a much broader audience (with maybe
even some decision makers).

By the way, did you get that ad for a Fax card for PCs that I slipped under
your door a few weeks ago?

				Alex

ps: I'm on the Nethax list 'cause I used to be a system manager for a couple
of VAXes at LOTS...  (Lancelot & Camelot, long dead.)

∂01-Sep-89  1516	CLT  
rwg@yukon.SCRC.symbolics.com

∂01-Sep-89  1518	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	re: private   
Received: from zaphod.es.llnl.gov by SAIL.Stanford.EDU with TCP; 1 Sep 89  15:18:23 PDT
Received: by zaphod.es.llnl.gov (4.0/1.26)
	id AA06821; Fri, 1 Sep 89 15:19:10 PDT
Date: Fri, 1 Sep 89 15:19:10 PDT
From: clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov (Cynthia Mason  422-8911)
Message-Id: <8909012219.AA06821@zaphod.es.llnl.gov>
To: JMC@sail.stanford.edu
Subject: re: private

OK...Well, I figured you might wanna know that Rich was coming, so
you could decide...It does seem we have a special kind of friendship
and its nice to have some freedom to chat between the two of us...

Re: Carol Doda's - No worries, MATE (spoken with a thick Aussie accent)

                        -------------------------------
                        |           Coupon            |
                        |             1               |
                        |                             |
                        |       Lasagne Dinner        |
                        |                             |
                        |        Mason's Cafe         |
                        |_____________________________|



Yours truly,

May K. Fist

∂01-Sep-89  1524	yeager@sumex-aim.stanford.edu 	Re: Final version of Networks considered harmful.
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 1 Sep 89  15:24:50 PDT
Received: from  (KSL-MAC-70.Stanford.EDU) by sumex-aim.stanford.edu (4.0/inc-1.0)
	id AA19052; Fri, 1 Sep 89 15:27:22 PDT
Date: Fri, 1 Sep 1989 15:25:58 PDT
From: Bill Yeager's Mac <yeager@sumex-aim.stanford.edu>
Subject: Re: Final version of Networks considered harmful.
To: John McCarthy <JMC@sail.stanford.edu>
In-Reply-To: Your message of 01 Sep 89 1415 PDT
Message-Id: <MacMM.20486.5758.yeager@sumex-aim.stanford.edu>

John,

An excellent set of ideas with respect to email and its delivery limitations.

We are really blocked by concepts that originated in the late 60's and early
70's and have so much momentum (vested interest), that they may be difficult
to overcome in the near future. Most likely, some startup will do what is
necessary to send/receive ascii email anywhere there is an associated phone
number, and a small piece of hardware (modem - simple mail receiver/sender in
PROM - minimum amount of disk storage), and the world of network mail will die
out from disuse.

Thanks for forwarding the message,

Bill

-------

∂01-Sep-89  1557	alex@jessica.Stanford.EDU 	re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  15:57:12 PDT
Received: by jessica.Stanford.EDU; Fri, 1 Sep 89 15:57:29 PDT
Date: Fri, 1 Sep 1989 15:57:28 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: ["Bill Yundt" <GD.WHY@Forsythe.Stanford.EDU> : controversial to 
In-Reply-To: Your message of 01 Sep 89 1449 PDT 
Message-Id: <CMM.0.88.620693848.alex@Jessica.Stanford.EDU>

Well, you already sent it to NETHAX@Jessica (just a short time ago) so that
takes care of that.

The moderator is: Geoff Goodfellow, crucible@fernwood.mpk.ca.us  
or so it says in the big msg he sent, although I don't understand that email
address.  If it fails, try looking up his company (given in the big msg)
in the phonbook... :)

				Alex

∂01-Sep-89  1600	MPS 	TGIF 
There is a TGIF in Nils conference room for the staff.
If you need me I will be in Nils conference room.
Have a good weekend.  See you Tuesday.

Pat

∂01-Sep-89  1717	alex@jessica.Stanford.EDU 	re: Goodfellow    
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  17:17:36 PDT
Received: by jessica.Stanford.EDU; Fri, 1 Sep 89 17:17:53 PDT
Date: Fri, 1 Sep 1989 17:17:52 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: Goodfellow 
In-Reply-To: Your message of 01 Sep 89 1646 PDT 
Message-Id: <CMM.0.88.620698672.alex@Jessica.Stanford.EDU>

Anterior Technology.

Note that the email address crucible@fernwood.mpk.ca.us  seems to exist.

				Alex

∂01-Sep-89  1732	alex@jessica.Stanford.EDU 	re: Goodfellow    
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  17:32:26 PDT
Received: by jessica.Stanford.EDU; Fri, 1 Sep 89 17:32:43 PDT
Date: Fri, 1 Sep 1989 17:32:42 PDT
From: Alex Bronstein <alex@jessica.stanford.edu>
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: Goodfellow 
In-Reply-To: Your message of 01 Sep 89 1724 PDT 
Message-Id: <CMM.0.88.620699562.alex@Jessica.Stanford.EDU>

Well, it does seem like a weird name, but there was a quite formal-looking
announcement on the network when he created it.  Officially he is in the
business of provided MODERATED discussion forums on the internet.  He claimed
that there was so much drivel on Unix news and other such bboards that he
thought people would be willing to pay for moderated service.  It may have
been a big joke, but it sounded reasonable.   If you ever find out the 
scoop, let me know.

				Alex

∂01-Sep-89  1804	@bridge2.ESD.3Com.COM:gpz@api.ESD.3Com.COM 	Re: Final version of Networks considered harmful.  
Received: from bridge2.ESD.3Com.COM by SAIL.Stanford.EDU with TCP; 1 Sep 89  18:04:47 PDT
Received: from api.ESD.3Com.COM by bridge2.ESD.3Com.COM with SMTP (5.61++/IDA-1.2.8)
	id AA06277 (for JMC@SAIL.STANFORD.EDU); Fri, 1 Sep 89 18:05:31 PDT
Received: by api.ESD.3Com.COM (3.2/SMI-3.0DEV3-890516)
	id AA18489; Fri, 1 Sep 89 18:04:46 PDT
Date: Fri, 1 Sep 89 18:04:46 PDT
From: gpz@api.ESD.3Com.COM (G. Paul Ziemba)
Message-Id: <8909020104.AA18489@api.ESD.3Com.COM>
To: JMC@SAIL.STANFORD.EDU
Subject: Re: Final version of Networks considered harmful.
Newsgroups: su.nethax
References: <1VVrQe@SAIL.Stanford.EDU>

Hi John,

this is just a nit-pick; do what you will :-)


In su.nethax you write:

>This is more-or-less accepted for publication late this year in CACM.
>Many thanks to everyone who commented.  Some of the comments
>have been incorporated.
>NETWORKS CONSIDERED HARMFUL - FOR ELECTRONIC MAIL

>	The reason why telefax will supplant email unless email
                   ↑↑↑

I believe you mean "the reason that..."
                               ----


yours for concise english,

 ~!paul

--
----

Paul Ziemba  ...!pyramid!zapi!gpz  gpz@bridge2.3com.com
		(415)940-7671 (w)

∂01-Sep-89  2000	JMC  
paper for Ursula

∂01-Sep-89  2051	BEDIT@Score.Stanford.EDU 	Summary of August computer charges.    
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Sep 89  20:51:04 PDT
Date: Fri 1 Sep 89 20:39:11-PDT
From: Billing Editor <BEDIT@Score.Stanford.EDU>
Subject: Summary of August computer charges.
To: MCCARTHY@Score.Stanford.EDU
Message-ID: <12522920926.10.BEDIT@Score.Stanford.EDU>

Dear Mr. McCarthy,

Following is a summary of your computer charges for August.

Account     System   Billed    Pct      Cpu    Job   Disk  Print   Adj   Total

JMC         SAIL     2-DMA816  100   377.12  21.48 ***.**  26.00  5.00 2718.06
MCCARTHY    SCORE    2-DMA816  100      .00    .00  30.47    .00  5.00   35.47

Total:                               377.12  21.48 ***.**  26.00 10.00 2753.53


University budget accounts billed above include the following. 

Account     Princip Inv      Title                      Comment             

2-DMA816    McCarthy         N00039-84-C-0211           Task 23              


The preceding statement is a condensed version of the detailed summary sheet 
sent monthly to your department. 

Please verify each month that the proper university budget accounts are paying 
for your computer usage.  Please also check the list of account numbers below 
the numeric totals.  If the organizations/people associated with that account 
number should NOT be paying for your computer time, send mail to BEDIT@SCORE. 

Please direct questions/comments to BEDIT@SCORE. 
-------

∂02-Sep-89  0445	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	this is a test     
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 2 Sep 89  04:45:25 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 416488; 2 Sep 89 07:46:50 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107616; Sat 2-Sep-89 04:41:16 PDT
Date: Sat, 2 Sep 89 04:40 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: this is a test 
To: "JMC@SAIL.Stanford.EDU"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <12VtUg@SAIL.Stanford.EDU>
Message-ID: <19890902114018.0.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: 01 Sep 89  1636 PDT
    From: John McCarthy <JMC@SAIL.Stanford.EDU>

    test

YOW, ya got me, podna!

∂02-Sep-89  1302	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk 	theorem proving list
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Sep 89  13:01:59 PDT
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU.#Internet) by SCORE.STANFORD.EDU with TCP; Sat 2 Sep 89 13:00:44-PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa03518;
          2 Sep 89 15:51 EDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU  2 Sep 89 15:50:08 EDT
Received: from lcs.mit.edu (CHAOS 15044) by MC.LCS.MIT.EDU  2 Sep 89 15:47:35 EDT
Received: from NSFNET-RELAY.AC.UK by mintaka.lcs.mit.edu id aa03467;
          2 Sep 89 15:45 EDT
Received: from prg.oxford.ac.uk by NSFnet-Relay.AC.UK   via Janet with NIFTP
           id aa07175; 2 Sep 89 20:32 BST
Received: from uk.ac.oxford.prg.client80 (client80) by uk.ac.ox.prg (4.12/prgv.37)
	id AA27481; Sat, 2 Sep 89 16:32:38 bst
Received: by uk.ac.oxford.prg.client80 (3.2/prg.1)
	id AA17291; Sat, 2 Sep 89 16:38:02 BST
Date: Sat, 2 Sep 89 16:38:02 BST
From: goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk
Message-Id: <8909021538.AA17291@uk.ac.oxford.prg.client80>
To: theorem-provers%mc.lcs.mit.edu@nsfnet-relay.ac.uk
Cc: goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk
Subject: theorem proving list

I hear that there is an emailist for theorem provers at MIT and that this
might be the address.  If so, could whoever is in charge please place me on
the list?  And so those of you who read these things wont be bored or upset,
let me also say that here at Oxford, we are making some substantial progress
on the use of OBJ3 as a theorem prover, and in particular, we are developing
techniques for compiling from other logics down into conditional order-sorted
equational rewriting.  Some early results are in the paper "OBJ as a Theorem
Prover" in the Birtwistle and Subrahmanyam edited volume "Current Trends in
Hardware Verification and Automated Theorem Proving" (Springer-Verlag, 1989).

Cheerio,
Joseph



∂02-Sep-89  1727	CLT  
have you been reimbursed for 255 to AAAI?

∂02-Sep-89  2158	root@NSFnet-Relay.AC.UK 	Mail Delivery Failure to uk.ac.hw.cs - Timeout    
Received: from NSFnet-Relay.AC.UK by SAIL.Stanford.EDU with TCP; 2 Sep 89  21:58:40 PDT
Date:     Sun, 3 Sep 89 5:48:24 BST
From:     root@NSFnet-Relay.AC.UK
Subject:  Mail Delivery Failure to uk.ac.hw.cs - Timeout
To:       JMC@sail.stanford.edu

The NIFTP process was unable to
deliver your mail to host uk.ac.hw.cs
over janet.

The reason given by the local host was:


The NIFTP process gave up after 6 attempts over 171 hours


Your message was not delivered to the following addresses:
	greg@uk.ac.heriot-watt.cs


	Your message begins as follows:

Received: from sail.stanford.edu by NSFnet-Relay.AC.UK   via NSFnet with SMTP
           id aa10301; 27 Aug 89 2:42 BST
Message-ID: <fSv2m@SAIL.Stanford.EDU>
Date: 26 Aug 89  1815 PDT
From: John McCarthy <JMC@edu.stanford.sail>
Subject: re: Functional language workshop   
To:   greg <greg%uk.ac.heriot-watt.cs@uk.ac.nsfnet-relay>

[In reply to message sent Mon, 21 Aug 89 09:54:15 BST.]

I'm well, but accepting your invitation would require combining
it with another trip.  When you have a date, let me know, and
I'll put it in my calendar and see what turns up.

∂03-Sep-89  0353	RFC 	Prancing Pony Bill  
Prancing Pony bill of     JMC   John McCarthy      3 September 1989

Previous Balance             8.34
Monthly Interest at  1.0%    0.08
Current Charges              0.30  (vending machine)
                           -------
TOTAL AMOUNT DUE             8.72


PAYMENT DELIVERY LOCATION: CSD Receptionist.

Make checks payable to:  STANFORD UNIVERSITY.
Please deliver payments to the Computer Science Dept receptionist, Jacks Hall.
To ensure proper crediting, please include your PONY ACCOUNT NAME on your check.

Note: The recording of a payment takes up to three weeks after the payment is
made, but never beyond the next billing date.  Please allow for this delay.

Bills are payable upon presentation.  Interest of  1.0% per month will be
charged on balances remaining unpaid 25 days after bill date above.

An account with a credit balance earns interest of  .33% per month,
based on the average daily balance.

Your last Pony payment was recorded on 6/22/89.

Accounts with balances remaining unpaid for more than 55 days are
considered delinquent and are subject to reduction of credit limit.
Please pay your bill and keep your account current.

∂03-Sep-89  1759	splat!root@ssyx.ucsc.edu 	regarding "Networks Considered Harmful"
Received: from labrea.stanford.edu by SAIL.Stanford.EDU with TCP; 3 Sep 89  17:49:53 PDT
Received: from ssyx.UCSC.EDU by labrea.stanford.edu with TCP; Sun, 3 Sep 89 17:49:15 PDT
Received: from splat.UUCP by ssyx.ucsc.edu (4.0/1.1)
	id AA17600; Sun, 3 Sep 89 17:50:08 PDT
Received: by Splat.Aptos.CA.US (smail2.5)
	id 09376; 3 Sep 89 17:28:13 PDT (Sun)
From: splat!root@ssyx.ucsc.edu (brad allen)
Date: Sun, 3 Sep 89 17:28:12 PDT
X-Mailer: Mail User's Shell (6.5.6 6/30/89)
To: JMC@sail.stanford.edu
Subject: regarding "Networks Considered Harmful"
Message-Id: <19890903.1728,09376@Splat.Aptos.CA.US>

It sounds somewhat pop-solutionary.
Most of it was basically correct.
You had a part near the middle of the article where
you imply that erradicating the current networking companies will help.

I'm not sure about that.
I think a better way is to offer something better,
and let the older ways disappear under its own weight.
The difficulty is offering these better things.

Basically I think it requires putting fiber to the house.
This isn't outrageously expensive for current computer geeks considering
most modem users pay the phone company significant amounts of money
for lousy equipment and mediocre service.  (It's fine for voice,
but why it's still used so much for digitial the way it is
is almost beyond me.)  Small companies could do this.
This in itself would offer a similar thing that your solution would
offer:  yet another company, yet another protocol, and yet another mess.
Offering decent (reasonable) rates for gatewaying from one service
to another is one way to bridge this gap:  conceivably, a system
could be automated with those new something-hundred numbers we dial
where digital, voice, fax, etc. could be converted via means of
robot and human operator.  (I mail to someone, a machine redials until
connect, a human operator tries to get ahold of the correct person until
she does, and then the operator or robot could give the message.)
Similar with US Mail:  Why can't I put a sticky stamp on something
which will eventually end up in my friend's Email box?

This would make all kinds of networks more accessible, and whatever
is better will more quickly stand up to existence, acceptance and use.

Well, I must comment that one of the most significant problems is
how to address the recipient.  Should it be by name and location, 
like US Mail?  I'm
Brad Allen
Aptos, Santa Cruz County, California, North America, Earth
but, what if I'm not in Aptos?
This is about the best I can get, too; saying things like this:
Brad Allen
1713 Jennifer Drive
Aptos, CA 95003
is potentially hazardous to my safety; saying
Brad Allen
USA SSN 564-91-4465
often is related to a military type culture, and there's probably some
reasoning behind that fear; what I say now is
Brad Allen
P.O. Box 1342
Aptos, CA 95001-1342
 (I have to check that mailbox periodically)
-or-
Brad Allen
phone# +1 408 688 8562 (this gets to a person which will tell the calling
party what phone # I'm staying at at the moment)
-or-
Brad Allen
<Ulmo@Splat.Aptos.Ca.US>
This last one is a prime example:  one week ago today, I would have said
Brad Allen
ulmo@ssyx.ucsc.edu
but they deleted my ssyx account.  I was using ssyx as my login host for
over one year.


I must be blabeling:  most people would have their own personal directory
of what "George" translates to, and the mail would go there; but the problem
is no longer that but which network to use to get to George.
I'm another good example of this.
I have personal mailboxes on Gorn, Splat, Cencom, Comix, SSYX, Zuul,
Pyrzqxgl, etc.  I prefer Zuul.  Unfortunately, Zuul just crashed
(after being up for over four years) four days ago.  Right now I'm at
Splat but that will change coming this September 7th, when I will have
no computer.  SSYX is defunct for me to use.  Gorn is always busy.
That's just the BBS/Internet world.  Then there's the choice between
phone, postal, MCI Mail, etc. etc. etc.


There are certain features which are more than just "conveniences"!!
That's why most of us don't use the telephone:  I'm not going to answer
something at the whim of someone else if I am otherwise occupied.


I have an ATM card.  Neat thing.  I stick it in this machine which says:

STAR PLUS MASTERTELLER 

Well, my card subscribes to all three of those.  Which network does the
card use?  I never tell it which one to use.  It just works -- the way
it SHOULD happen, but something strange politically must happen.


Those are simple questions with fairly simple answers, I think,
but it seems pretty odd of you to go jumping to a solution without
at least backing it up, considering that there are so many problems with it.

Certainly something of your size and effort would could conceivably
work pretty well, but I ask that you make sure it is reasonably
gatewayed with the known world.


brad allen
5 year bbs user
ulmo@splat.aptos.ca.us or Ulmo on Pyrzqxgl (408) 476-4633

∂04-Sep-89  1459	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	simplification challenge
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 4 Sep 89  14:57:44 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 416605; 4 Sep 89 10:02:29 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107640; Mon 4-Sep-89 04:21:02 PDT
Date: Mon, 4 Sep 89 04:20 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: simplification challenge
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, macsyma-i@RUSSIAN.SPA.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890904112000.4.RWG@TSUNAMI.SPA.Symbolics.COM>

  /
 |    1
 | ------- dx = 
 |      24
/  1 - x

                               2
                            2 x  - (sqrt(6) + sqrt(2)) x + 2           (sqrt(6) + sqrt(2)) x
   (sqrt(6) + sqrt(2)) (log(--------------------------------) + 2 atan(---------------------))
                                2                                                2
                            2 x  + (sqrt(6) + sqrt(2)) x + 2                 2 (x  - 1)
 - ---------------------------------------------------------------------------------------------
                                                96

                               2
                            2 x  + (sqrt(6) - sqrt(2)) x + 2           (sqrt(6) - sqrt(2)) x
   (sqrt(6) - sqrt(2)) (log(--------------------------------) - 2 atan(---------------------))
                               2                                                2
                            2 x  + (sqrt(2) - sqrt(6)) x + 2                2 (x  - 1)
 + -------------------------------------------------------------------------------------------
                                               96

                2                                 2                         2
               x  - sqrt(3) x + 1                x  - sqrt(2) x + 1        x  - x + 1
   sqrt(3) log(------------------) + sqrt(2) log(------------------) + log(----------)
                2                                 2                         2
               x  + sqrt(3) x + 1                x  + sqrt(2) x + 1        x  + x + 1
 - -----------------------------------------------------------------------------------
                                           48

        sqrt(3) x         sqrt(2) x           x           x - 1
   atan(---------)   atan(---------)   atan(------) + log(-----)
          2                 2                2            x + 1
         x  - 1            x  - 1           x  - 1                 atan(x)
 - --------------- - --------------- - ------------------------- + -------
      8 sqrt(3)        12 sqrt(2)                 24                 12

∂04-Sep-89  1543	balzer@vaxa.isi.edu 	Publication of ISAT Prototyping White Paper 
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 4 Sep 89  15:42:44 PDT
Received: from vaxa.isi.edu by vax.darpa.mil (5.61/5.61+local)
	id <AA08097>; Mon, 4 Sep 89 18:35:10 -0400
Posted-Date: Mon, 04 Sep 89 15:39:57 PDT
Message-Id: <8909042240.AA21868@vaxa.isi.edu>
Received: from LOCALHOST by vaxa.isi.edu (5.61/5.61)
	id AA21868; Mon, 4 Sep 89 15:40:16 -0700
To: cpl@vax.darpa.mil
From: balzer@VAXA.ISI.EDU
Subject: Publication of ISAT Prototyping White Paper
Date: Mon, 04 Sep 89 15:39:57 PDT
Sender: balzer@vaxa.isi.edu

Last Spring, while we were working on the CPS report, ISAT (an advisory
panel for ISTO) asked me to prepare a draft White Paper on Prototyping.
I composed one by extracting the substance of our evolving report and
added information related to their requests for Grand Challenges, Potential
Breakthroughs, etc., and sent the draft around last summer before presenting
it at ISAT last August (Dick Gabriel attended that session with me).

ISAT has been collecting and reviewing similar White Papers from many
areas, and is now ready to publish some of them in a special section
of the Annual Review of Computer Science.  

They've asked me to submit a revised version of our draft. I reread it
and performed minor surgery to reflect the final version of our report
(especially dropping the requirement to support Common Lisp in addition 
to Ada).

I've enclosed the revised version below. Please check it and indicate
any additonal changes that should be made.

Regards,
Bob
P.S. Rumor has it that announcement of the awards for the first round of
the CPS program will soon be made.

\input macros
\drafttrue
\tolerance=2500
\def\bookline{ISAT Prototyping White Paper DRAFT}
\def\chapline{ISAT Prototyping White Paper DRAFT}

\beginChapter 1.{ISAT Prototyping White Paper DRAFT}%
{ISAT Prototyping White Paper DRAFT}{ISAT Prototyping White Paper DRAFT}

Authors: Robert Balzer, Frank Belz, Robert Dewar, David Fisher, \hfil\break
Richard P. Gabriel, John Guttag, Paul Hudak, Mitchell Wand.

Draft Dated: \timestamp\hfil\break
Limited Distribution and Not for Publication\hfil\break
All Rights Reserved

\notableofcontents
\endTitlePage

\beginSection{Scope}
\beginsubSection{Definition}


{\bit Prototyping} is the process of writing programs for the purpose of
learning.  It is a tool for obtaining information early to reduce the
cost of rework.  Despite its benefits, no strong prototyping tradition
exists within the commercial or military software development
communities, and no widely available prototyping language exists. Also
lacking are tools for extracting information from prototypes and for
integrating prototypes with other prototypes and existing systems.  With
few exceptions, prototyping relies on conventional implementation
languages and software development tools.

Prototyping is distinguished from other software development activities
by its use of a high-level language and code composition techniques (for
aggregation, encapsulation, and interfacing) to quickly construct and
evolve computational models to answer questions about system behavior
(both functionality and performance).  Prototyping is concerned with the
extraction and interpretation of information from such computational
models.  Efficiency and completeness are usually sacrificed to achieve
these capabilities, but only to the extent that the resulting models
contain the necessary behavior and structure needed to answer the posed
questions while retaining sufficient performance to gather that
information.  Prototyping thus involves distinctly different demands
from those which have driven the progression of either programming or
specification languages and their environments.

In the absence of prototyping languages and environments, prototyping
has occurred in implementation languages and their development
environments.  The typical prototyping process is an initial build
followed by repeated cycles of instrument, test, and revise. Huge
differences exist between the ease with which computational models
(prototypes) can be constructed, instrumented, tested, and revised in
these different languages and environments. In fact, much of the
popularity of some environments (such as Lisp, Smalltalk, and various
Expert System shells) arises from its support of this iterative process,
often to a greater extent than the appropriateness of the language
itself.

\endsubSection%{Definition}
\beginsubSection{The Grand Challenges}

There are two grand challenges for prototyping. The first is to develop
the technology for maximizing the amount of insight gained from a
computational model while minimizing the effort required to build, run,
and evolve that model, and the second is to change the software
lifecycle and acquisition processes to utilize that technology to manage
and reduce risk for both the client and the developer.

While the first challenge is open-ended with no well defined endpoint,
it is quite clear that there is enormous room for improvement.  Programs
of all forms (implementations, prototypes, and specifications) are
difficult to construct, hard to understand, and treacherous to modify.
Prototyping often involves interfacing to existing systems.
Constructing such interfaces is currently one of the most idiosyncratic
programming activities.  Gaining insight is even more problematical.
While many low level instrumentation packages exist, it is very
difficult to combine and abstract them, or to interpret them in
application terms.  Often the effort required, exceeds that of
constructing the program itself.

Once effective prototypes can be cheaply produced, the software
lifecycle and acquisition processes should change to enable all parties
to use these prototypes to reduce risk by obtaining information earlier,
without the expense and delay of full implementation. Opportunities
include determining real user needs and the feasibility of satisfying
them preaward; comparing design alternatives during development; and
determining the impact of proposed enhancements during maintenance. Risk
can be further reduced by developing, fielding, and maintaining systems
through a larger number of small evolutionary steps; by allowing
(nonperformance critical) prototype components to remain in fielded
systems; and by utilizing previously developed components.

\endsubSection%{The Grand Challenges}

\beginsubSection{Missing Science and Technology}

{\bit Prototyping Language:} We have no examples of languages designed for
prototyping. Rather, we have many implementation languages {\it used}
for prototyping. The key additional concerns raised by prototyping
are discussed in the next section. Since these concerns have not been
addressed together, it is unclear what interactions and interference
will arise.

{\bit Evolution Technology:} We have no tools other than editors to
evolve programs, i.e. change their functionality. The editors we use are
either text- or structure-based. The latter ensure syntactic
correctness. But nothing ensures that the revised semantics are correct
with respect to some (revised) criteria; nothing provides
a mechanism to directly assist the user in achieving his true objective;
nothing uses any information (such as design plans or multi-level
models) beyond what can be deduced from the program itself to help effect
the set of modifications required to achieve some evolution objective.

{\bit Program Visualization:} Our ability to determine what, how, and
how well a program is doing at appropriate levels of abstraction and in
application-specific terms is severely limited. Conceptually, the issue
is one of presentation (including focus) rather than the gathering of the
information needed for the presentation(s). This latter instrumentation
capability should exist within any well-designed prototyping language.

{\bit Guided Optimization:} Once prototypes have been constructed, they
must be sufficiently optimized so that they can be used (tested). Much
work has occurred in fully automatic optimization of high-level
languages, and this should be directly applicable to prototyping
languages. Nevertheless, many prototypes will require optimizations
beyond the state of the art of full automation. Rather than requiring
the prototype to be expressed in lower level terms (as is the norm in
compiled implementation languages), we need to allow the user to provide
additional information to guide the optimization. This may take the form
of providing data frequency, size, or distribution information,
selecting an implementation, applying a transformation, or defining a
metaprogram of optimizations. Guided optimization could broaden the
domain of prototypes that can be sufficiently optimized, as well as
provide better optimizations for prototypes within that domain.

{\bit Component Technology:} We have no engineering technology base upon
which to construct prototyping systems. Although most of the language
and environment capabilities need for prototyping have been built, some
many times, they are not accessible. Rather, they exist within
monolithic vertically integrated systems. Without any well defined interfaces,
extracting subportions is a daunting task. Furthermore, in the absence
of those interfaces, most capabilities have received neither the variety
of use nor the conceptualization needed to identify the appropriate
abstractions for expressing those interfaces.

\endsubSection%{Missing Science and Technology}
\beginsubSection{Potential Breakthroughs in 3--10 year horizon}

{\bit Common Prototyping System (\PS\
) [4 years]:} Build a single widely
available and widely applicable prototyping language using existing
prototyping technology that supports the Ada
community and facilitates the incorporation of components written
in other languages, particularly Common Lisp.

{\bit Common Prototyping Technology [8 years]:} Build a system
incorporating advanced evolution, guided optimization, and program
visualization capabilities to support prototyping in any of a broad
class of definable prototyping languages.

The research focus of the first effort is language centered: defining a
prototyping language; blending multiple computation models; spanning
real time, concurrent, distributed, and knowledge-based applications;
incorporating suitable abstraction mechanisms; and supporting delayed
and/or defaulted commitments. It relies upon existing technology for
testing, monitoring, browsing, debugging, and modifying prototypes and
for interfacing them with each other and with foreign language systems,
but requires their integration into a well-engineered, modularized,
open-architecture framework so that many contractors can contribute to
the creation of a scalable software infrastructure.

The second effort builds upon the first by focusing on prototyping
technology itself -- advancing the state of the art in gathering
information from the prototype, altering it on the basis of that
information, and improving its performance (to enable more extensive
testing) -- and in making that technology applicable to a wide range
of definable prototyping languages.

With respect to the first effort, many issues have been considered, some 
resulting in conclusions. A few of the more important issues (both resolved
and unresolved) are:

\beginlist
\item{\bull} {\bit Blended Computation Models:} No single computation model
is satisfactory. Only by employing an imperative computation model can 
structural prototypes for the intended target language, Ada,
be constructed. But imperative models sacrifice the simplicity of reasoning
and modification found in declarative models. An imperative model with
constraints on the way side-effects are used (such as limiting their scope)
which incorporates declarative capabilities (e.g. for deriving information and
maintaining consistency) might be appropriate.

\item{\bull} {\bit Delayed Binding}: Arbitrarily late binding of any 
prototyping decision must be supported. Simultaneously, early binding must
also be permitted. Furthermore, such early bindings must be exploited. 

\item{\bull} {\bit Tightly Coupled with Target Environment:} To allow 
prototypes to incorporate (sub)systems written in the target language,
and vise-versa, Call-In and Call-Out must be supported, including the
passing of all types of data and the handling of any exceptions raised
across this boundary.

\item{\bull} {\bit Environment Accessibility and Extensibility:} The
environment itself (including the ability to instrument all control
transitions and all data accesses and modifications) must be written in
the prototyping language. This ensures an open architecture for the
environment by exposing and documenting the interfaces to the
computation model, so they can be manipulated by the language. It also
ensures that the language is sufficiently powerful to represent these
interfaces and to write tools of the complexity found in environments.
It increases the comprehensibility and adaptability of the environment
because users are necessarily familiar with the prototyping language,
and it increases the extensibility of the environment because these
users have access to all the same mechanisms needed to build the
environment.  This is especially important for the application specific
portions of the environment such as instrumentation, test harnesses, and
scenario generation.  It is not by accident that all the best modern
environments (Lisp, Smalltalk, MESA, and Rationale-Ada) share this
property.  These environments are more coherent, uniform, powerful, and
integrated than ones built in a separate implementation language.

\item{\bull} {\bit Single Abstract and Surface Syntaxes}: Modularity and
coherence of the environment necessitates a single abstract syntax. A
single surface syntax is predicated by technology insertion (training)
and the desire to facilitate interaction and cooperation within the
Ada community. The difficulty is that users and tool builders have widely
different and strongly held preferences for structured versus flat surface
syntaxes.

\item{\bull} {\bit Wide-spectrum}: A wide range of styles in specifying
computations, which might loosely be characterized as ranging from very
high level to much lower level, must be allowed. This range arises from
the need to both rapidly construct prototypes and to construct detailed
structural prototypes.

\endlist
\endsubSection%{Potential Breakthroughs in 3--10 year horizon}
\endSection%{Scope}
\beginSection{Background}
\beginsubSection{Assessment of Field}

Programming is an expensive, labor-intensive activity subject to error,
and it has been estimated that the cost of correcting an error or bad
decision increases by a factor of ~10 for each phase of the waterfall
lifecycle through which it passes undetected (thus errors found during
requirements analysis cost one-hundredth those that are discovered during
coding, and one-thousandth those uncovered during testing). Many errors
result from a lack of information about the nature of and requirements
for the final program.

Thus, prototyping is a tool for obtaining such information.
Prototypes are routinely used throughout the physical sciences, where
they are called mockups, breadboards, or simulations.  Given a proposed
solution to a problem, prototyping is used to answer three types of
types of question: is this a method for achieving the solution, does
the proposed implementation have acceptable performance, production
cost, and reliability, and is it a good solution?

The need to answer these questions during software development is just
as great as in the physical sciences, particularly the last
question---whether the problem has been properly posed---because software
systems are bound neither by physical laws nor by manufacturing-imposed
similarity to previous systems.  As software systems become increasingly
complex the need for such information during all pre-implementation
phases becomes even more critical.

Despite its apparent importance, no strong prototyping tradition exists
within either the commercial or military software development
communities.  In fact, its importance contrasts strongly with the
existing fragmented support for prototyping.  No widely available
prototyping language exists.  Also lacking are tools for extracting
information from prototypes and for integrating prototypes with other
prototypes and existing systems.  Finally, no tools exist for directly
re-using existing (either prototype or target language) components or
for encapsulating them in wrappers to adapt them to particular re-use
situations.  With few exceptions, prototyping relies on conventional
implementation languages and software development tools.  This limits
both the use of and benefits from prototyping.

In fact, it is fair to conclude that {\bit Prototyping does not currently
exist as a distinct field.}

When prototypes are developed, common practice is to regard them as
throw-aways.  This practice precludes many additional benefits derivable
from prototyping during the iterative processes that occur during the
definition of a system and during its enhancement after delivery.

As a model, a prototype can continue to answer questions during these
iterative processes, but only if it tracks those iterations.  This
need to track strongly discourages throw-away prototypes.  Prototypes
must be well-structured and well-constructed, but written in a style,
language, and environment facilitating change, rather than in styles,
languages, and environments which were designed to facilitate runtime
performance.

\endsubSection%{Assessment of Field}
\beginsubSection{Relationship to Other Fields}

Almost all the experience we have is with {\it programming} languages.
That is, languages used to produce operational production programs.
Over the years, these languages have become higher level by
incorporating abstractions whose implementations are determined by the
language's compiler, thus freeing the programmer from this level of
concern.  This progression will surely continue.

Many programming languages have been used for prototyping.  In fact,
this is where most of our experience with prototyping comes from
because few languages, with the possible exception of simulation and
artificial intelligence languages, have been specifically designed for
prototyping.  Most prototyping experience has been pragmatic and
opportunistic, making use of existing technology because no
alternative existed, and it has demonstrated that prototyping can be done
within a programming language.

But this experience also clearly demonstrates the limitations and
inherent conflicts arising from such use of a programming language for
prototyping.  Though most programming languages are formally
equivalent in the class of computations they can express, languages
are distinguished by the convenience with which those computations can
be expressed. 

We believe that the expressiveness of \PL\ is very important, because
that expressiveness along with the ability to compose programs is what
enables prototypes to be created quickly and easily.  But we believe
that in prototyping, the convenience of obtaining information from a
prototype is as important as the expressiveness of \PL.

We do not believe that \PL\ should represent an advance in the
progression of programming languages.  Rather we believe it
should be an attempt to initiate a parallel but separate progression
of prototyping languages.

Currently, there is much activity in the research community to develop
specification languages and integrate them into the software
lifecycle.  This is a young but rapidly emerging field, developing, in
our view, its own distinct progression of languages. These languages
differ markedly from programming languages: Rather than
addressing efficiency concerns, they have abandoned efficiency
entirely in an effort to describe the desired effect of a computation.

\PL\ is  not an attempt to advance the Specification Language
progression.  An important purpose of prototyping is answering
questions concerning the {\it dynamics} of the system. Those questions
can relate to the behavior and performance of the system, but each
concerns its dynamics.  To answer those questions the prototype must
be executed, and its execution must be sufficiently rapid to support
gathering data. Prototyping thus involves executing models, not just
describing them, and is therefore distinct from specification efforts.

Driven by the needs of prototyping during specification and by the
absence of available prototyping languages, executable specification
languages were developed.  Though much more recent than the attempts
to incorporate prototyping in programming languages, the results seem
to be quite similar.  Experience indicates that prototyping can be
done in executable specification languages, but many of the
prototyping concerns are inadequately addressed by the predominately
descriptive nature of specification languages.

As with programming and specification, prototyping is heavily dependent
on a supportive environment, both for its instrumentation, and for low
inertia evolution. It is less concerned with performance instrumentation
than programming is, but much more concerned with behavioral
instrumentation (determining what the prototype is doing), especially in
abstracted, application specific terms. Similarly, as insight inevitably
leads to evolution, making these evolution changes with {\bit low
interia} is crucial.

\endsubSection%{Relationship to Other Fields}
\beginsubSection{Centers of Excellence}

Since Prototyping does not exist as a distinct field, it is not surprising
to find no centers of excellence for it. Rather, we find centers of
excellence for the important aspects of prototyping we've identified.
These are listed below in alphabetic order (by system) within each
category. For each system, we've tried to indicate the particular aspect(s)
of interest to Prototyping.

Almost all the systems identified below suffer from the same
common ailment: while they contain one or more interesting ideas, concepts,
or capabilities (the reason for including them), these interesting aspects
are embedded in a larger idiosyncratic framework from which they can not
be easily extracted. They thus do not form composable components from
which new systems can be constructed. Much of the effort of \PS\ will
be devoted to resolving this endemic ailment.

\beginlist

\item{\bull} For languages, abstractions, and abstraction mechanisms:
\itemitem{--} AP5 (ISI): computational relations, data activated integration
\itemitem{--} CIPL (Munich): transformational semantics
\itemitem{--} Common Objects: abstract interface for objects
\itemitem{--} EL (Harvard): extensible typed infix "Lisp"
\itemitem{--} LCS (Edinburgh Univ): concurrency abstractions
\itemitem{--} ML (Edinburgh Univ): higher order abstraction, polymorphism
\itemitem{--} Paisley (Bell Labs): formal performance requirements
\itemitem{--} Refine (Kestrel): maps, optimization
\itemitem{--} Setl (NYU): sets, finite differencing
\item{\bull} For instrumentation:
\itemitem{--} Features (CMU): view definition
\itemitem{--} Garden (Brown Univ): multiple simultaneous views
\itemitem{--} Smalltalk (Xerox Parc): tool kit of probes
\itemitem{--} Snobol-4 (Bell Labs): language based primitive instrumentation hooks
\itemitem{--} TSL (Stanford): Concurrent process instrumentation
\item{\bull} For flexible programming environment:
\itemitem{--} Lisp Machine (Symbolics): comprehensiveness, accessible computation model, low inertia modification cycle
\itemitem{--} Smalltalk (Xerox Parc): coherency

\endlist
\endsubSection%{Centers of Excellence}
\endSection%{Background}
\beginSection{Research Opportunities}
\beginsubSection{Opportunities and Possible Breakthroughs}

{\bit Evolution Technology:} Construct an interactive tool which helps
effect evolution changes (i.e. functionality changes) by converting
user-stated goals---expressed at a suitable level of abstraction within
a multi-level design---into methods for achieving them at that and all
successively lower levels.  The system would successively re-express
these changes in lower levels of the model by performing known mappings
between the levels.  Further interaction with the user would be required
if either the lower level or the mapping employed requires additional
information, or if no set of mappings could be found. This
transformational approach to evolution would express modifications as
{\it semantic altering} transformations and allow them to be invoked by
the effects they produce. Much of the effort would be devoted to
determining a basis set of such evolution transformations from which further
transformations could be composed, the design information upon which
they depended, and how such design information could be gathered,
represented, and used to effect the desired semantic alterations.

{\bit Program Visualization:} Automatically animate a prototype
expressed as a multi-level design using that multi-level structure and
its dynamic behavior as the basis for organizing the animation as an
explanation of the behavior. Much of this effort would be devoted to
determining the space of presentations, methods of animating them,
useful points in this space, {\it what} they showed, and when they were
useful.  Techniques for instantiating them with the particular pieces of
the prototype to be visualized and for (automatically) gathering the
information needed to drive the animation need to be developed so that
once selected (either manually or automatically), they could be used.

{\bit Guided Optimization:} Extend automatic optimization systems to
follow a manually prepared declarative optimization plan and fill in the
remainder automatically. The declarative optimization plan should be
relatively immune to program evolution so that it can be re-used on
successive versions. This builds upon existing work in transformations,
metaprograms, and program annotations, and couples it with automatic
optimization through a partially instantiated declarative plan.

\endsubSection%{Opportunities and Possible Breakthroughs}
\beginsubSection{Measures of Progress}

Since the success of \PS\ is predicated upon its real use and since it
is a relatively short term project, one measure of its progress ought to
be such real use. For the Ada community, the criteria for acceptance is
expected to be the prototyping functionality, the engineering quality of
the system, its documentation, performance, and low initial training
requirements.

Actual usage should also be the basis for determining whether \PS\
facilitates the development and transition of Knowledge-Based
applications into military systems.

However, these measures just show subjective acceptance of \PS\
, they
don't quantify the benefits arising from such usage. Normally, this
can only be determined by controlled experiments with parallel teams
performing the same task with and without the technology. Such experiments
are difficult to design, administer, and interpret, and expensive to run.
However, because prototyping is a risk reduction technique for uncovering
and utilizing design altering information, its benefits 
can be directly estimated with existing techniques without any duplicative
effort. The actual cost of development including the cost of prototyping
can be compared with the estimated cost (using commercial software development
cost models) of implementing the original system (unaltered by the insights
arising from the prototyping) and {\it then} making the changes found during
prototyping. The difference represents the (estimated) quantitative benefit
for prototyping that particular system. Averaging these figures across
many systems would provide an overall measure of the benefit gained.

In addition to these usage measures of progress, technical measures
could be identified for the various \PS\ capabilities, such as the
degree of coupling between prototypes and Ada, or the degree to which
all control transitions and all data accesses and modifications are
accessible for instrumentation.

\endsubSection%{Measures of Progress}
\endSection%{Research Opportunities}
\beginSection{Impact}
\beginsubSection{Potential Impact}

The main potential impact is the development of increasingly appropriate
software systems with reduced cost and delay. Almost all software
disasters were failures of either appropriateness or performance. Both
are directly addressable by risk reduction techniques during prototyping.
Furthermore, the much larger set of delays, as opposed to outright failure,
caused by appropriateness or performance problems, is similarly addressable.
Reducing such delays and failures would have an immense impact, for it is
not just the cost of the software involved (which by itself is 7 percent
of the DOD budget), but also of the hardware system it prevents from being
used. In addition, there is the even wider set of non-crucial appropriateness
and performance problems which don't prevent a software system from being
fielded, but do limit the utility derived from it. Such problems are now
corrected during maintenance, but many could be identified during prototyping
with fixes incorporated in the initial release. This would both eliminate
the reduced utility of the fielded system and reduce the cost of arriving
at a fully functional solution.

Much of the design of the prototyping language is predicated on widening
its scope so that a broader set of military applications can be addressed.
This is why concurrent, distributed, and real-time capabilities were required.

For similar reasons, knowledge-based capabilities were also required. But this
case is special because, whereas all the others have a long history within
military systems and are explicitly supported by Ada, knowledge-based
capabilities are much more recent, are not supported by Ada, and are
being developed as stand alone applications in Lisp, Prolog, or Expert
System shells. For the others, \PS\ just provides prototyping capabilities.
For knowledge based applications, it also provides the means for coupling
them with existing components both during prototype development and as
fielded (sub)systems. Thus, \PS\ provides a method for developing and
embedding knowledge based applications in military systems.

\PS\ is also an attempt to create a software infrastructure of well
engineered, carefully modularized, open-architecture components so that
the system can accommodate future research and commercial improvements
through additive growth, be used as a platform for multiple purposes,
and be user tailorable. Unix is a good example of such an extensible
open architecture system. Success in this endeavor is as important for
the software research community as \PS\ itself is to the military.

\endsubSection%{Potential Impact}
\beginsubSection{Transition to the Real World}

Real use is an integral part of the \PS\ plan. Under Darpa sponsorship,
military contractors will experimentally use \PS\ to prototype systems
they are actually developing.

Besides creating a {\it product} for such real use, \PS\ also is creating
a {\it software infrastructure} as the basis for future technology transfer.

\endsubSection%{Transition to the Real World}
\endSection%{Impact}
\beginSection{Conclusions and Recommendations}

Prototyping is an area ripe for a technology push. It has a high
potential impact on a wide range of military software by providing the
means for managing and reducing the risk associated with the appropriateness,
feasibility, and performance of the target system. 

Much of the technology is in hand, but is available only in various
separate vertically integrated monolithic systems. It needs to be
extracted, modularized into a set of well engineered components, and
made available in a widely accessible system. An open architecture 
software infrastructure must be created to accommodate the research
undertaken as it becomes available.

Two efforts are recommended. The first, Common Prototyping System (\PS\
),
is a four year effort to produce a single widely accessible, widely applicable
prototyping language using existing prototyping technology for use by
the Ada community for military applications.
Its research will be mainly language centered focused on defining a 
prototyping language, blending computation models, managing delayed bindings,
and incorporating capabilities for concurrent, distributed, real-time, and
knowledge-based applications. It will create a set of modular well-engineered
components incorporating existing prototyping technology and supporting
the defined Common Prototyping Language. These components will form
an open architecture software infrastructure for subsequent technology
transfers.

The second effort, Common Prototyping Technology, is a longer range (8 year)
effort to create advanced prototyping technology for use within the \PS\
infrastructure. Its research will focus on automated tools for evolution,
program visualization, and guided optimization.
\endSection%{This Report}

%\endChapter
\bye





∂05-Sep-89  0802	chandler@Polya.Stanford.EDU 	Pat Simmons.... 
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Sep 89  08:02:35 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA12750; Tue, 5 Sep 89 08:02:57 -0700
Date: Tue, 5 Sep 1989 8:02:57 PDT
From: "Joyce R. Chandler" <chandler@Polya.Stanford.EDU>
To: clt@sail, jmc@sail, sloan@Polya.Stanford.EDU, davis@Polya.Stanford.EDU
Subject: Pat Simmons....
Message-Id: <CMM.0.88.621010977.chandler@Polya.Stanford.EDU>

just called ... she is having car trouble and may not be in today.  She said
she would call me back in about an hour and let me know.

∂05-Sep-89  1223	peters@russell.Stanford.EDU 	[MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: User unknown] 
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Sep 89  12:22:56 PDT
Received: from localhost by russell.Stanford.EDU (4.0/inc-1.0)
	id AA11892; Tue, 5 Sep 89 12:24:13 PDT
Message-Id: <8909051924.AA11892@russell.Stanford.EDU>
To: jmc@sail.stanford.edu, tw@sail.stanford.edu
Subject: [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: User unknown]
Date: Tue, 05 Sep 89 12:24:09 PDT
From: peters@russell.Stanford.EDU


------- Forwarded Message

Return-Path: MAILER-DAEMON
Return-Path: <MAILER-DAEMON>
Received: by russell.Stanford.EDU (4.0/inc-1.0)
	id AA11779; Tue, 5 Sep 89 12:17:40 PDT
Date: Tue, 5 Sep 89 12:17:40 PDT
>From: MAILER-DAEMON (Mail Delivery Subsystem)
Full-Name: Mail Delivery Subsystem
Subject: Returned mail: User unknown
Message-Id: <8909051917.AA11779@russell.Stanford.EDU>
To: <peters>

   ----- Transcript of session follows -----
Connected to polya.Stanford.EDU:
>>> RCPT To:<tw@cs.stanford.edu>
<<< 550 <tw@cs.stanford.edu>... User unknown
550 tw@cs... User unknown
>>> RCPT To:<jmc@cs.stanford.edu>
<<< 550 <jmc@cs.stanford.edu>... User unknown
550 jmc@cs... User unknown

   ----- Unsent message follows -----
Return-Path: <peters>
Received: from localhost by russell.Stanford.EDU (4.0/inc-1.0)
	id AA11776; Tue, 5 Sep 89 12:17:40 PDT
Message-Id: <8909051917.AA11776@russell.Stanford.EDU>
To: jmc@cs, nilsson@cs, shoham@cs, tw@cs
Subject: Natural Language Algorithms
Date: Tue, 05 Sep 89 12:17:38 PDT
>From: peters

John, Nils, Yoav and Terry,

A group of us at CSLI comprising several faculty members and graduate
students from Linguistics and several visitors are embarking on a
project this year to develop some new algorithms for parsing,
interpreting and translating natural languages and also (taking
advantage of a number of visitors from Japan) a reasonably sized
grammar of Japanese.  Do any of you have, or know of, Computer Science
graduate students who would be interested in taking part in this
project?  We can offer plenty of intellectual stimulation and
opportunity to contribute and to learn -- but, unfortunately, no
salary.  We'd welcome more participation from CS.

Stanley

------- End of Forwarded Message

∂05-Sep-89  1257	chandler@Polya.Stanford.EDU 	Pat Simmons
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Sep 89  12:57:23 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA26773; Tue, 5 Sep 89 12:57:40 -0700
Date: Tue, 5 Sep 1989 12:57:38 PDT
From: "Joyce R. Chandler" <chandler@Polya.Stanford.EDU>
To: clt@sail, jmc@sail
Cc: sloan@Polya.Stanford.EDU, davis@Polya.Stanford.EDU
Subject: Pat Simmons
Message-Id: <CMM.0.88.621028658.chandler@Polya.Stanford.EDU>

I just talked to Pat.  She will not be in today.  She will be in tomorrow but
because she probably won't have a car she is not sure yet what time tomorrow
she will be in.  She said she'd call me and let me know.  When I hear I'll
let you-all know.

∂05-Sep-89  1313	clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov 	Re:  Have you read 
Received: from zaphod.es.llnl.gov by SAIL.Stanford.EDU with TCP; 5 Sep 89  13:13:29 PDT
Received: by zaphod.es.llnl.gov (4.0/1.26)
	id AA09152; Tue, 5 Sep 89 13:14:16 PDT
Date: Tue, 5 Sep 89 13:14:16 PDT
From: clm%zaphod.es.llnl.gov@lll-lcc.llnl.gov (Cynthia Mason  422-8911)
Message-Id: <8909052014.AA09152@zaphod.es.llnl.gov>
To: JMC@sail.stanford.edu
Subject: Re:  Have you read

Nope.  What's it about????  I've heard of "Seth"  but I'm not sure
its the same one.  

I hope your labor day weekend wasn't TOO laborious.  I did a lot of
catch-up reading, and have some good ideas on formulating some sections
of the dissertation (crispness is next to godliness).  You know you've
had too many years as a grad student when you read conference proceedings
before going to bed.  YUCK!  I'd much rather be reading The Choking Doberman!

Cheers,

Al T. Tude

∂06-Sep-89  1207	MPS 	My car    
I have to leave at 4:15 in order to get my car out
of hock before they close the shop for the day.

Pat

∂06-Sep-89  1814	RWF 	sail 
What is the current status of SAIL?  Will I be able to continue
using it?

∂06-Sep-89  1840	GOODMAN%uamis@rvax.ccit.arizona.edu 	Electronic News article
Received: from Forsythe.Stanford.EDU by SAIL.Stanford.EDU with TCP; 6 Sep 89  18:39:52 PDT
Received: by Forsythe.Stanford.EDU; Wed,  6 Sep 89 18:38:23 PDT
Date: Wed, 6 Sep 89 18:16 MST
From: GOODMAN%uamis@rvax.ccit.arizona.edu
Subject: Electronic News article
To: DUANE.ADAMS@C.CS.CMU.EDU, MBLUMENT@NAS.BITNET, DONGARRA@ANL-MCS.ARPA,
 GANNON%RDVAX.DEC@DECWRL.DEC.COM, JAHIR@ATHENA.MIT.EDU, HEARN@RAND-UNIX.ARPA,
 JLH@SIERRA.STANFORD.EDU, JMC@SAIL.STANFORD.EDU, MCHENRY@GUVAX.BITNET,
 CWEISSMAN@DOCKMASTER.ARPA, TROYWIL@IBM.COM, RALSTON@MCC.COM
X-VMS-To: @NAS

The Electronic News had a long artcile (July 24, pp. 1 and 42) on
the new Dept of Commerce decontrols of ATs etc. To quote: "The
liberalization move was said to stem from a report of the National
Academy of Sciences.... [2 paragraphs on us]". We may have
actually had some influence on a major decision directly involving
2 cabinet members.

∂06-Sep-89  1844	ARK  
To:   JMC, CLT    
 ∂06-Sep-89  1825	JMC  
To:   CLT, ARK    
 ∂06-Sep-89  1814	RWF 	sail 
What is the current status of SAIL?  Will I be able to continue
using it?

Replying-To: RWF
Reply-Subject: re: sail

Reply-Text:

[In reply to message rcvd 06-Sep-89 18:14-PT.]

It is planned to continue till February.  It is proposed that each
user pay the same percentage share of the reduced costs as he
was paying over the last n months.  Carolyn (clt) and Arthur Keller (ark)
know  n.  I believe the Department was paying your costs and I hope
it will continue to pay your share.

ARK - The base period is Jan-Mar 89.  It is adjusted for people whose
usage changed drastically, upon request.

Arthur

∂06-Sep-89  2127	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	simplification challenge
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 6 Sep 89  21:27:52 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 417683; 7 Sep 89 00:19:36 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107863; Wed 6-Sep-89 21:20:05 PDT
Date: Wed, 6 Sep 89 21:18 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: simplification challenge
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, macsyma-i@RUSSIAN.SPA.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "rwf@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890904112000.4.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890907041857.2.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Mon, 4 Sep 89 04:20 PDT
    From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

      /
     |    1
     | ------- dx = 
     |      24
    /  1 - x

                                   2
                                2 x  - (sqrt(6) + sqrt(2)) x + 2           (sqrt(6) + sqrt(2)) x
       (sqrt(6) + sqrt(2)) (log(--------------------------------) + 2 atan(---------------------))
                                    2                                                2
                                2 x  + (sqrt(6) + sqrt(2)) x + 2                 2 (x  - 1)
     - ---------------------------------------------------------------------------------------------
                                                    96

     + . . .

Actually, I'd already done the hard part of pairing the conjugate roots in the
partial fraction expansion, and then denesting the roots of cos(pi/12).  All
that remained was symmetrizing, and removal of an imaginary additive constant
due to inconsistent branch choices.

        /
        [    1
(D2236) I ------- dx = 
        ]      24
        / 1 - x

                           (sqrt(6) + sqrt(2)) x         (sqrt(6) + sqrt(2)) x
(sqrt(6) + sqrt(2)) (atanh(---------------------) + atan(---------------------))
                                        2                             2
                                2 (1 + x )                    2 (1 - x )
--------------------------------------------------------------------------------
                                       48

                              (sqrt(6) - sqrt(2)) x         (sqrt(6) - sqrt(2)) x
   (sqrt(6) - sqrt(2)) (atanh(---------------------) + atan(---------------------))
                                           2                             2
                                   2 (1 + x )                    2 (1 - x )
 + --------------------------------------------------------------------------------
                                          48

         sqrt(3) x         sqrt(3) x          sqrt(2) x         sqrt(2) x
   atanh(---------) + atan(---------)   atanh(---------) + atan(---------)
               2                 2                  2                 2
          1 + x             1 - x              1 + x             1 - x
 + ---------------------------------- + ----------------------------------
               8 sqrt(3)                            12 sqrt(2)

           x              x
   atanh(------) + atan(------)
              2              2
         1 + x          1 - x     atanh(x) + atan(x)
 + ---------------------------- + ------------------
                24                        12

(c2237) RATSIMP(DIFF(%,X));

Time= 3602 msecs
                                     1           1
(D2237)                         - ------- = - -------
                                   24          24
                                  X   - 1     X   - 1

∂07-Sep-89  1029	rpg@lucid.com 	Lunch
Received: from lucid.com by SAIL.Stanford.EDU with TCP; 7 Sep 89  10:29:50 PDT
Received: from challenger ([192.9.200.17]) by heavens-gate id AA08667g; Thu, 7 Sep 89 10:29:45 PDT
Received: by challenger id AA04446g; Thu, 7 Sep 89 10:26:47 PDT
Date: Thu, 7 Sep 89 10:26:47 PDT
From: Richard P. Gabriel <rpg@lucid.com>
Message-Id: <8909071726.AA04446@challenger>
To: jmc@sail.stanford.edu
Subject: Lunch

I'm feeling a little under the weather today and would rather
sleep than eat. Another time, perhaps.

			-rpg-

∂07-Sep-89  1039	VAL  
Lunch tomorrow at MacArthur Park?

∂07-Sep-89  1055	MPS 	Russian Visitor
 ∂07-Sep-89  1034	gilberts@Polya.Stanford.EDU 	Russian Visitor 
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Sep 89  10:34:42 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA15985; Thu, 7 Sep 89 10:35:07 -0700
Date: Thu, 7 Sep 1989 10:35:06 PDT
From: Edie Gilbertson <gilberts@Polya.Stanford.EDU>
To: MPS@Sail, CLT@Sail
Cc: Gilbertson@Polya.Stanford.EDU
Subject: Russian Visitor
Message-Id: <CMM.0.88.621192906.gilberts@Polya.Stanford.EDU>

CSLI just called about getting a long dist auth code for a Russian
visiting John McCarthy.  Do you know what account number should be
used for the phone calls, and do you have the name of the visitor
(who has an office at CSLI?)   Thanks.  -Edie

Edie,

I will have to ask him when he comes in.  This is news to me.

Pat

∂07-Sep-89  1111	VAL 	re: reply to message
[In reply to message rcvd 07-Sep-89 11:09-PT.]

1130 is fine.

∂07-Sep-89  1107	weening@Gang-of-Four.Stanford.EDU 	Qlisp
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Sep 89  11:07:19 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.0/25-eef) id AA22962; Thu, 7 Sep 89 11:07:19 PDT
Date: Thu, 7 Sep 89 11:07:19 PDT
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8909071807.AA22962@Gang-of-Four.Stanford.EDU>
To: jmc@sail, clt@sail
Subject: Qlisp

Can we meet sometime to talk about Qlisp for the upcoming year?  Also,
I talked to Tom D. about future equipment, and he is checking into the
possibility of DEC gift funds being available to buy a Unix system for
Formal Reasoning.  (DEC gave a $8 million equipment gift to Stanford.)

∂07-Sep-89  1120	shoham@Polya.Stanford.EDU     
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Sep 89  11:20:54 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA27689; Thu, 7 Sep 89 11:21:18 -0700
Date: Thu, 7 Sep 1989 11:21:14 PDT
From: Yoav Shoham <shoham@Polya.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
In-Reply-To: Your message of 07 Sep 89 1119 PDT 
Message-Id: <CMM.0.88.621195674.shoham@Polya.Stanford.EDU>

Ok, thanks - I'll get from someone else.

Yoav

∂07-Sep-89  1120	gilberts@Polya.Stanford.EDU 	re: Russian Visitor  
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Sep 89  11:20:53 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA27654; Thu, 7 Sep 89 11:21:10 -0700
Date: Thu, 7 Sep 1989 11:21:08 PDT
From: Edie Gilbertson <gilberts@Polya.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Cc: MPS@Sail, Gilberts@Polya, CLT@Sail
Subject: re: Russian Visitor 
In-Reply-To: Your message of 07 Sep 89 1107 PDT 
Message-Id: <CMM.0.88.621195668.gilberts@Polya.Stanford.EDU>

I've just spoken to CSLI again, and the name of the visitor is Andre Scedrov,
from the Univ of Penn.  Do you know anything about him?  Should he get a
long dist auth code under one of your accounts?  -Edie

∂07-Sep-89  1128	VAL 	Index

Ability, 25, 29-33, 46, 52, 82, 88, 244
	law of, 3, 6, 39, 42-43, 210
Abnormality, 198, 201-204, 207-215, 232-233, 242-243
Accessibility relation, 54-58, 139, 158-161, 164
Action 3, 9, 14-15, 39, 58-59, 65, 79-82, 112, 191-192, 230-231
	deliberate 98-99
	free 113
Advice taker, 9-15, 238
Airport problem, 1-3, 13-20, 
Alternativeness relation, see Accessibility relation
Anthropomorphism, 179-180, 183
Approximate theory, 90, 97-100,
Artificial intelligence problem, 64-65, 246
Artificial intelligence research, 187-188
	deficiences of, 68-69, 187-188
	progress in, 22, 69, 180
Aspect
ATM, see Automatic teller
Autoepistemic logic
Automatic teller
Beauty, appreciation of
Behavioral definition, see operational definition
Behaviorism
Belief
Best object
Biology
Blocks world
Body
Brain
Causality
CBCL, see Common Business Communication Language
Cancellation of inheritance
Canning
Cellular automaton
Chess program
Chinese room
Circumscription
	applications of
	collapsible
	domain
	formula
	predicate
	prioritized
Circumscription policy
Circumscriptive inference
Closed world assumption
Cohistorical situations
Combinatorial explosion
Common Business Communication Language
Common sense
Commonsense inertia
Commonsense knowledge
Commonsense physics
Commonsense reasoning
Communication
Communication convention
Computational complexity
Concept
Conciousness
Concurrency
Conjecture
Context
Continuous time
Control
Counterfactual
Creativity
DARPA
Declarative sentence
Deduction
Default
Denotation
Deontic logic
Design stance
Desire
Determinism
Duties of machines
Education by brain surgery
EKL
Elaboration tolerance
Emotion
Empiricism
EMYCIN
Epistemic logic
Epistemic alternative
Epistemological adequacy
Event
Event calculus
Exception
Existence
Expert system
Expert system shell
Extensionality
Fact
Finite automaton
First order language
Fluent
	propositional
FOL
Formal literature
Frame
Frame problem
Free will
Future operator
General Problem Solver
General purpose database
GPS, see General Problem Solver
Hate
Herbrand universe
Heuristic
Heuristic adequacy
History
Horn clause
Humor
Imperative sentence
Incrementalism
Induction, mathematical
Information storage convention
Inheritance
Intellect
Intelligent backtracking
Intention
Intentional stance
Introspection
Is-a hierarchy, see Inheritance
Kempe transformation
Knowledge
	joint
Knowledge engineer
Language
	abstract
	artificial natural
	natural
Learning
Leibniz's law
Lexicographic ordering
Life automaton
LISP
Logic
Logic programming
Love
Map coloring
Martian
Meaning
Mental qualities
Mental state
Metaepistemology
Metaphilosophy, see Metaepistemology
Metaphysical adequacy
Metaphysics
MICROPLANNER
Mind
Minimal inference
Minimal models
Minimal entailment
Minimization
Missionaries and cannibals
Missoury program
Modal function
Modal logic
Modal operator, see Modal logic
Monotonicity, see Nonmonotonic reasoning
Morality
Motor parameter
MYCIN
Name
National Science Foundation
Natural selection
Necessity
Nominalism
Nonmonotonic logic
NP-completeness
NSF, see National Science Foundation
Objectivity
Observation
Occam's razor
Ontology
Operational definition
Parallel programs, see Concurrency
Parameter
Philosophy
	naive
Physical stance
Physics
Plan
PLANNER
Policy
Positivism
Possible worlds
Postponable variable
Predicate calculus, see First order logic
Prediction
Present object
Probability
	and nonmonotonic reasoning
Production system
PROLOG
Priority, see Circumscription, prioritized
Proof checker
Proposition
Psychology
	introspective
Qualification
Qualification problem
Rationalism
Rationality
Realism
Reasoning
	conjectural
	commonsense
	nondeductive
	nonmonotonic
Reasoning program
Referential opacity
Reification
Representation problem
Resolution
Rights of machines
Robot
Rule
Search
Second order language
Self
Self-conciousness
Self-knowledge
Sense
Senses
Separable formula
Sequence extrapolation
Simple abnormality theory
Simultaneous actions, see Concurrency
Situation
Situation calculus
State
Strategy
	feasible
STRIPS
Structural definition
Tense logic
Thermostat
Thought
Time
Time-sharing system
Tool
Transitive closure
Turing test
Ultranatural deduction
Uncertainty
Understanding
Unique names assumption
Unintended model
Variable predicates
Visual images
Will
Yale shooting

∂07-Sep-89  1746	VAL 	Nonmonotonic workshop    
To:   "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU   
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>


			 CALL FOR PARTICIPATION

	  THIRD INTERNATIONAL WORKSHOP ON NONMONOTONIC REASONING


The third international workshop on nonomontonic reasoning, sponsored by
AAAI and CSCSI, will be held May 31-June 3, 1990, at the Stanford Sierra
Camp in South Lake Tahoe, California.  The aim of the workshop is to
bring together active researchers in the area of nonmonotonic reasoning
to discuss current research, results, and problems of both theoretical
and practical nature.


Topics (not necessarily exhaustive):

	General theories of defeasible inference
	Comparison of formal systems
	Applications to planning, commonsense reasoning
	Knowledge update and truth maintenance
	Relation to probability models
	Theories of inheritance with exceptions
	Argument-based systems
	Proof theory, complexity, and automation
	

Attendance will be limited to 30-40 people, by invitation only.  Those
wishing to attend should submit five copies of a detailed abstracts of
current research to:

Kurt Konolige
SRI International EJ272
333 Ravenswood Avenue
Menlo Park, Ca. 94025
Phone: (415) 859-2788  
E-mail: konolige@ai.sri.com

Electronic mail submissions are encouraged (one copy only!).  Abstracts
should consist of no more than 10 double-spaced pages when printed (4000
words) and should include enough information to enable the program
committee to judge the contribution of the work.  Abstracts will be
accepted on the basis of quality, originality, and significance.  The
deadline for submission of papers is ***December 17, 1989*** (note that
this is later than the original announcement).  Notification of
acceptance will be made by February 26, 1990.  Accepted authors will be
asked to send a preprint for distribution at the workshop.  Papers
presented at the workshop can be submitted for publication in a
collection to appear later.

Program Committee:

	Johan DeKleer, Xerox Parc	(dekleer.pa@xerox.com)
	Jon Doyle, MIT			(doyle@zermatt.lcs.mit.edu)
	David Etherington, AT&T		(ether@research.att.com)
	Matt Ginsberg, Stanford 	(ginsberg@polya.stanford.edu)
	Hector Geffner, UCLA		(hector@cs.ucla.edu)
	David Israel, SRI 		(israel@ai.sri.com)
	Henry Kautz, AT&T		(kautz@research.att.com)
	Vladimir Lifschitz, Stanford	(val@sail.stanford.edu)
	David Poole, UBC		(poole@vision.cs.ubc.ca)
	Erik Sandewall, Linkoping	(enea!lisbet.liu.se!e-sandewall@uunet.uu.net)   
	Richmond Thomason, Pittsburgh	(rich.thomason@cad.cs.cmu.edu)

In addition to accepted authors, we have a limited number of slots for
students who have shown promise in the area.  Sponsors for such students
should send a short justification to Kurt Konolige at the above address.

∂08-Sep-89  0451	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	eater3   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 8 Sep 89  04:50:55 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 418315; 8 Sep 89 07:39:37 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107989; Fri 8-Sep-89 04:40:15 PDT
Date: Fri, 8 Sep 89 04:39 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: eater3
To: "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hpm@rover.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890908113909.5.RWG@TSUNAMI.SPA.Symbolics.COM>

Dean Hickerson hadn't seen Buckingham's eater3, which means
I may be the only available source.  Even more interestingly,
Buckingham's letter has been buried for years in a box somewhere
(I hope) in Weyhrauch's basement, and I couldn't find it (nor
rederive it) when I thought I needed it to make a gun from the
p100.

However, a couple of consecutive meals of broiled salmon induced
enough inverse Alzheimer's that I was able to rederive it!
Here it is, with three vanilla eaters, doing something incredibly
annoying to lumps-of-muck.

 . . . . . . . . . . . . . . . . . . . . . . . . . o o . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . o . . . .
 . . . . . . . . . . . . . . . . . . . . . . . o . o . . . .
 . . . . . . . . . . . . . . . . . . . . . . . o o . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . o o o . . . . . . . . . . . . .
 . . . . . . . . . . . . o o o . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . o o . . . . .
 . . . . . . . . . . . . . . . . . . . . . . o . . o . . o o
 . . . . . . . . . . . . . . . . . . . . . . o . o . . . . o . . o
 . . . . . . . . . . . . . . . . . . . . . . . o . . . . . o . o . o
 . . . . . . . . . . . . . . . . . . . . . . . . . . o o . o . . o
 . . . . . . . . . . . . . . . . . . . . . . . . . . o . . o
 . . . . . . . . . . . . . . . . . . . . . . . o . . . . o .
 . . . . . . . . . . . . . . . . . . . . . . . o o o o o . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . o . . . .
 . . o o . . . . . o o . . . . . . . . . . . . . o . o . . .
 . o . o . . . . . o . o . . . . . . . . . . . . . o . . . .
 . o . . . . . . . . . o . . . . . . . . . . . . . . . . . .
 o o . . . . . . . . . o o . . . . . . . . . . . . . . . . .

Eater3 presumably oscillates with period 8 if the loaf is
symmetrically pricked on its other (NW) side.  It is less
versatile than vanilla eater, but more robust, by virtue
of lying low longer before completing self repair.

(Aside:  while string-scanning my malignant mailoma for
HPM's net address, ZMail stumbled over ACW's High Precision
Mandelbrot function!)

∂08-Sep-89  0512	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	account       
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 8 Sep 89  05:12:11 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 418324; 8 Sep 89 07:53:08 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 107991; Fri 8-Sep-89 04:53:52 PDT
Date: Fri, 8 Sep 89 04:52 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: account   
To: "JMC@SAIL.Stanford.EDU"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: mlb@RUSSIAN.SPA.Symbolics.COM
In-Reply-To: <nXtz6@SAIL.Stanford.EDU>
Message-ID: <19890908115247.6.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: 05 Sep 89  1644 PDT
    From: John McCarthy <JMC@SAIL.Stanford.EDU>

    As of Sept 1, SAIL went private, i.e. is no longer part of the
    Computer Science Department facility with charges for all use.
    Therefore, if you have any use for an account, you can have one.

I'd be honored, of course.  But the only use I can think of making
might be the simplification of retrieving old files, e.g. Pox and
Life.  Or do tapes from that era crumble to dust upon touch?

Do you know if mail to RWG@SAIL still forwards?  I haven't got any
lately, but there is a persistent rumor that I returned to Stanford,
and someone may ping me there.

Also, is there any truth to the claim that Kennedy wanted to finish
deChristianizing the Chapel because he thought that IHS meant I Hate
Stanford?  Institute Had The Finest Professors!

∂08-Sep-89  1238	CLT 	hardware discussion      
To:   tom@POLYA.STANFORD.EDU, weening@Gang-of-Four.Stanford.EDU,
      JMC@SAIL.Stanford.EDU
Tuesday morning at 10.

∂08-Sep-89  1309	MLB@WHITE.SWW.Symbolics.COM 	Lisp history questions    
Received: from WHITE.SWW.Symbolics.COM ([128.81.57.24]) by SAIL.Stanford.EDU with TCP; 8 Sep 89  13:09:05 PDT
Received: from CHROME.SWW.Symbolics.COM by WHITE.SWW.Symbolics.COM via CHAOS with CHAOS-MAIL id 18110; Fri 8-Sep-89 13:06:19 PDT
Date: Fri, 8 Sep 89 13:06 PDT
From: mArQ lE bRuN <MLB@WHITE.SWW.Symbolics.COM>
Subject: Lisp history questions
To: jmc@sail.stanford.edu
cc: rwg@WHITE.SWW.Symbolics.COM
Message-ID: <19890908200610.6.MLB@CHROME.SWW.Symbolics.COM>

Hi, recent mail reminded me I've been curious lately about some small points of
Lisp history:

1. What was the real intent, role and (d)evolution of the use of M-expressions?
Were they actually motivated just by logical rigor, was it partly to avoid the
potential confusion of using only S-expressions to discuss themselves or perhaps
was it just the unfamiliar minimal syntax of S-expressions?  Did people really
"use" M-expressions, or was it a formalism intended mainly for publications and
such?  I never see M-expressions used (although I suppose logicians or somebody
might) -- what was the historical process of their abandonment?  Did everybody
just stop one day, or was it gradual, with the "die-hard M-hackers" slowly
fading away?

I'm curious about all this because there's a real value in the lack of syntax of
the S-expression forms.  I'm wondering why they aren't more pervasively used
throughout the computer sciences, and what might be learned from Lisp history
about how the idea might be assimilated more widely.  (For example, people are
to this day struggling along with arcane and/or weak macro facilities, because
they can't fraction off the syntactic component of their languages).

2. This is more trivial: what can you tell me about the historical process
whereby the variable T came to be self-evaluating (instead of evaluating to the
self-evaluating *T*), yet somewhat inconsistently, F became supplanted by the
direct use of NIL?

(3. I don't suppose I get extra credit for noticing the mispelling near the
bottom of p21 of 2nd edition of the 1.5 manual: "Boolian"?)

∂08-Sep-89  1707	dresser@csli.Stanford.EDU 	re: Math question      
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 8 Sep 89  17:07:25 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA20385; Fri, 8 Sep 89 17:08:47 PDT
Date: Fri 8 Sep 89 17:08:45-PDT
From: Charles G. Dresser <DRESSER@CSLI.Stanford.EDU>
Subject: re: Math question   
To: JMC@SAIL.Stanford.EDU
Message-Id: <621302925.0.DRESSER@CSLI.Stanford.EDU>
In-Reply-To: <1VYyTK@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

Thanks for 1/e.  I'm curious about your reference to lim (1+1/n)**n as the 
classical definition of e.  I'm familiar with the definition of e as the
solution to ln x = 1.  But not this classical definition.  What motivated
this definition?  (A quick look at Thomas confirms the definition I am
familiar with.  The other result is derived.)

Charles

(By the way, if you are busy, you don't need to bother with my
curiosity.)
-------

∂08-Sep-89  2017	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 8 Sep 89  20:17:50 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 418763; 8 Sep 89 23:13:03 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108076; Fri 8-Sep-89 20:13:47 PDT
Date: Fri, 8 Sep 89 20:12 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
To: "JMC@SAIL.Stanford.EDU"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <bYX1p@SAIL.Stanford.EDU>
Message-ID: <19890909031242.8.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: 08 Sep 89  0907 PDT
    From: John McCarthy <JMC@SAIL.Stanford.EDU>

    test

Well, that answers that.  Did Caroline show you that -pi↑2/12e↑3
sum?  I collected a few in a typesot letter to Askey, in case you
want a copy . . .

∂09-Sep-89  0814	CLT  
timothy's car seat is outside the front door

∂09-Sep-89  0953	dresser@csli.Stanford.EDU 	re: Math question      
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 9 Sep 89  09:53:06 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA28289; Sat, 9 Sep 89 09:54:12 PDT
Date: Sat 9 Sep 89 09:54:11-PDT
From: Charles G. Dresser <DRESSER@CSLI.Stanford.EDU>
Subject: re: Math question   
To: JMC@SAIL.Stanford.EDU
Message-Id: <621363251.0.DRESSER@CSLI.Stanford.EDU>
In-Reply-To: <bY$mV@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

Thanks.

Charles
-------

∂09-Sep-89  0955	mrc@SUMEX-AIM.Stanford.EDU 	re: Another Quayle shot    
Received: from hanna.cac.washington.edu by SAIL.Stanford.EDU with TCP; 9 Sep 89  09:55:10 PDT
Received: from tomobiki-cho.cac.washington.edu by hanna.cac.washington.edu
	(5.61/UW-NDC Revision: 2.4 ) id AA16257; Sat, 9 Sep 89 09:55:20 -0700
Date: Sat, 9 Sep 1989 9:53:54 PDT
From: Mark Crispin <mrc@SUMEX-AIM.Stanford.EDU>
Subject: re: Another Quayle shot
To: John McCarthy <JMC@sail.stanford.edu>
In-Reply-To: <tYa2P@SAIL.Stanford.EDU>
Message-Id: <MS-C.621363234.1103527590.mrc@SUMEX-AIM.Stanford.EDU>

I too wish it had been Bush/Bentsen vs. Dukakis/Quayle.

-------

∂11-Sep-89  0912	MPS  
I got a call from Prof. Garcia of the Polytechnic Univ of Madrid
today.  She was not clear regards the message.  Evidently, you met
a Prof. Pazos at an AI conference.  It has something to do with an
honorary member of the university.  Do you know anything about it?
She also said there had been a letter sent, but I do not remember
receiving such a letter.  Her home number is 91-256-7949.  She
would like you to call her.

Pat

∂11-Sep-89  1000	JMC  
Call Craig.

∂11-Sep-89  1105	perrault@ai.sri.com 	Richard Waldinger  
Received: from Warbucks.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 11 Sep 89  11:05:45 PDT
Received: from agate.ai.sri.com by Warbucks.AI.SRI.COM with INTERNET ;
          Mon, 11 Sep 89 11:06:27 PDT
Received: by agate.ai.sri.com (4.1/4.16)
	id AA03356 for jmc@sail.stanford.edu; Mon, 11 Sep 89 11:07:41 PDT
Date: Mon 11 Sep 89 11:07:39-PDT
From:     PERRAULT@Warbucks.AI.SRI.COM (Ray Perrault)
Subject: Richard Waldinger
To:       jmc@sail.stanford.edu
cc:       perrault@Warbucks.AI.SRI.COM
Message-Id: <621540459.0.PERRAULT@ai.sri.com>
Mail-System-Version: <SUN-MM(229)+TOPSLIB(128)@ai.sri.com>

John,

SRI awards each year two SRI Fellowships to employees who have demonstrated
technical leadership in their field.  The award includes 3 months leave
with no constraints.  I would like to nominate Richard for one of these.
Would you be prepared to write him a letter of recommendation?  
I realize you must get reams of these requests, so I'll try to make this
easier:  if you don't think you can write an enthusiastic letter in 
short time, I'll withdraw the request!  

If you agree, which I hope you will, I'll send you a few more details.
I'd appreciate the letter by Sept 28.

Many thanks,

ray
-------

∂11-Sep-89  1211	Mailer 	re: Another Quayle shot    
To:   JMC@SAIL.Stanford.EDU, LES@SAIL.Stanford.EDU,
      su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>

[In reply to message from JMC rcvd 08-Sep-89 19:23-PT.]

Quayle is a new invention, the all-purpose presidential
life insurance policy.  NOBODY IN THE WORLD wants George
Bush to die.  In future elections, presidents remembering
the fates of McKinley, Kennedy, Nixon, and the attempts
on Ford, Roosevelt, etc., will take out such policies.
The policy must be a virtual unknown, with a voting
record that can be interpreted as a sop to a major
voting bloc, whose incompetence has not yet been made
spectacularly visible.  His intelligence must be enough
for him to seem plausible to political reporters, yet
low enough for him to believe in astrology or the
existence of canals on Mars.  Considering that people
were willing to vote for Carter for president, the fact
that the policy is an unknown will cause no problem.
The solution is to amend the constitution so that 
a new election is held within four months.

∂12-Sep-89  0430	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	flakiness
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 12 Sep 89  04:29:55 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 419528; 12 Sep 89 07:19:51 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108172; Tue 12-Sep-89 04:20:16 PDT
Date: Tue, 12 Sep 89 04:19 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: flakiness
To: fun@RUSSIAN.SPA.Symbolics.COM, math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "Dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    spa-hackers@RUSSIAN.SPA.Symbolics.COM
Message-ID: <19890912111931.1.RWG@TSUNAMI.SPA.Symbolics.COM>

I just had to boot due to a soft disk ecc in a paging partition.
Why, just last March, I booted for a power supply replacement.

A few screens I wish I hardcopied:  "... 86 inboxes deleted",
Peek idle times >3000 hrs, Macsyma command numbers reaching
several thousand several times, menus of editor buffers, mail
buffers (>9MW), and sent messages stretching on for pages...,
and, of course, EGC work units, given a couple of week-long
double-floating plots.  The session also survived a dozen or so
performances of the 1986 Stanford math animation, a Richter 5.1,
medfly spraying, 6 megabit window hardcopying, rendering, 3 DGCs...

How do they expect us to get any work done if we have to boot
these damned things all the time?  Support our boys in Diablo
Canyon.

∂12-Sep-89  0537	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	another attempt    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 12 Sep 89  05:37:12 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 419540; 12 Sep 89 08:26:34 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108175; Tue 12-Sep-89 05:27:18 PDT
Date: Tue, 12 Sep 89 05:26 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: another attempt
To: "george@euler.math.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <8909120240.AA07753@euler.math.nwu.edu>
Message-ID: <19890912122631.3.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Mon, 11 Sep 89 21:40:06 CDT
    From: George Gasper <george@euler.math.nwu.edu>

    I hope that your address
    rwg@yukon.scrc.symbolics.com

    works.  

After only three years, I finally tell you my correct address!

    I received the copy of your Aug. 27 th letter to Askey, and was at first
    surpised by the trig. funcitons of square roots of quadratics in the index
    of summation.  But then I recalled that there are integrals of Bessel functions
    of square roots of quadratics in the variable of integration contained in 
    section 13.47 of Watson's Theory of Bessel Funtions book.  These are continuous
    analogues of corresponding series, but I have not seen any of the anologous
    series published anywhere.  Your sums written in terms of Bessel functions
    of orders -1/2  and 1/2 suggest that there should be more general sums 
    containing Bessel functions of arbitrary order.  Perhaps your techniques
    extend to the Bessel functions  cases.  These should provide a continuous
    connection between the sine  and cosine series.  A series analogue of Sonine's
    formula 13.47 (5) on page 417 would be of particular interest.

YOW, what an observation!  Unfortunately, I can't quite make it work.
I can punt the sqrt(t↑2+z↑2) from the denom via a linear combination of
the nu=1/2 case and its derivative wrt a.  But that leaves us the analog
of an alternating series (mine are not), and requires us to choose mu = -3/2,
in violation of R(mu)>-1, (which the integral needs for convergence).

Maybe we can stop the alternation by switching to some linear operation on
13.47 (1), with suitable half-integers and derivatives for b as well as a.

    Can your technique handel such a series?

My cosines came out of infinite products.  Now that you mention it, there
ought to be nice products for Bessels.  But where?  I guess the problem is
with the roots going hairy for the non-trig case.

					      Concerning continuous analogues of
    series, see the Boas and Pollard paper on pages 18-25 of the Amer. Math Monthly
    80(1973).
    How did you discover your surprising "bozo bibasic"?  It sure looks strange.

From Mr. Bibasic, that's a compliment!  Just like I said, plug in two geometric
sequences, and let the sums (and product) go to infinity.

    It would be interesting if some more parameters could be added!!!!

More parameters you just plug in.  The trick is to specialize to something
nonhideous.

∂12-Sep-89  1100	bh%anarres.Berkeley.EDU@Berkeley.EDU 	Re:  Might you be in a position to help ME in getting old SAIL dump
Received: from anarres.Berkeley.EDU by SAIL.Stanford.EDU with TCP; 12 Sep 89  11:00:24 PDT
Received: by anarres.Berkeley.EDU (5.57/1.25)
	id AA20527; Tue, 12 Sep 89 11:01:15 PDT
Date: Tue, 12 Sep 89 11:01:15 PDT
From: bh%anarres.Berkeley.EDU@berkeley.edu (Brian Harvey)
Message-Id: <8909121801.AA20527@anarres.Berkeley.EDU>
To: JMC@sail.stanford.edu
Subject: Re:  Might you be in a position to help ME in getting old SAIL dump

I'll discuss it with Marty and see what's involved.

(P.S.  I still think we should get the Smithsonian to pay to
keep it running forever!)

∂12-Sep-89  1315	@Score.Stanford.EDU,@ai.ai.mit.edu,@mc.lcs.mit.edu:goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk 	Research Assistantships Available  
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Sep 89  13:15:11 PDT
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU.#Internet) by SCORE.STANFORD.EDU with TCP; Tue 12 Sep 89 13:14:01-PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa26186;
          12 Sep 89 15:50 EDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 12 Sep 89 15:52:08 EDT
Received: from lcs.mit.edu (CHAOS 15044) by MC.LCS.MIT.EDU 12 Sep 89 15:48:10 EDT
Received: from NSFNET-RELAY.AC.UK by mintaka.lcs.mit.edu id aa25998;
          12 Sep 89 15:41 EDT
Received: from prg.oxford.ac.uk by NSFnet-Relay.AC.UK   via Janet with NIFTP
           id aa12622; 12 Sep 89 19:34 BST
Received: from uk.ac.oxford.prg.client80 (client80) by uk.ac.ox.prg (4.12/prgv.37)
	id AA24581; Tue, 12 Sep 89 18:09:15 bst
Received: by uk.ac.oxford.prg.client80 (3.2/prg.1)
	id AA24030; Tue, 12 Sep 89 18:14:35 BST
Date: Tue, 12 Sep 89 18:14:35 BST
From: goguen%prg.oxford.ac.uk@nsfnet-relay.ac.uk
Message-Id: <8909121714.AA24030@uk.ac.oxford.prg.client80>
To: theorem-provers%mc.lcs.mit.edu@nsfnet-relay.ac.uk
Subject: Research Assistantships Available


                           RESEARCH ASSISTANTSHIPS


Applications are invited for 2 Research Assistants to join two projects in the
Programming Research Group at the University of Oxford, under the direction of
Professor Joseph Goguen.

The aim of the first project is to build an extensible theorem prover, based
on the OBJ3 functional language, that can be programmed to support any desired
logical system, by providing rules of deduction and proof strategies for
combining them.  Applications will include hardware specification
verification.  This project is supported by the U.K. Information Engineering
Directorate, and will run in collaboration with Inmos Ltd., the Royal Signals
and Radar Establishment, and Royal Holloway and Bedford New College.

The second project is to extend the OBJ3 system into an object oriented
language called FOOPS.  It is anticipated that the most interesting problems
will involve concurrency.  Applications will include the specification and
rapid prototyping of software systems.  This project is supported by the U.K.
Science and Engineering Research Council.

These posts are tenable for 3 years at a salary scale of aproximately 11,000
to 16,000 pounds per year, depending on qualifications.  Applicants should
have a upper Honors undergraduate degree, and a doctorate or equivalent
experience in Computing Science, Computer Engineering, or Mathematics and
Computing Science.  A good background in algebra, Lisp programming, and
hardware and/or software engineering is desirable.  The positions are tenable
from October 1989, and applications should be sent as soon as possible to:

  Prof. Joseph A. Goguen
  Programming Research Group
  Oxford University Computing Laboratory
  11 Keble Road
  Oxford OX1 3QD
  United Kingdom

  Email: goguen@uk.ac.ox.prg (JANET)
         goguen@prg.ox.ac.uk (Internet)
  Tel: (0865) 272567

Applications should include a full Curriculum Vita and the names of two
referees.



∂12-Sep-89  1332	CLT 	hardware  
To:   JMC@SAIL.Stanford.EDU, weening@Gang-of-Four.Stanford.EDU
CC:   tom@Polya.Stanford.EDU 

Here is what I plan to send to DARPA (the price estimates may change
if I get new data from Tom).  Any comments?

-----------------------------------------------------------------------------
Our ancient Dec Kl10, SAIL, will be retired (to old and expensive to
maintain) soon (hopefully in February).  This together with the Alliant
provides the main computing facilities used by our group.  Given the
excessively high maintanence cost (approx 70k/yr including overhead), the
lack of available software, etc.  the Alliant is not a viable replacement
for SAIL.  Thus we need to purchase replacement hardware.

We want to continue with a multi-user time-sharing system and we have
come up with the following rough plan:

A Dec (mips) or Sun (spark) with
  64meg memory
  4gig disk  (1-controller 4-disks)
  8mm tape drive
  ethernet controller
  serial line board
The estimated cost for this package is around 87k (Sun with super discount).

In addition we will need terminals for office and home.  (Sail terminals
won't work).  Here we are thinking of the NCD terminals (about 2-2.5k each).
A minimum of 14 terminals is needed for the existing researchers
and additional terminals will be needed for secretaries and students.
This system would provided facilities for John McCarthy's group
(McCarthy, Lifshitz, Talcott, Mason, Weening, Pehoushek)
and Zohar Manna's group (Manna and Waldinger). 

The yearly cost for maintaining the new system is estimated to
be 17k for maintenance plus 50k for systems programmer, backup, etc.
This is to be contrasted with the SAIL computer charges for the two
groups which were running at approximately 17k/month (200k/year).


The Qlist project recently requested approval to purchase the 4 processors
on loan (since the loan officially terminated last April and Alliant is
pushing for a decision).  It is important to have more than 4 processors
to get reasonable experimental data.  We are investigating the possibility
of using 8 processor Alliants at Argonne or elsewhere for gether final
data.  If this is feasible then it would probably be better not to
purchase the 4 processors, but to use that money for the proposed new
equipment.


∂12-Sep-89  1346	weening@Gang-of-Four.Stanford.EDU 	Re: hardware   
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Sep 89  13:46:10 PDT
Received: from localhost by Gang-of-Four.Stanford.EDU (5.0/25-eef) id AA03853; Tue, 12 Sep 89 13:46:09 PDT
Message-Id: <8909122046.AA03853@Gang-of-Four.Stanford.EDU>
To: Carolyn Talcott <CLT@SAIL.Stanford.EDU>
Cc: JMC@SAIL.Stanford.EDU, tom@Polya.Stanford.EDU
Subject: Re: hardware 
In-Reply-To: Your message of 12 Sep 89 13:32:00 -0700.
             <1Vwwik@SAIL.Stanford.EDU> 
Date: Tue, 12 Sep 89 13:46:07 PDT
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>

    Date:  12 Sep 89  1332 PDT
    From:  Carolyn Talcott <CLT@SAIL.Stanford.EDU>

    Here is what I plan to send to DARPA (the price estimates may change
    if I get new data from Tom).  Any comments?

I spotted a few typos; let me know if you want them pointed out.
Other comments follow:
    -----------------------------------------------------------------------------
    Our ancient Dec Kl10, SAIL, will be retired (to old and expensive to
    maintain) soon (hopefully in February).  This together with the Alliant
    provides the main computing facilities used by our group.  Given the
    excessively high maintanence cost (approx 70k/yr including overhead), the
    lack of available software, etc.  the Alliant is not a viable replacement
    for SAIL.  Thus we need to purchase replacement hardware.
I'm not sure what you mean by the "lack of available software".  Alliant
runs pretty much the same code as other Unix machines.  I agree with
the rest of the points, though.  You might also add that the Alliant
is now a 4-year-old design and would have to be retired in a few years
in any case.

    We want to continue with a multi-user time-sharing system and we have
    come up with the following rough plan:

    A Dec (mips) or Sun (spark) with
      64meg memory
      4gig disk  (1-controller 4-disks)
      8mm tape drive
      ethernet controller
      serial line board
    The estimated cost for this package is around 87k (Sun with super discount).

    In addition we will need terminals for office and home.  (Sail terminals
    won't work).  Here we are thinking of the NCD terminals (about 2-2.5k each).
You should say these are "NCD X terminals".
    A minimum of 14 terminals is needed for the existing researchers
    and additional terminals will be needed for secretaries and students.
    This system would provided facilities for John McCarthy's group
    (McCarthy, Lifshitz, Talcott, Mason, Weening, Pehoushek)
    and Zohar Manna's group (Manna and Waldinger). 
We need modems also.

    The yearly cost for maintaining the new system is estimated to
    be 17k for maintenance plus 50k for systems programmer, backup, etc.
Does this include the cost of maintaining the terminals?  Tom had
estimated about $100/month per terminal, which covers the maintenance
and the service of booting and providing fonts over the network.
    This is to be contrasted with the SAIL computer charges for the two
    groups which were running at approximately 17k/month (200k/year).

    The Qlist project recently requested approval to purchase the 4 processors
    on loan (since the loan officially terminated last April and Alliant is
    pushing for a decision).  It is important to have more than 4 processors
    to get reasonable experimental data.  We are investigating the possibility
    of using 8 processor Alliants at Argonne or elsewhere for gether final
    data.  If this is feasible then it would probably be better not to
    purchase the 4 processors, but to use that money for the proposed new
    equipment.

∂12-Sep-89  1443	tom@Polya.Stanford.EDU 	hardware   
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Sep 89  14:43:51 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA16955; Tue, 12 Sep 89 14:44:16 -0700
Date: Tue, 12 Sep 89 14:44:16 -0700
From: Tom Dienstbier <tom@Polya.Stanford.EDU>
Message-Id: <8909122144.AA16955@Polya.Stanford.EDU>
To: weening@Gang-of-Four.Stanford.EDU
Cc: CLT@SAIL.Stanford.EDU, JMC@SAIL.Stanford.EDU
In-Reply-To: Joe Weening's message of Tue, 12 Sep 89 13:46:07 PDT <8909122046.AA03853@Gang-of-Four.Stanford.EDU>
Subject: hardware 



Here are some numbers that are pretty accurate.
I woukld like to get a quote from DEC on the next model up from the
DECserver 5400. This would gives us more room for future expansion.
Soon as I get them I will forward them to you. 

Tom

			SUN Microsystems 4/390



1 ea. 4/390-S-56-SPARCserver390	150900	152400
      56MB memory
      4GB disk drives
      serial ports		  1500

1 ea. SYSLUN SunSystem Software		  2000

1 ea. SS1-07 Sun OS standard software	   250

1 ea. SS-09  Full System Software
             Documentation		   450

1 ea. 801A   Exabyte tape drive 	  6500

				total    161600 

			Less 43%          69488

			  Grand Total	 $92112.00



			DEC DECsystem5400	


1 ea. DU-54FT1-AA			83,028
      16MB memory
      1.2GB disk Drive
      296MB Cart.Tape Drive
      DSSI Ethernet Cntrl.

3ea.  MS650-BA			12000	36,000	
      16MB memory
      

3ea.  RA90
      1.2GB disk		25840   77,520

				total $196,548  

		Less discount 16%       31,447.68

   		Grand total   	      $165,100.32	





		MIPS/NCD Xwindow terminals

NCD 16P11 1MB memory With full ROM set		$2800 list

NCD 16P1  1MB memory				$2500  "

Current discount is 30%.

∂12-Sep-89  1533	CLT 	cinda
To:   JMC@SAIL.Stanford.EDU, weening@Gang-of-Four.Stanford.EDU,
      tom@Polya.Stanford.EDU
We have an appointment with her 10am thursday

∂12-Sep-89  1536	tom@Polya.Stanford.EDU 	cinda 
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Sep 89  15:36:51 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA19431; Tue, 12 Sep 89 15:37:19 -0700
Date: Tue, 12 Sep 89 15:37:19 -0700
From: Tom Dienstbier <tom@Polya.Stanford.EDU>
Message-Id: <8909122237.AA19431@Polya.Stanford.EDU>
To: CLT@SAIL.Stanford.EDU
Cc: JMC@SAIL.Stanford.EDU, weening@Gang-of-Four.Stanford.EDU
In-Reply-To: Carolyn Talcott's message of 12 Sep 89  1533 PDT <twyZh@SAIL.Stanford.EDU>
Subject: cinda

Great..


tom

∂12-Sep-89  1647	RWF 	re: proposal for hardware
To:   JMC, RWF
CC:   CLT   
[In reply to message from JMC rcvd 12-Sep-89 11:23-PT.]

Sure. I'll take anything that's free.

∂12-Sep-89  1839	TYSON@AI.SRI.COM 	Networks considered harmful - Email  
Received: from Warbucks.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 12 Sep 89  18:39:39 PDT
Received: from ELCAPITAN.AI.SRI.COM by Warbucks.AI.SRI.COM with INTERNET ;
          Tue, 12 Sep 89 18:40:23 PDT
Date: Tue, 12 Sep 89 18:41 PDT
From:     TYSON@Warbucks.AI.SRI.COM (Mabry Tyson)
Subject: Networks considered harmful - Email
To:       jmc@sail.stanford.edu
Message-ID: <19890913014151.8.TYSON@ELCAPITAN.AI.SRI.COM>

Hi...  I saw your "Final version of Networks considered harmful"
(September 1) message on su.nethax.  It sounds like a good idea in
general (though there's always junk mail...).  

What I wanted to know is whether you would see any problem with me
mailing a copy of that to Symbolics.  (They do have something called
DIALNET which is not UUCP but is an over-the-phone mail system.)  I
thought they might be interested but I wanted to check with you whether
you might object or have any copyright (yours or CACM's) concerns.

∂13-Sep-89  0048	TYSON@Warbucks.AI.SRI.COM 	re: Networks considered harmful - Email    
Received: from Warbucks.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Sep 89  00:48:34 PDT
Date: Wed 13 Sep 89 00:49:14-PDT
From:     TYSON@Warbucks.AI.SRI.COM (Mabry Tyson)
Subject: re: Networks considered harmful - Email 
To:       JMC@SAIL.STANFORD.EDU
Message-ID: <621676154.0.TYSON@AI.SRI.COM>
In-Reply-To: <Bx0KJ@SAIL.Stanford.EDU>
Mail-System-Version: <VAX-MM(229)+TOPSLIB(126)@AI.SRI.COM>
Reply-To: Tyson@Warbucks.AI.SRI.COM

Ok, I won't send them the text of what was in SU.NETHAX.  From your note,
it sounds like you would prefer my not telling Symbolics anything about
your article.  So I won't.  (If I am wrong, let me know...  I just figured
that if I said anything to them, it could get out and be discussed before
your article is published (or, worse yet, someone might get something
else published on the subject before CACM's publication).)
-------

∂13-Sep-89  0129	ME 	Boyer 
To:   JMC
CC:   CLT   
 ∂13-Sep-89  0030	JMC  
To:   ME, CLT
Any idea how to mail to R. Boyer at utexas?

ME - Try boyer@cs.utexas.edu.  That forwards to some machine which may
be a workstation, but anyway it looks like a good address.

∂13-Sep-89  0458	cross@vax.darpa.mil 	Reschedule DARPA Visit  
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 13 Sep 89  04:58:22 PDT
Received: from sun46.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA05515>; Wed, 13 Sep 89 07:53:33 -0400
Posted-Date: Wed 13 Sep 89 07:59:48-EDT
Received: by sun46.darpa.mil (4.1/5.51)
	id AA09376; Wed, 13 Sep 89 07:59:50 EDT
Date: Wed 13 Sep 89 07:59:48-EDT
From: Steve Cross <CROSS@DARPA.MIL>
Subject: Reschedule DARPA Visit
To: jmc@sail.stanford.edu
Cc: cross@vax.darpa.mil
Message-Id: <621691188.0.CROSS@SUN46.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@SUN46.DARPA.MIL>

John:  Because of the need to stay here next week to work budget matters, we must 
cancel our visit.  Could we reschedule for the week of 15 October.  Tuesday afternoon 
or Thursday morning would be the best bets. Steve Cross
-------

∂13-Sep-89  0659	CLT 	boyer
To:   JMC
CC:   ME    
boyer@cli.com

∂13-Sep-89  0700	JMC  
balch

∂13-Sep-89  0945	CLT 	revised hardware proposal for DARPA
Here is the current version (with Joes input integrated).  Shall I ship it?

-----------------------------------------------------------------------------
Our ancient Dec KL10, SAIL, will be retired (too old and expensive to
maintain) soon (hopefully in February).  This together with the Alliant
provides the main computing facilities used by our group.  The Alliant is
not a viable replacement for SAIL since the maintainence cost is
excessively high (approx 70k/yr including overhead), sharing software with
non-Alliant users is difficult (big systems do not port easily), and the
Alliant itself is almost 4 years old and will soon be obsolete and
ready for retirement.  Thus we need to purchase replacement hardware.

We want to continue with a multi-user time-sharing system and we have come
up with the following rough plan:

A Dec system5400 or Sun 4/390 with (roughly)
    64meg memory
    4gig disk  (1-controller 4-disks)
    8mm tape drive
    ethernet controller
    serial line board
The estimated cost for this package is
  92k (Sun with 43% discount) or 165k (Dec with 16% discount)
  [maybe Dec can be induced to make a better deal!]

In addition we will need terminals for office and home.  (Sail terminals
won't work).  Here we are thinking of the NCD X terminals (about 2k each).
A minimum of 14 terminals (and 7 modems at .5-.9k each) is needed
for the existing researchers and additional terminals will be needed for
secretaries and students.  

This system would provide facilities for John McCarthy's group (McCarthy,
Lifschitz, Talcott, Mason, Weening, Pehoushek) and Zohar Manna's group
(Manna and Waldinger).

The yearly cost for maintaining the new system is estimated to be 75k (17k
for computer maintenance, 8k for terminal maintainence, and 50k for
systems programmer, backup, etc.)  This is to be contrasted with the SAIL
computer charges for the two groups which were running at approximately
200k/year.

The Qlisp project recently requested approval to purchase the 4 Alliant
processors currently on loan (the loan officially terminated last April
and Alliant is pushing for a decision).  Program development can be done
on 4 processors, but it is important to have more than 4 processors to get
reasonable experimental data.  We are investigating the possibility of
using 8 processor Alliants at Argonne or elsewhere for gathering final
data.  If this is feasible then it would probably be better not to
purchase the extra Alliant processors, but to use that money for the
proposed new equipment.

∂13-Sep-89  1150	TYSON@Warbucks.AI.SRI.COM 	re: Networks considered harmful - Email    
Received: from Warbucks.AI.SRI.COM by SAIL.Stanford.EDU with TCP; 13 Sep 89  11:50:34 PDT
Date: Wed 13 Sep 89 11:51:08-PDT
From:     TYSON@Warbucks.AI.SRI.COM (Mabry Tyson)
Subject: re: Networks considered harmful - Email 
To:       JMC@SAIL.STANFORD.EDU
Message-ID: <621715868.0.TYSON@AI.SRI.COM>
In-Reply-To: <Rx1ps@SAIL.Stanford.EDU>
Mail-System-Version: <VAX-MM(229)+TOPSLIB(126)@AI.SRI.COM>
Reply-To: Tyson@Warbucks.AI.SRI.COM

He recieves mail as either BOYER@CLI.COM or BOYER@CS.UTEXAS.EDU.
(I see he logged into the UT machine today from CLI so either should work.)
-------

∂13-Sep-89  1220	MPS 	Book Review    
Alex Madrigal, San Jose Merc, 777-8439, would like to
talk to you about reviewing the book

The Improbable Machine by Jeremy Campbell

Pat

∂13-Sep-89  1651	VAL 	Elephant programs   

Reversing a string:

	for i=1 to 10 do read(x);
	for j=10 downto 1 do write(x when i was j)
	
Factorial:

	n:=0; x:=1;
	for n=1 to 10 do x:=(x when n was n-1)*n

∂13-Sep-89  1702	VAL 	re: what was   
[In reply to message rcvd 13-Sep-89 16:50-PT.]

The best example: Read a travel diary and find out when the owner visited
Paris last time.

	while not eof read(year,city);
	write(year when city was Paris)

∂13-Sep-89  1919	VAL 	Second hard example 
[In reply to message rcvd 13-Sep-89 18:40-PT.]

The following program reads a string of left and right parentheses
and then points to the last unmatched left parenthesis (I think).

	i:=0; depth:=0;
	repeat
	    i:=i+1;
	    read(x);
	    if x='(' then depth:=depth+1
		     else depth:=depth-1
	until eof;
	write(i when (depth was depth) and not (i was i))

(That is, write the value that i had when depth was the same as now last time).

∂16-Sep-89  2246	rdz@Gang-of-Four.Stanford.EDU 	Someone's plan file, for your amusement
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Sep 89  23:27:45 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.0/25-eef) id AA13575; Wed, 13 Sep 89 22:50:18 PDT
Date: Wed, 13 Sep 89 22:50:18 PDT
From: Ramin Zabih <rdz@Gang-of-Four.Stanford.EDU>
Message-Id: <8909140550.AA13575@Gang-of-Four.Stanford.EDU>
To: jmc@Gang-of-Four.Stanford.EDU
Subject: Someone's plan file, for your amusement

"The workers ceased to be afraid of the bosses.  It's as if they suddenly
 threw off their chains." -- a Soviet journalist, about the Donruss coal strike

∂16-Sep-89  2257	restivo@Neon.Stanford.EDU     
Received: from Neon.Stanford.EDU by SAIL.Stanford.EDU with TCP; 16 Sep 89  22:47:12 PDT
Received:  by Neon.Stanford.EDU (5.61/25-eef) id AA09869; Fri, 15 Sep 89 12:32:13 -0700
Date: Fri, 15 Sep 1989 12:32:12 PDT
From: Chuck Restivo <restivo@Neon.Stanford.EDU>
To: jmc@sail
Message-Id: <CMM.0.88.621891132.restivo@Neon.Stanford.EDU>

Would you recommend if you know of a particularly comprehensive index of
symbols and their meaning - the sort of thing in most of the usual into
to math logic texts is not enough.

Thank you.

		Chuck Restivo

∂17-Sep-89  0231	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Life query . . .   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 17 Sep 89  02:31:42 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 420413; 15 Sep 89 06:34:07 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108316; Fri 15-Sep-89 03:34:56 PDT
Date: Fri, 15 Sep 89 03:34 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Life query . . .
To: TREV@STONY-BROOK.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "Rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hpm@rover.ri.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "acad!megalon!rudy@uunet.UU.NET"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rem@suwatson.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890914144217.3.TREV@BARNABUS.SCH.Symbolics.COM>
Message-ID: <19890915103409.8.RWG@TSUNAMI.SPA.Symbolics.COM>

    Date: Thu, 14 Sep 89 07:42 PDT
    From: trev@VERMITHRAX.SCH.Symbolics.COM

	Date: Thu, 14 Sep 89 07:12 PDT
	From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>

	without actually having to fondle the dots.

	In an Oct 77 letter, David Buckingham, Life magician,
	listed the known oscillator periods as

	[....]

    Hi Bill,
    I'm saving all these reports in a life-archive, against
    the day I'll be able to understand them better.  I've
    been giving the hacks:demo life program random dot fields
    for years now, without ever studying the patterns closely.
    Is there any kind of coherent publication on the subject,
    maybe an encyclopedia of life-forms, or a detailed history
    of the game?

The best (but poor) approximation is the collection of 11(?)
Lifeline newsletters issued in 1971-73.  "The Recursive Universe"
by William Poundstone, actually has a lot of good Life stuff.
Martin Gardner's early columns, which started the craze, were
reprinted, with updates, in "Wheels, Life, and Other
Mathematical Amusements" Freeman, 1983.  There is also a good
section in Volume 2 of Winning Ways, by Berlekamp, Conway, & Guy.
My unbounded, reversible algorithm is in Physica 10D (1984) p75-80.

		  On-line, even?

Most Life transpired in the days of mainframes, with personal
disk allocations smaller than current physmems.

				  Or an editor that lets you
    setup initial fields with the mouse (this might be a good
    weekend project).  If you have any authoritative references,
    please let Math-Fun know too...some of my friends are just
    as puzzled by the names of these creatures!

    trev

There are several problems with hacks:life.

0)  It has edges.  (I have a version that doesn't.)

1)  While some edge-related bogosity is unavoidable, the
    usual edge conventions are relatively benign:  "Toroidal
    boundary conditions" (the glider wraps around), or
    "zero boundary conditions", where the glider turns into
    block.  I also have an experimental Life where the edges
    are read-only.  They stay how you set them.  But the top
    (at least) edge in hacks:life can be weird and malignant,
    with vigorous noise sources moving gradually to the right.

2)  1 pixel/bit causes eye- and brain-strain.

3)  It can be done in about half that many bitblts.

4)  As you say, no (e.g. mouse) control of dots.

However, mousing in dots, while useful, is neither necessary
nor sufficient to do interesting experiments.  And watching
random soups has a lower expectation than the state lottery.

This is not completely fair.  The gun shuttle, and gliders
themselves, were found by soup-watching.  But despite the
unlimited possibilities, the *probabilities* in soups are
skewed astronomically in favor of a few trite events and
outcomes.

Thus, you really need to be able to collect and build a library
of modules.  E.g., that stretchable period 5 I sent was generated by

(defcell rwg-sek (let ((frob (make-bloc :rot-180 :shift 2 -1 :merge (make-bloc))))
		   (make-eater :merge (send frob :shift -2 0)
			       :merge (send frob :shift -6 -4)
			       :merge (send frob :shift -10 -8)
			       :merge (make-eater :rot-180 :shift -10 -9))))

and that's the *low level*.  I've actually got a sort of CAD system
for the harder stuff, e.g.,

(defmethod (:init p40-g-puffer) (plist)
  (push (send (car inferiors) :poise-preblock
	      (send (send (make-instance 'middleweight-poofs-block-from-left)
			  :inverse-poise :poise-block)
		    :swap-xy :poise-middleweight
		    (make-instance 'smart-spaceship :age (- (zl:get plist :delay)))))
	inferiors))     ; (Pardon the archaic phraseology.)

i.e., a separate flavor for every event!

Then there's complaint number

5)  If you do happen to witness a miracle, or more probably *think*
    you might have seen one, you can't back up and capture it.

My unbounded version can back up.  (By the stratagem of never actually
going forward in the first place, since the Life rule is obviously
irreversible.)  However, the program is rotten on soups, and clunky to
watch:  It is usually faster to peek ahead 16 steps than 1.

Also, it lacks a nice mousey editor for moving stuff around.

Ideally, we'd make some life-toolkit-in-the-sky hybrid of bitblts, mice,
and flavors.  But if the DEA ever snuffs cocaine, they'd probably come
after us next.

Random addendum:  The crux of the p1100 gun was a spark (from a p100)
turning a reproducible block into a glider.  A pentadecathlon can do
the same job, with the benefit of having an odd period.  The only
missing component, then, to a p450 (or maybe even 330) gun made
entirely from pre-gun1 technology is something to turn p30 (say)
gliders 90 degrees.  The simplest such mechanism I know is a
shuttle+eater found by the Canadians long after gun1.  But if there
had been no gliderogenous shuttle collision, the Canadian reflector
would have given us a tantalizing gun design that was almost certain
to work, but too big to build!  Call it gun3andahalf.

∂17-Sep-89  0231	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Buckingham's p10 (unique period)  
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 17 Sep 89  02:31:44 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 420426; 15 Sep 89 09:11:44 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108320; Fri 15-Sep-89 06:12:36 PDT
Date: Fri, 15 Sep 89 06:11 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Buckingham's p10 (unique period) 
To: "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "Rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hpm@rover.ri.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "acad!megalon!rudy@uunet.UU.NET"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rem@suwatson.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890915131149.0.RWG@TSUNAMI.SPA.Symbolics.COM>

 . . . . . . o o . o . . .
 . . . . . o . . o o . . .
 . . . . . . o . . . . . .
 . . . o o o . o o o . . .
 . . o . . . . o . . o . o
 . . . o o o . o . o . o o
 o o o . . . . o . o . . .
 o . . o o o o o . o . . .
 . . . o . . . . o . . . .
 . . . . o o o o . . . . .
 . . . . . . . . . . . . .
 . . . . . . o o . . . . .
 . . . . . . o o . . . . .

Ramblings about the size of oscillators:  Life squanders
real estate.  A given area counts through a minute fraction
of its potential states.  So oscillators are large compared
to their periods.  Can this be quantified?

How many states can a given area achieve?  (Harder than the
Garden of Eden problem?)  What is the largest area that can
be counted through all states?  (= Minimum GoE -1 ?  Could a
"grandorphan" be smaller?)  Achieved by surrounding it with
an appropriate glider storm?

Clearly, above some size, maximum period as a function of area must
grow exponentially, because one could build a Turing type, but
finite "eons counter" only a fixed size increment over its memory.
What is the base of that exponential?  I.e., the densest possible
memory.

∂17-Sep-89  0327	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Period p guns imply period n p oscillators.  Magic.   
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 17 Sep 89  03:27:34 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 421308; Sun 17-Sep-89 06:26:36 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108468; Sun 17-Sep-89 03:20:43 PDT
Date: Sun, 17 Sep 89 03:19 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Period p guns imply period n p oscillators.  Magic.
To: "Rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hpm@rover.ri.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "acad!megalon!rudy@uunet.UU.NET"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "gumby@sumex-aim.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19890915103409.8.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19890917101955.7.RWG@TSUNAMI.SPA.Symbolics.COM>

Dean Hickerson unwedged me.  You only need one glider annihilation
geometry.  If you use it as a negating 90 deg turn at all four
corners of a square circulating memory, you are constrained by
symmetry to 4 n bits, and period 4 n p.  But now all you need is
to load the memory with four copies of a period n bit sequence!

Lest anyone doubt that Buckingham was from another planet,

 . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 o . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . o o . . . . . . . . . . . . . . . . . . . . . . o . o . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . o o . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . o . . . . .
 . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . o . . . . . . . . . . . . . . . . . . . . . . . . . . o .
 . o o o . . . . . . . . . . . . . . . . . . . . . . . . . . o . o
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . o o
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . o . o . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . o o . . . . . . . . . . . . .
 . . . . . o . . . . . . . . . . . . o . . . . . . . . . . . . .
 . . . . . . o o . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . o o . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . o . o . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . o o . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . o . . . . . . o . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . o o . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . o o . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . o o . . . . . . . . o o . . . . . . . . o . . . . . . . . .
 . . . o o . . . . . . . . o o . . . . . o o . . . . . . . . . .
 . . o . . . . . . . . . o . . . . . . . . o o . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . o . . . . . . . . . o o
 . . . . . . . . . . . . . . . . . . . o o . . . . . . . . o o .
 . . . . . . . . . . . . . . . . . . . o . o . . . . . . . . . o

And this is only his second best.

∂17-Sep-89  0716	pamela@cs.columbia.edu 	WGBH/BBC project
Received: from cs.columbia.edu by SAIL.Stanford.EDU with TCP; 17 Sep 89  07:15:58 PDT
Received: by cs.columbia.edu (5.59++/0.1) 
	id AA12420; Sun, 17 Sep 89 10:16:50 EDT
Date: Sun, 17 Sep 1989 10:16:49 EDT
From: Pamela McCorduck <pamela@cs.columbia.edu>
To: Feigenbaum@sail.stanford.edu, jmc@sail.stanford.edu
Cc: pamela@cs.columbia.edu
Subject: WGBH/BBC project
Message-Id: <CMM.0.88.622045009.pamela@cs.columbia.edu>

As you may know, Boston's WGBH and the BBC are planning a joint-venture TV
series tentatively called "The Age of Information", the backers being the
ACM and Unisys.  I've been asked to consider writing the accompanying book.

When I saw the prospectus I suggested that instead of the usual chronology
(Part I: Babbage invents the computr'; Part 4: two guys invent the Apple)
it might be better to organize arouhnd major topics.  That way both past
and future can be addressed, especially the largescale and possibly
radical influences of information technology.

I'm asking people who've obviously given this some thought to help me
select the best topics.

My own tentative and unordered list is: Simulation; Visualization; 
Miniaturization; Complexity; AI (which could come under Simulation, but
is important enough to devote a chapter/episode to in itself, especially
if it includes Robotics); Co-Evolution (human machine symbiosis); Dis-
tributed Processing; Randomization; Global Telecommunications...

Surely not a complete list, much overlap.  Can you improve it?  Do
you have favorite examples that would be folly to exclude?

Many thanks.

Pamela

∂17-Sep-89  0912	@Score.Stanford.EDU,@AI.ai.mit.edu,@REAGAN.ai.mit.edu:dam@ai.mit.edu 	Justifying Resolution    
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Sep 89  09:12:35 PDT
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU.#Internet) by SCORE.STANFORD.EDU with TCP; Sun 17 Sep 89 09:11:22-PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa12506;
          17 Sep 89 12:10 EDT
Received: from REAGAN.AI.MIT.EDU (CHAOS 13065) by AI.AI.MIT.EDU 17 Sep 89 12:13:20 EDT
Received: from ELVIS.AI.MIT.EDU by REAGAN.AI.MIT.EDU via CHAOS with CHAOS-MAIL id 261276; Sun 17-Sep-89 12:08:15 EDT
Date: Sun, 17 Sep 89 12:07 EDT
From: "David A. McAllester" <dam@ai.mit.edu>
Subject: Justifying Resolution
To: theorem-provers@AI.ai.mit.edu
Message-ID: <19890917160735.3.DAM@ELVIS.AI.MIT.EDU>

I'm teaching a graduate course in AI and plan to cover resolution
theorem proving.  I have never studied resolution very deeply.  I would
like to present technical results that support the intuition that
various restrictions and/or variants of resolution are superior to basic
binary resolution.  Often restrictions are defined and proven complete
but no justification for their superiority is ever given.  Here are some
more specific questions.

In the ground case simple binary resolution can be run to completion in
time O((3↑n)↑2) where n is the number of atomic formulas (there are at
most 3↑n clauses and each pair of clauses represents a possible
resolution).  Is there a tighter simple bound on the running time of
resolution in the ground case?

Input resolution (every resolution involves an original clause) has a
ground-case running time of O(|C|3↑n) where |C| is the size of the imput
clause set.  This is better than (3↑n)↑2 so we have a justification for
the superiority of input resolution.  Does this techical result capture
the intuition behind input resolution?

Hyper-resolution seems to have a ground-case running time O((|C|↑n)2↑n)
where |C| is the number of original clauses.  Does this capture the
appropriate intuition behind hyper-resolution?  I hope not, this seems
like a pretty bad bound.

Are the above bounds correct?

The above bounds indicate that in the ground case various versions of
resolution have running time on the order of 2↑{cn} with c equal to 2 or
greater.  The best known decision procedure for SAT has running time
bounded by 2↑{cn} where c is roughly 1/4.  Thus, resoltion seems to be a
horrible technique in the ground case.  Is there some technical result
(theorem) that intuitively justifies resolution in general?

An alternative to ground-case analysis is ``acceleration theorems''.
For example, I would like to prove that some particular variant of
resolution, say variant X, NEVER uses more refutation steps in some
standard breadth-first search strategy than does standard binary
resolution.  I have heard that there are some, but not many, results
like this.  Can anyone tell me what they are?

Does anyone know of a technical result justifying unit resolution (other
than the fact that it is complete for Horn clauses, which does not
impress me).  The number of literals in a clause (in the ground case)
provides a lower bound on the number of resolutions needed to derive the
empty clause from the given clause.  Can one simply apply the A* search
technique to resolution using literal-count as the lower bound?  This
would seem to yield a version of unit resolution with the same technical
justification as A* in general.  Does an A* (or AO*) analysis make sense
for resolution?

How about a technical result justifying set-of-support resolution?

	Thanks, David McAllester

∂17-Sep-89  1042	qphysics-owner@neat.cs.toronto.edu 	Call for papers: AIJ Special Issue on Qualitative Physics  
Received: from neat.cs.toronto.edu by SAIL.Stanford.EDU with TCP; 17 Sep 89  10:42:37 PDT
Received: by neat.cs.toronto.edu id 2139; Sun, 17 Sep 89 13:39:19 EDT
Subject: Call for papers: AIJ Special Issue on Qualitative Physics
To:	qphysics@ai.utoronto.ca
cc:	dekleer.PA@xerox.com, bwilliams.PA@xerox.com
Resent-From: qphysics-owner@cs.toronto.edu
Resent-To: qphysics@cs.toronto.edu
Resent-Reply-To: bwilliaws.PA@xerox.com
Resent-Message-Id: <89Sep17.133919edt.2139@neat.cs.toronto.edu>
Resent-Date: Sun, 17 Sep 89 13:38:35 EDT


    CALL FOR PAPERS
 
    Artificial Intelligence Special Volume on
    ``Qualitative Reasoning about Continuous Physical Systems''

    The birth of `qualitative physics', as the field is commonly called,
    dates to a special volume of Artificial Intelligence that appeared in
    December 1984.  Since then much has happened, including three workshops
    on qualitative physics, and it is time for another special volume.  In
    particular, we seek reports of progress which both:

    o make advances in automated reasoning about continuous physical systems
      using qualitative representations.

    o AND provide a rationale in terms of tasks, such as diagnosis, design,
      explanation and process monitoring, or fundamental problems in
      qualitative physics.

    Note that this special volume has a narrower focus than the one that
    appeared in 1984 -- the 1984 issue incorporated papers on model-based
    reasoning in general, while the current issue focuses specifically on
    qualitative reasoning about continuous physical systems.  For example,
    the 1984 papers on diagnosis and design verification would not be
    included in the current volume, because they do not address issues
    specific to qualitative representations of continuous physical systems.
    However, we solicit papers that both address these issues and are set
    within the context of a task.

    Authors should submit 6 copies of a manuscript to the special volume
    editors:

    Johan de Kleer and Brian C. Williams
    Xerox PARC
    3333 Coyote Hill Road
    Palo Alto, CA 94304

    Manuscripts must be received by February 1, 1990.  To expedite the
    production of the special volume the contributors will be required to
    rigidly adhere to dealines and to help facilitate the reviewing process.
-------

∂18-Sep-89  0033	weening@Gang-of-Four.Stanford.EDU 	Re: sd to go4 when home  
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  00:33:37 PDT
Received: from LOCALHOST by Gang-of-Four.Stanford.EDU with SMTP (5.61/25-eef) id AA09740; Mon, 18 Sep 89 00:33:34 -0700
Message-Id: <8909180733.AA09740@Gang-of-Four.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: sd to go4 when home 
In-Reply-To: Your message of 17 Sep 89 23:47:00 -0700.
             <BzDjT@SAIL.Stanford.EDU> 
Date: Mon, 18 Sep 89 00:33:33 PDT
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>

If you type "stty rows 22" (or whatever the number of lines is) before
starting Emacs, it should adjust properly.

You might want to consider using Telnet in transparent mode instead of
Supdup, to make use of all 24 lines of the screen.  I have a character
macro on SAIL that I use for this.  After doing "telnet go4" and
getting a shell prompt, I run the character macro, which does:
	source .dmwaits<return> <meta>t
The <meta>t is at the end because SAIL doesn't seem to read the rest
of the macro if it is earlier.  My file .dmwaits on Gang-of-Four does:

set term=dmwaits
stty 2400
echo -n '↑]↑↑'

The characters in quotes make it enter roll mode and clear the screen.
I'm representing them here by ↑-combinations, but to actually get them
in the file with Emacs, one types <ctrl>Q <ctrl>] <ctrl>Q <ctrl>↑.

∂18-Sep-89  0932	weening@Gang-of-Four.Stanford.EDU 	Gang of Four   
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  09:32:47 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.61/25-eef) id AA11050; Mon, 18 Sep 89 09:32:39 -0700
Date: Mon, 18 Sep 89 09:32:39 -0700
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8909181632.AA11050@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Gang of Four

Do you know the names of the Gang of Four?  When we get four X
terminals, I think it would look neat to name them after those people.

∂18-Sep-89  1253	restivo@Neon.Stanford.EDU 	Re: reply to message   
Received: from Neon.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  12:53:10 PDT
Received:  by Neon.Stanford.EDU (5.61/25-eef) id AA00952; Mon, 18 Sep 89 12:53:42 -0700
Date: Mon, 18 Sep 1989 12:53:42 PDT
From: Chuck Restivo <restivo@Neon.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: reply to message 
In-Reply-To: Your message of 17 Sep 89 1126 PDT 
Message-Id: <CMM.0.88.622151622.restivo@Neon.Stanford.EDU>

Dear Professor McCarthy,


      BC9      Feys, R. & Fitch, F.B.
      F4         Dictionary of symbols of mathematical logic. North-
               Holland, 1969.

      QA9      Greenstein, Carol Horn.
      G698       Dictionary of logical terms and symbols. Van Nostrand
               Reinhold Company, 1978.

I found these.  They are what I was after.  

Best,
Chuck Restivo

∂18-Sep-89  1329	MPS 	Re:  JMC again (again)   
 ∂18-Sep-89  1321	drb@cscfac.ncsu.edu 	Re:  JMC again (again)  
Received: from ncsuvx (ncsuvx.ncsu.edu) by SAIL.Stanford.EDU with TCP; 18 Sep 89  13:21:22 PDT
Received: by ncsuvx (5.57/2 4/27/87)
	id AA06446; Mon, 18 Sep 89 16:20:50 EDT
Posted-Date: Mon, 18 Sep 89 16:22:46 edt
Received: by cscfac.ncsu.edu (1.2/Ultrix2.0-B)
	id AA07988; Mon, 18 Sep 89 16:22:46 edt
Date: Mon, 18 Sep 89 16:22:46 edt
From: drb@cscfac.ncsu.edu (Dennis R. Bahler)
Message-Id: <8909182022.AA07988@cscfac.ncsu.edu>
To: MPS@sail.stanford.edu
Subject: Re:  JMC again (again)

Pat, I just took a look at the hall and I see no reason to fool with
35mm slides.  Overheads will do fine visibility-wise assuming they're
sized 14 pt or so or the equivalent.

Gee, I can't think of anything else today.

Dennis Bahler

This is the message I got from Dennis.  Do you still want to make
slides?

Pat

∂18-Sep-89  1341	Mailer 	re: New Workers  
Received: from rml2.sri.com by SAIL.Stanford.EDU with TCP; 18 Sep 89  13:41:47 PDT
Date: Mon 18 Sep 89 13:38:42-PST
From: Ric Steinberger <RIC@RML2.SRI.COM>
Subject: re: New Workers  
To: JMC@SAIL.STANFORD.EDU
Cc: RIC@RML2.SRI.COM, comments@KL.SRI.COM, su-etc@SAIL.STANFORD.EDU
Message-ID: <622154322.870000.RIC@RML2.SRI.COM>
In-Reply-To: <16zZFF@SAIL.Stanford.EDU>
Mail-System-Version: <VAX-MM(246)+TOPSLIB(136)@RML2.SRI.COM>

>	From: John McCarthy <JMC@SAIL.Stanford.EDU>
>	Since that time [middle 1960s] the
	technology of productivity has improved, but few people have such
	an opportunity today.  Please list all the reasons you know for
	the change which has gotten worse since 1970.
	
* As you say, the technology of production has improved.  Thus 1 worker and
  some "intelligent" machinery can produce more goods or control more processes
  than previously possible.  
		
* Foreign competition.  The Japanese are the prime example, but the West 
  Germans, and the Koreans, and other countries have gained increasing
  percentages of many world markets (e.g., automobiles, steel) that used
  to be primarily "American."

* Foreign competition II.  Many countries, especially in South East Asia
  are able to offer factories and labor at prices that essentially makes
  production in the US prohibitive.

* US labor unions.  They have sometimes demanded, and been given, conditions
  of employment that have resulted in US companies loosing competitiveness.

* US investment policy.  All to often, US corporations have to act to
  make the next quarterly report look "good", instead of being able to
  invest in longer term research and development.

* US investment policy II.  It has been very profitable to merge and divest
  companies in the 1980s.  Little net wealth is created, but Wall Street
  has done well.

* There has been few laws passed that would result in worker retraining
  for those former employees of places like steel mills.

* The lack of quality control in American manufacturing.  In many many
  industries from automobiles to semiconductors, the Japanese and West
  Germans are producing products with quality superior to those made
  in America.  Why are 50% of the cars in CA imports?  Why are all the
  DRAMS made offshore?

That's not *all* I can think of, but I need to get back to *my* work. %-)

-ric steinberger

-------

∂18-Sep-89  1412	hayes@kanga.parc.xerox.com 	Vlad Dabija 
Received: from arisia.Xerox.COM by SAIL.Stanford.EDU with TCP; 18 Sep 89  14:08:38 PDT
Received: from kanga.parc.Xerox.COM by arisia.Xerox.COM with SMTP
	(5.61+/IDA-1.2.8/gandalf) id AA02376; Mon, 18 Sep 89 14:05:19 -0700
Received: by kanga.parc.xerox.com
	(5.61+/IDA-1.2.8/gandalf) id AA20729; Mon, 18 Sep 89 14:09:11 PDT
Message-Id: <8909182109.AA20729@kanga.parc.xerox.com>
Date: Mon, 18 Sep 89 14:09:11 PDT
From: Pat Hayes <hayes@parc.xerox.com>
To: nilsson@score.stanford.edu, McCarthy@sail.stanford.edu
Cc: hayes@parc.xerox.com
Subject: Vlad Dabija


Nils and John,

this guy ( Vlad Dabija ) is a Romanian who has been in LA for five
months and wants to take a graduate degree in AI at Stanford.  He sent
me copies of his papers and an interesting letter.  He certainly seems
very smart ( GRE Analytical 800, Quantitative 790 ) and very
motivated, and his English is fine. He is all primed to take the TOEFL
exams at their next incarntion in October.  He wants to come up here
and talk to me and any faculty who might be interested in him, and I
think he might well be very interesting to several people ( he is
especially interested in Krep and machine learning ).

What should I tell him to do?  Is there some kind of procedure for
sheduling such a visit, or are such cases handled on a one-by-one
basis?  Should I circulate the faculty to see if anyone is interested,
or is there some more formal process?

I have his resume ( and copies of all his papers ): he is currently
working for Ashton-Tate.  While in Romania he implemented an ATN
compiler, a learning discovery system, a resolution TP, a coloring
puzzle-solver, and a rule-oriented expert systems shell, and applied
it to CAD and diagnostics, all on his own.  If anyone wants to talk
directly to him, phone him at (213)839-4881.

Pat Hayes



∂18-Sep-89  1441	weening@Gang-of-Four.Stanford.EDU 	Sun salesman visit  
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  14:41:42 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.61/25-eef) id AA11736; Mon, 18 Sep 89 14:41:35 -0700
Date: Mon, 18 Sep 89 14:41:35 -0700
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8909182141.AA11736@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Sun salesman visit

I'm not sure if Carolyn or Tom mentioned this to you yet, but Tom says
he has confirmed 2:00 this Wednesday for our appointment with the Sun
salesman.  Also, I just received a message saying that Sun has
increased their discount to Stanford from 30% to 38%.

∂18-Sep-89  1442	nilsson@Tenaya.Stanford.EDU 	re: Vlad Dabija 
Received: from Tenaya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  14:42:05 PDT
Received:  by Tenaya.Stanford.EDU (NeXT-0.8/25-eef) id AA04016; Mon, 18 Sep 89 14:40:20 PDT
Date: Mon, 18 Sep 89 14:40:20 PDT
From: Nils Nilsson <nilsson@tenaya.stanford.edu>
Message-Id: <8909182140.AA04016@Tenaya.Stanford.EDU>
To: Pat Hayes <hayes@parc.xerox.com>
Subject: re: Vlad Dabija
Cc: nilsson@score.stanford.edu, McCarthy@sail.stanford.edu,
        hayes@parc.xerox.com

You might arrange an appt. for Dabija with Sharon
Hemenway (hemenway@cs; our graduate programs
administrator) who can give him forms
and tell him about how to apply to our PhD program.
Beyond that, we have no formal procedure for having
prospective students meet faculty.  You can send
a note around to the faculty mentioning this guy and
see who is interested in talking to him.  I would be 
glad to see him (as Dept. chairman) for a few minutes.

-Nils

∂18-Sep-89  1640	weening@Gang-of-Four.Stanford.EDU 	Re: Sun salesman visit   
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  16:40:44 PDT
Received: from LOCALHOST by Gang-of-Four.Stanford.EDU with SMTP (5.61/25-eef) id AA12109; Mon, 18 Sep 89 16:40:40 -0700
Message-Id: <8909182340.AA12109@Gang-of-Four.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: Re: Sun salesman visit 
In-Reply-To: Your message of 18 Sep 89 15:04:00 -0700.
             <fzyvr@SAIL.Stanford.EDU> 
Date: Mon, 18 Sep 89 16:40:39 PDT
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>

Ed Washington.

∂18-Sep-89  2100	JMC  
vtss discrepancies

∂18-Sep-89  2303	harnad@clarity.Princeton.EDU 	Visual Search & Complexity: BBS Call for Commentators  
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 18 Sep 89  23:03:22 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA16912; Tue, 19 Sep 89 01:50:37 EDT
Received: by psycho.Princeton.EDU (4.0/1.85)
	id AA11047; Tue, 19 Sep 89 01:35:09 EDT
Date: Tue, 19 Sep 89 01:35:09 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8909190535.AA11047@psycho.Princeton.EDU>
To: harnad@elbereth.rutgers.edu
Subject: Visual Search & Complexity: BBS Call for Commentators

To: BBS Associates

Below is the abstract of a forthcoming target article to appear in
Behavioral and Brain Sciences (BBS). Commentators must be current BBS
Associates or nominated by a current BBS Associate. To be considered as
a commentator on this article or to suggest other appropriate
commentators please send email to:
	 harnad@princeton.edu              or write to:
BBS, 20 Nassau Street, #240, Princeton NJ 08542  [tel: 609-921-7771]
____________________________________________________________________

Analyzing Vision at the Complexity Level

John K. Tsotsos

Department of Computer Science,
University of Toronto and
The Canadian Institute for Advanced Research

The general problem of visual search can be shown to be computationally
intractable in a formal complexity-theoretic sense, yet visual search
is widely involved in everyday perception and biological systems manage
to perform it remarkably well. Complexity level analysis may resolve
this contradiction. Visual search can be reshaped into tractability
through approximations and by optimizing the resources devoted to
visual processing. Architectural constraints can be derived using the
minimum cost principle to rule out a large class of potential
solutions. The evidence speaks strongly against purely bottom-up
approaches to vision. This analysis of visual search performance in
terms of task-directed influences on visual information processing and
complexity satisfaction allows a large body of neurophysiological and
psychological evidence to be tied together.

∂19-Sep-89  0854	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	p135, 255, etc.    
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 19 Sep 89  08:54:28 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 422072; 19 Sep 89 11:10:26 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108646; Tue 19-Sep-89 08:04:37 PDT
Date: Tue, 19 Sep 89 08:03 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: p135, 255, etc.
To: "Rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "hpm@rover.ri.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "acad!megalon!rudy@uunet.UU.NET"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19890919150343.5.RWG@TSUNAMI.SPA.Symbolics.COM>

Mixing an idea from Buckingham with unsymmetrical gliderpong,

 . . . . . . . . . . . . . . . . . o o o . . . . . . . . o o o . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . o . . . o . . . . . . o . . . o . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . o . . . o . . . . . . o . . . o . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . o o o . . . . . . . . o o o . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . o o o . . . . . . . . o o o . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . o . . . o . . . . . . o . . . o . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . o . . . o . . . . . . o . . . o . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . o o o . . . . . . . . o o o . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . o o . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . o o . . . . . . . . . . . . . . . . . . . . . . .
 o o o o o o o o o o . . . . . . . . . . . . . o o . . . . . . . . . . . . . o o o o o o o o o o
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . o o . . o o . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . o o . . o o . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . o o . . o o . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . o o . . . . . . o o . . . . . . . . . . . . . . . . . . . . . . . . . . o o . . . . . . o o .
 o . . o . . . . o . . o . . . . . . . . . . . . . . . . . . . . . . . . o . . o . . . . o . . o
 o . . o . . . . o . . o . . . . . . . . . . . . . . . . . . . . . . . . o . . o . . . . o . . o
 o . . o . . . . o . . o . . . . . . . . . . . . . . . . . . . . . . . . o . . o . . . . o . . o
 . o o . . . . . . o o . . . . . . . . . . . . o o . . . . . . . . . . . . o o . . . . . . o o .
 . . . . . . . . . . . . . . . . . . . . . . . o o . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

(Funsters:  In a couple of days I'll distribute a 1.5 page Life pgm that
 will eat these dot formats, and run patterns this size at about 10 hz.)

In my copy of LiFanatic, Buckingham(?) labels the symmetrical version
of this as 240+60n.  If this is a mistake, it would be his first known
to me.  But I see no way of fudging such geometries by < 120 ticks, =
lcm(8,15).  (Any stretch of one cell will lengthen the glider orbits
 by 2, and thus 8 ticks.  But this will no longer sync with the
 pentadecathlons.  So you have to stretch by 15 cells = 120 ticks.)

Maybe he meant that, since you get 60+120n with vanilla gliderpong,
the union of the two hacks is . . . no, he wouldn't be so imprecise.

Dean had the temerity to ask for Buckingham's best:

 . . . . . . . . . . . . . o . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . o . o . . . . . . . . . . . . .
 . . . . . . . . . . . . . o o . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . o . . . . . . . . . . . . . . . . . . . . . .
 . . . . o . o . . . . . . . . . . . . . . o . o . . . . .
 . . . . . o o . . . . . . . . . . . . . . o o . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . o . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . o . . . . . . . . . . o . . . . . . .
 . . . . . . . . . . . o o . . . . . . . . o . o . . . . .
 . . . . . . . . . . o o . . o o o . . . . o o . . . . . .
 . . . . . . . . . . . . . . o . . . . . . . . . . . . . .
 . . . . . . . . . . . o . . . o . . . . . . . . . . . . .
 . . . . . . . . . . . o o . . . . . o o o . . . . o o . .
 . . . . . . . . . . o . o . . . . . o . . . . . . o . o .
 . . . . . . . . . . . . . . . . . . . o . . . . . o . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 o o . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . o o . . . . . . . . . . . . . . . . . . . . . . . . . .
 o . . . . . . . o o . . . . . . . . . . . . . . . . o o .
 . . . . . . . . . o o . . . . . . . . . . . . . . . o . o
 . . . . . . . . o . . . . . . . . . . . . . . . . . o . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . o o . . . . . . . .
 . . . . . . . . . . . . . . . . . . . o . o . . . . . . .
 . . . . . . . . . . . . . . . . . . . o . . . . . . . . .
 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
 . . . . . . . . . o o . . . . . . . . . . . . . . . . . .
 . . . . . . . . o . o . . . . . . . . . . . . . . . . . .
 . . . . . . . . . . o . . . . . . . . . . . . . . . . . .

This is among the most astonishing intellectual achievements that
I have ever been able to appreciate.  The guy was *dangerous*.

On the other hand, when you see Dean's latest, you will see why
he is entitled to all the temerity he wishes.

∂19-Sep-89  1028	DIGEX@ai.ai.mit.edu 	Electronic Mail and networks 
Received: from lcs.mit.edu (MINTAKA.LCS.MIT.EDU) by SAIL.Stanford.EDU with TCP; 19 Sep 89  10:28:12 PDT
Received: from ai.ai.mit.edu by mintaka.lcs.mit.edu id aa15784;
          19 Sep 89 13:12 EDT
Date: Tue, 19 Sep 89 13:13:17 EDT
From: Doug Humphrey <DIGEX%AI.AI.MIT.EDU@mintaka.lcs.mit.edu>
Subject:  Electronic Mail and networks
To: jmc@sail.stanford.edu
cc: MTD%AI.AI.MIT.EDU@mintaka.lcs.mit.edu, 
    RS%AI.AI.MIT.EDU@mintaka.lcs.mit.edu, 
    DIGEX%AI.AI.MIT.EDU@mintaka.lcs.mit.edu
Message-ID: <645993.890919.DIGEX@AI.AI.MIT.EDU>


After reading your message concerning the problems of 'networking'
being linked to 'electronic mail' and how it is limiting email,
I figured I had to send you a message and let you know that your
views on this match my own very closely.  So much so that we (a small 
start-up company known as Digital Express, Inc.) are building just 
such an Electronic Mail product.  I think all of us were a little
shocked at how directly you hit the nail on the head with your
message.  I think that we are close to being ready to go into an
alpha test phase; if you feel strongly about all of this, maybe we
could discuss things, and include you in on our testing of the alpha code?
Though we are a commercial concern, we are very much interested in
doing the Right Thing, if we can just figure out what that might be...

Sincerely yours,

Doug Humphrey

digex @ ai.ai.mit.edu
digex @ tumtum.cs.umd.edu

∂19-Sep-89  1229	pamela@cs.columbia.edu 	re: WGBH/BBC project 
Received: from cs.columbia.edu by SAIL.Stanford.EDU with TCP; 19 Sep 89  12:29:37 PDT
Received: by cs.columbia.edu (5.59++/0.1) 
	id AA22845; Tue, 19 Sep 89 15:30:25 EDT
Date: Tue, 19 Sep 1989 15:30:23 EDT
From: Pamela McCorduck <pamela@cs.columbia.edu>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: WGBH/BBC project
In-Reply-To: Your message of 18 Sep 89 2032 PDT
Message-Id: <CMM.0.88.622236623.pamela@cs.columbia.edu>

John, would love to see you.  Tuesday and Wednesday are pretty awful, but
other times look good.  You can call me when you arrive: 222-9057.

Pamela

∂19-Sep-89  1316	bergman@Polya.Stanford.EDU 	new NSF grant    
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Sep 89  13:16:25 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA25186; Tue, 19 Sep 89 13:16:38 -0700
Date: Tue, 19 Sep 1989 13:16:37 PDT
From: Sharon Bergman <bergman@Polya.Stanford.EDU>
To: jmc@sail
Cc: clt@sail, mps@sail, littell@Polya.Stanford.EDU, bscott@Polya.Stanford.EDU,
        bergman@Polya.Stanford.EDU, val@sail
Subject: new NSF grant 
Message-Id: <CMM.0.88.622239397.bergman@Polya.Stanford.EDU>

John,   Here is the account number for your new NSF grant (BASIC
RESEARCH IN AI):

	Account number:		2-DMA531
	Fund number:		163D006
	NSF reference number:	IRI-8904611
	Performance period:	8/1/89-7/31/89 (plus 6 months flex. period)

-Sharon Bergman

∂19-Sep-89  1427	bergman@Polya.Stanford.EDU 	re: new NSF grant     
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Sep 89  14:27:01 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA29197; Tue, 19 Sep 89 14:27:29 -0700
Date: Tue, 19 Sep 1989 14:27:28 PDT
From: Sharon Bergman <bergman@Polya.Stanford.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: new NSF grant 
In-Reply-To: Your message of 19 Sep 89 1320 PDT 
Message-Id: <CMM.0.88.622243648.bergman@Polya.Stanford.EDU>

John,   This is a 3-year continuing grant.  If the 2nd and 3rd years are
funded, the end date is 7/31/92.  2nd and 3rd year funding is
contingent on the availability of funds and the scientific progress of
the project.
-Sharon

∂19-Sep-89  1727	CLT 	visit
Sol says `the Matyasevitch visit has been (temporarily?) aborted as
his visa application didn't make it in time.'

∂20-Sep-89  0342	harnad@clarity.Princeton.EDU 	BBS & APS 
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 20 Sep 89  03:42:13 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA16125; Tue, 19 Sep 89 15:15:05 EDT
Received: by psycho.Princeton.EDU (4.0/1.85)
	id AA12647; Tue, 19 Sep 89 14:49:50 EDT
Date: Tue, 19 Sep 89 14:49:50 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8909191849.AA12647@psycho.Princeton.EDU>
To: harnad@elbereth.rutgers.edu
Subject: BBS & APS

To:  BBS Associates
Re:  Sponsorship by American Psychological Society
    
Please let me know your reactions and suggestions.

----
> Date:   Tue Sep 19 13:43:04 1989
> From: Sandra Scarr <ss9v@mendel.acc.virginia.edu>
> To: Stevan Harnad <harnad@clarity>
> 
> Dear Stevan:
> 
> As you may know, I am the chair of the Publications Committee for the
> new American Psychological Society. We have gotten off to a great start
> with the upcoming publication of Psychological Science (Jan. 1990), and
> with Bill Estes as Editor. APS is now venturing in two directions:
> Starting a second journal, yet to be determined, and adopting journals
> that are not now sponsored by any society. Your journal is one of the
> most luminous of those not now sponsored by a scientific society. There
> are advantages to sponsorship by a society for the Editor, the
> publisher, and the society. I will be happy to discuss them with you at
> your convenience.
> 
> Would you favor sponsorship for Behavioral and Brain Sciences by APS?
> Because Cambridge University Press is the publisher of your journal and
> of Psychological Science, they may be amenable to such an arrangement.
> I will discuss it with Jim Alexander, if you think it would be a good
> idea.
> 
> Please forgive the bad line interference. I look forward to hearing
> from you.
> 
> Sandra Scarr                   Department of Psychology 
> ss9v@virginia.edu              University of Virginia 
> Office: (804)-924-3374         Charlottesville, VA 22903 
> Home: (804)-979-2047           FAX: (804)-924-7185 
> 

To:   Sandra Scarr
From: Stevan Harnad

Sandra:

The idea sounds interesting, and I certainly support the aims of APS.
I would like to know more about what sponsorship by the APS would
entail, however. Three immediate questions arise:

(1) How would this effect BBS's current Associateship of ~5000 (which
grows by about 3-400 annually)? There is currently no fee for becoming
an Associate, and no requirement to subscribe to the journal (although
Associates are eligible for a preferred rate) and I would want to keep
it that way.

(2) BBS's constituency is not limited to psychology; it includes
neuroscience, computer science, philosophy, linguistics, and portions
of biological and social sciences. How would sponsorship by APS affect
BBS's interdisciplinary status?

(3) BBS is also international; almost half of the Associateship (and
probably more than half of the readership) is non-US. How would
sponsorship by APS affect BBS's international status?

I think questions like these can probably be answered satisfactorily,
but I've taken the liberty of branching your letter and my reply to the
current BBS email list (only 750 strong so far, unfortunately, but
growing rapidly as I encourage all Associates to get email addresses
and send them to BBS), in keeping with BBS's policy of allowing the
Associateship to share in decisions about the future course of the
project.

Sincerely,

Stevan Harnad
Editor, BBS
Psychology Department
Princeton University

∂20-Sep-89  0404	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	rudimentry pgm + [HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu: More c/3's]  
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 20 Sep 89  04:04:46 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 422426; 20 Sep 89 07:02:20 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108696; Wed 20-Sep-89 03:56:41 PDT
Date: Wed, 20 Sep 89 03:55 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: rudimentry pgm + [HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu: More c/3's]
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "Rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hpm@rover.ri.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "acad!megalon!rudy@uunet.UU.NET"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "wri-tech@wri.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
File-References: SPA|C:>user>rwg>nbwlife.lisp.newest
Included-msgs: The message of 17 Sep 89 20:49 PDT from HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu,
               The message of 17 Sep 89 20:49 PDT from Dean Hickerson
Message-ID: <19890920105546.6.RWG@TSUNAMI.SPA.Symbolics.COM>

The ref'd file is a marginal improvement on HACKS:LIFE.  You say, e.g,
(string-to-array "
        ...................o........o........o.........o...........
        .................oo........o.o......o.o......oo............
        .ooo.......o.....oo.......oo........o.o......oo.........o.o
        .oo..o.oo.oo.......o........o.......o..........o.......oooo
        ...ooo....o......o...o....oo............o....o...o....oo...
        .o..o.o...o...o.o...oooo..oooooo...oo..o..o.o...oooo..oo...
        o....o....o..o.oo...o.o..o.o...o.o..o.oo...oo...o.o..o.....
        o....o....o..o.oo...o.o..o.o...o.o..o.oo...oo...o.o..o.....
        .o..o.o...o...o.o...oooo..oooooo...oo..o..o.o...oooo..oo...
        ...ooo....o......o...o....oo............o....o...o....oo...
        .oo..o.oo.oo.......o........o.......o..........o.......oooo
        .ooo.......o.....oo.......oo........o.o......oo.........o.o
        .................oo........o.o......o.o......oo............
        ...................o........o........o.........o...........
" 69)
                              ;  Figure 1

This enters the pattern shifted 69 to the right.  Then (bwlife).  For
some reason, the running time is dominated by displaying.  Hence the
commands are

+     = larger dots, slower;
-     = smaller dots, faster;
Space = stop, single step;
p     = run, proceed;
End   = quit.

The file defines globals BW-WINDOW and LIFE-ARRAY, which may be
obnoxious.  If you already have a fairly vanilla window lying
around, you might supply your own definition of GET-BW-WINDOW.

Regarding Figure 1 above, Hickerson says

    Date: Sun, 17 Sep 89 20:49 PDT
    From: Dean Hickerson <HUL%PSUVM.BITNET@CORNELLC.cit.cornell.edu>
    Subject: More c/3's
    To: rwg@yukon.scrc.symbolics.com
    cc: fermat!r@la.tis.com
    
	 Sorry about the previous message.  I accidentally hit the SEND key
    while I was editing.  The complete message follows:

I would be delirious, too.

	 I've searched for some thinner orthogonal spaceships with speed c/3.
    The one in Figure 0 is the smallest I've found.  It has width 10 in the
    phase shown and 12 in its other phases.  It has a tail spark which can
    protect it from many attacks from behind.  In Figure 0 it's about to
    destroy a lightweight spaceship.
    
	    .ooo.......o
	    .oo..o.oo.oo
	    ...ooo....o......o...
	    .o..o.o...o.....o....
	    o....o....o.....o...o
	    o....o....o.....oooo.
	    .o..o.o...o.
	    ...ooo....o.
	    .oo..o.oo.oo
	    .ooo.......o
    
	      Figure 0
    

        Figure 1 shows a horizontal wick being grown at c/3 on the left and
    burning cleanly at c/3 on the right.  Thus, we can build an arbitrarily
    long spaceship with width 14.  The wick has period 28 and reappears in
    generation 1 shifted right 9 units (or 37, 65, ... or left 19, 47, ...).

So in 3 gens, it goes right 27 units = left 1 unit, as it must!  Ergodisia!
You are trashing my reality predicates!  Such behavior is unprecedented!
Did you really ask your program to look for a phase velocity of -9c??
    
    Figure 2 shows an alternate way to connect the spaceship to the left of
    the wick and another way to end it on the right.
    
	    ..................o.........o........o.........o.o
	    .................o.o......oo........o..ooo...o..o.
	    .ooo.......o.....o.o......oo.......oo....o.o.o..oo
	    .oo..o.oo.oo.....o..........o..........o.o........
	    ...ooo....o..........o....o...o....oo.oo..........
	    .o..o.o...o...o.oo..o..o.o...oooo..ooo............
	    o....o....o..o...o.oo...oo...o.o..o.oo............
	    o....o....o..o...o.oo...oo...o.o..o.oo............
	    .o..o.o...o...o.oo..o..o.o...oooo..ooo............
	    ...ooo....o..........o....o...o....oo.oo..........
	    .oo..o.oo.oo.....o..........o..........o.o........
	    .ooo.......o.....o.o......oo.......oo....o.o.o..oo
	    .................o.o......oo........o..ooo...o..o.
	    ..................o.........o........o.........o.o
    
				  Figure 2
    
    The bottom 5 rows and rightmost 9 columns can also be changed to:
    
	    .........        .........        ......o.o
	    o.....o.o        o...o..oo        o...o..o.
	    o.o.o..o.   or   o.o.o..o.   or   o.o.o..oo
	    o...o..oo        o.....o.o        o........
	    .........        .........        .........
    
			     Figure 3
    
	 I've tried unsuccessfully to find a way to make the right end into
    a c/3 fuse which leaves stuff behind, in order to make a puffer.  Or a
    slower than c/3 fuse, to make an ever lengthening spaceship or puffer.
    Unfortunately, every random thing I've tried for the right end turns
    into a 33c/43 fuse which produces 2 blocks, 2 blinkers, and 2 boats every
    43 generations.

The rarity of sub-speedlimit ships is related to strangulation by apron
strings.  For similar reasons, your c/2 ships should be easier to pufferize.

Among the extreme peculiarities of your new devices is the failure to
period-tuple, a strong proclivity of known puffers, even conventional
spaceships, and nonlinear systems generally.  That is, a small window fol-
lowing just the nose of a conventional puffer will see period 2, but succes-
sively larger ones will see period 4, then maybe 20, then maybe 140, as in 

 o o o . . . . . . . . . . . o o o .
 o . . o . . . . . . . . . . o . . o
 o . . . . . . o o o . . . . o . . .
 o . . . . . o . . o . . . . o . . .
 . o . . . . . . . o . . . . . o . .
 . . . . . . . . . . . . . . . . . .  .

It is clearly easier for your program to look for strict periodicities.
The wonder is that it finds them.  That practically guarantees small,
period tupling puffers with your exotic ships as noses, if only we
knew how to search.  Buckingham, where are you?

	 Here's yet another c/3:
    
	    .........o...
	    ........o.o..
	    ........o....
	    .ooo....o...o
	    .oo..o....o.o
	    ...ooo.oo.oo.
	    .o..o.o.ooo..
	    o....o.....o.
	    o....o.....o.
	    .o..o.o.ooo..
	    ...ooo.oo.oo.
	    .oo..o....o.o
	    .ooo....o...o
	    ........o....
	    ........o.o..
	    .........o...
    
	      Figure 4

Ho hum!

∂20-Sep-89  0457	@ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM 	Oops!  And shred that cultural litmus! 
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 20 Sep 89  04:57:35 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 422432; 20 Sep 89 07:56:36 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 108698; Wed 20-Sep-89 04:50:55 PDT
Date: Wed, 20 Sep 89 04:50 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Oops!  And shred that cultural litmus!
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "rcs@la.tis.com"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "hul@psuvm.psu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
    "hpm@rover.ri.cmu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "acad!megalon!rudy@uunet.UU.NET"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
File-References: SPA|C:>user>rwg>life>nbwlife.lisp.newest
Message-ID: <19890920115002.7.RWG@TSUNAMI.SPA.Symbolics.COM>
Character-Type-Mappings: (1 0 (NIL 0) (NIL :ITALIC NIL) "CPTFONTI")
Fonts: CPTFONT, CPTFONTI

That nbwlife file (and bin) is under life>.  Sorry.
    
    Professor Igarashi

(or perhaps Igerashi)
		       tells me that in Japan, gliders are called
    scooters.  And Marco Aiello tells me they're bullets in Italy.

Upon closer reading, Marco retracts this.  The article he saw
introduced the gun (ε1canoneε0) before the glider (an option unavailable to
M. Gardner), and referred metaphorically to the output as ε1ballistiε0-
something-or-other.  But later, the article switched to calling them
ε1aliantiε0, which presumably loses Conway's glide-reflection pun.

∂20-Sep-89  0745	crucible@fernwood.MPK.CA.US 	THE INTERNET CRUCIBLE - Volume 1, Issue 2
Received: from fernwood.MPK.CA.US by SAIL.Stanford.EDU with TCP; 20 Sep 89  07:44:11 PDT
Received: from localhost by fernwood.MPK.CA.US at Wed, 20 Sep 89 07:32:14 -0700.
	(5.61.14/XIDA-1.2.8.34) id AA20970 for jmc@sail.stanford.edu via SMTP
Message-Id: <8909201432.AA20970@fernwood.MPK.CA.US>
To: Distribution:;
Subject: THE INTERNET CRUCIBLE - Volume 1, Issue 2
Date: Wed, 20 Sep 89 07:32:08 PDT
From: crucible@fernwood.mpk.ca.us

THE CRUCIBLE                                                  INTERNET EDITION
September, 1989                                             Volume 1 : Issue 2

   In this issue:
	- Explanation of THE CRUCIBLE moderation policy
	- Corrigendum to THE CRUCIBLE Volume 1, Issue 1
	- Letters in response to THE CRUCIBLE Volume 1, Issue 1
	- The Changing Nature of Managing the Internet

   THE CRUCIBLE is a moderated forum for the discussion of Internet issues.
   Contributions received by the moderator are stripped of all identifying
   headers and signatures and forwarded to a panel of referees.  Materials
   approved for publication will appear in THE CRUCIBLE without attribution.
   This policy encourages consideration of ideas solely on their intrinsic
   merit, free from the influences of authorship, funding sources and
   organizational affiliations.

   THE INTERNET CRUCIBLE is an eleemosynary publication of Geoff Goodfellow.

Mail contributions to:                            crucible@fernwood.mpk.ca.us

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

		  THE INTERNET CRUCIBLE MODERATION POLICY

	THE CRUCIBLE is a refereed periodical irregularly published on the
Internet.  All contributions and editorial comments to THE CRUCIBLE are
reviewed and published without attribution.

	Each contribution is refereed by a range of networking experts from
academia, research, and industry.  As with refereed professional journals,
the referees are responsible for ensuring that a contribution is credible.
Even though one or more of the referees may not agree with the position
taken by the authors of a contribution, the referees may still recommend
that the contribution be published.  If not, the referees make cogent
suggestions as to how the contribution might be improved. 

	Publication without attribution is a time-honored means for
advancing positions solely on the basis of their content.  Unlike
professional journals that exist both to serve the community and contribute
to the authors' reputations, THE CRUCIBLE exists solely to serve the
community.  THE CRUCIBLE moderator, a member of the network community since
1973, feels that the Internet is best served by fostering a forum in which
ideas stand solely on their intrinsic merit, not on the standings of the
authors advancing the ideas.

	In addition to providing a neutral forum in which to examine the
ideas, publication with refereeing but without attribution provides a means
for contributors to express ideas that may be controversial.  It is an
unfortunate but all too common situation that many organizations
(commercial or research) naturally avoid association with controversial
positions.  Even if an author prepares a contribution on personal time and
publishes it without affiliation, the author's employer, funding source,
(and parts of the community) will most likely view the contribution as
being associated with the author's organization.  Thus, many potentially
controversial contributions of merit are never submitted.

	THE CRUCIBLE, by publishing without attribution, prevents prejudice
towards contributions on the basis of authors' standings or their 
affiliations, and encourages contributors to speak freely, without
organizational entanglements or jeopardizing funding sources.  THE CRUCIBLE
relies on a wide cross-section of referees to filter contributions that are
not of a meritorious nature.

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

			   CRUCIBLE Corrigendum

	The contributors of THE CRUCIBLE Volume 1, Issue 1, "A Critical
Analysis of the Internet Management Situation: The Internet Lacks
Governance" incorrectly referred to Mr. Robert Braden, IAB Executive
Director as "Dr."  The contributors regret the error.

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

	   LETTERS IN RESPONSE TO THE CRUCIBLE VOLUME 1, ISSUE 1

Sir:

	Who was the author of this article?  The only name I noticed was
Goodfellow, but he is listed as moderator.  I have no objection to the
views.  Indeed they are sensible.  But I do have objections to anonymous
editorials.

Yours etc.,

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

Sir:

	Good job.  This showed up in my mailbox while I was trying to 
write something making many of the same points.  You outdid me.

Yours etc.,

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

Sir:

	Congratulations on a great article!  While I disagree with some
specific points you made, the vast majority of what you said is right on
target, and badly needed to be said.  (As a case in point, I've been unable
to reach SRI-NIC.ARPA (aka NIC.DDN.MIL) for a week now, and I can't even
get an email response from our beloved regional NSF net.)

Yours etc.,

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

Sir:

	Fascinating article.  Too bad the author did not have the guts to
sign his/her name.  My only nit-pick is to reply to this:
   > Meanwhile, the outside world goes on about developing economically viable
   > and efficient networking technology without the benefit of direct
   > participation on the part of the Internet.

	With "oh really?  What?"

Yours etc.,

[The contributors respond:

   At the beginning of the article we noted technologies such as PDNs, UUCP
   and CSNET dial-up networks.  While these lack the glamor of IP-style
   networking on the extended Internet, they nonetheless are economically
   viable.  One might argue that OSI is producing technology roughly
   comparable to the Internet suite, but there is very little technology
   transfer in this regard.  Considering that OSI is standardized in The
   Open, we find it difficult to blame the OSI proponents for this lack of
   cross-over.]

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

Sir:

	I dislike intensely receiving a controversial article by email when
I cannot tell either who wrote the article, or to whom it was distributed.

Yours etc.,

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

Sir:

	Good article.  Many valid points were raised.  I, like many others,
suffer from the problems pointed out in this article.  I see how things
could be better and am frustrated when it takes so long for things to
improve.  It will be interesting to see what kind of responses you receive.

Yours etc.,

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

Sir:

You stated:
   > While everyone in the market says they want OSI, anyone
   > planning on getting any work done today buys Internet technology.

	We're getting work done today using OSI technology, on Intel's
OpenNET network. It's got superior UNIX file-system semantics to either NFS
or RFS. We'd like to go TCP/IP, since it's got more *real* standards, but
we certainly don't plan to take a step backwards to do so.

	So, for us, the situation is the reverse...

Yours etc.,

[The contributors respond:

   In limited, pilot and laboratory settings, OSI may provide solutions today
   that Internet technology does not.  It has been our experience that this is
   the exception and not the rule.  The OSI file service is called FTAM, not
   OpenNet.  OpenNet, while based on the OSI transport service, is nonetheless
   a proprietary vendor solution as it has not been standardized in The Open.
   Presumably one could easily host OpenNet technology on top of TCP, and
   outside of the increase in performance, robustness, and throughput, you'd
   never notice the difference!  All humor aside, it is important to
   distinguish between OSI standards and things resting on top of OSI
   standards. ]

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

Sir:

	A friend forwarded Vol 1 No 1 of CRUCIBLE to me recently.  I was
feeling good about the text all through the preamble, saying to myself that
this sounded like a useful contribution to an area that could benefit from
"critical analysis" especially since you advertised that it was going to be
refereed.

	However, almost none of the body of this article has merit.  It is
a rehash of uninformed flaming. Neither the article nor any of the missing
refereed commentary shows any indication of knowledge of a great deal of
hard work that has been going on for the last three or four years to
design, promote and implement an improved national research and education
network based on the strengths of the Internet.  There is no reference to
at least a half dozen major studies of the requirements for deployment of
advanced technology in an upgraded research net.  Nor even one word about
two annual conferences on the need for a national net that were held in
Washington in 88 and 89 that have helped develop the NREN plan and the
necessary federal legislation to support it.

	I was particularly annoyed by the pernicious foolishness about
competitive alternatives to the Internet.  Is whoever is the author of this
piece seriously proposing that the nation should fund TWO advanced networks
on the theory that the costs will be lower overall after they compete with
each other?  Or do I have it wrong, and the worthy scholar of your piece
doesn't care whether we use world class networking in our allegedly world
class national research enterprise but just wants cheap reliable email?  If
so, we already have it in the form of BITNET which reaches millions of
individuals worldwide for pennies a message and is 100% supported by
member/user fees.

	I've spent almost all of my career in academia, some twenty-four
years at last count.  In that arena, if you have the courage of your
convictions, you put your name on them. I don't see any excuse for unsigned
articles in CRUCIBLE.  This isn't an AIDS hotline we are talking about.

	I'd like to see CRUCIBLE succeed and have a positive influence on
the shaping of public policy for the NREN that will be taking place over
the next couple of years.  But, candidly, you're off to a bad start.

Yours etc.,

[The contributors respond:

   It is important to appreciate that the thrust of the contribution had
   nothing whatsoever to do with the NREN.  It was neither a criticism of the
   NREN, and it was certainly not a covert appeal for funding an alternative
   to the NREN.  The contribution was about the Internet.  This is a network
   which is operational today, which many thousands of people depend on as a
   normal part of their work environment.

   While we applaud the hard work, studies, conferences, etc., that
   demonstrate the commitment to fund and implement the NREN, these are, quite
   bluntly, immaterial to the argument at hand.  There is a notable exception
   however: if the NREN is to be based on Internet technology, then we hope
   that the NREN management learns from the lessons of the existing Internet
   and is sensitive to economic issues in the technology it uses.  Recent
   history in numerous other industries demonstrates that, despite broad
   agendas and good intentions on the part of government, commercial
   forces are often vital to producing vibrant, useful services.
 
   The simple fact is that economic models work and others, such as the
   current Internet model, don't work.  For example, there have been massive
   innovations in dial-up modem technology in recent years.  This is solely
   because the users of dial-up modems, by and large, have to pay for
   the phone calls they make. If dial tone and connect time were free, 
   we'd likely still be using 1200bps today!  There is no correspondent 
   accountability in Internet technology, and half-attempts to provide this,
   such as policy-based routing, are fraught with potentially disastrous
   performance side-effects.

   Finally, was this contribution really "a rehash of uninformed flaming"?
   Can anyone ever recall seeing the Internet structure being taken to task 
   in a public forum?  We think not.  This is the first time anyone has ever
   bothered to fire a shot off the bow of the old boy's club that's been
   non-managing Internet technology for the last few years.]


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

Sir:

One minor quibble with an otherwise fine and level-headed document:

      This bug flies in the face of the Internet parable
      "be generous in what you accept and rigorous in what you send".
   
Perhaps you meant to say "rule of thumb", or "proverb"?

Yours etc.,

[The contributors respond:

   Quite right.  We stand corrected.]

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

Sir:

	All very well said and well put.  I've just come back from a week
on the West Coast trying to use the Internet for real work while I was
away, and separate problems at BARRNET and JVNC/NEARNET made things
literally unusable.  By the time I yelled long enough at different people,
things were fixed--by Friday, the end of the week.

	The balkanization of the Internet into regional nets has been like
the breakup of AT&T, only worse (because of their isolation from economic
pressures and because "service" doesn't matter.)

	Will there be a BOF on your topics here at INTEROP?

Yours etc.,

[The contributors respond:

   No BOF is planned, but it would be interesting to have an open forum for a
   face-to-face meeting between the IAB, NSF regional network directors and
   an Internet constituency.]

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

	       The Changing Nature of Managing the Internet:
			 A Paradox in Governance


				ABSTRACT

The development of Internet technology was well-served by a directed agenda
sponsored by the US Government.  However, today's development of Internet
technology is largely stagnant because the technology is providing useful
service for many of the research and development communities needs.
Although several developmental efforts are underway, seemingly not much
progress is being made.  The ill-defined distinction between Internet
research, engineering, and development has led to additional confusion.

The first issue of THE INTERNET CRUCIBLE dealt with Internet technical and
accountability failures.  This second issue is complementary in that it
examines the paradoxes and failures inherent in the current Internet
management structure.  A new structure for Internet management is proposed.


	    A Brief History of Internet Technical Management

	As a part of its charter in the 70's, DARPA (then known as ARPA)
was responsible for funding high-risk, high-potential payoff research.
DARPA's guiding rule was fairly simple: ideas were researched, proven, and
then deployed.  In order to advance DARPA's ambitious agenda in networking,
the government recruited far-sighted individuals with an interest in
networking, each with personal research agendas in networking and
communications.  The DARPA "program managers" then oversaw the technical
direction taken by the researchers which produced, among other things,
Internet technology.  Research efforts in areas of networking continue to
be funded and directed by program managers DARPA and NSF independent of the
existing Internet technology base.

	As Internet technology passed from idea, to research, to deployed
technology, there were two side-effects:

	1) Internet technology ceased to be high-risk research--it became
	   engineering; and,

	2) the networking agenda became less of a priority research 
	   item as it became developed, proven, operational technology.

	Several production internets now exist, providing useful service to
their respective communities.  Problems with Internet technology today are
largely second-order problems, ones which are easily solvable by
responsible engineering and modest evolutions in the technology.

	A cogent argument could be made that any networking "research"
being done these days must be substantively distanced from current
technology.  Using this perhaps controversial definition of "research",
topics such as OSI, Message Handling Systems and Directory Service
applications, while important to production networks in the short- and
medium-term, are "development" and not "research".

	This leads us to note that the nature of Internet technical
direction has changed: it is now one of managing engineering disciplines,
not one of managing research.  It is not clear that the management
structure used to guide the technical direction of Internet technology
reflects this change.  To be sure, there have been recent changes in the
management structure (i.e., the recent IAB reorganization and the creation
of the IRTF and IESG), but these changes appear to be superficial only--the
old thinking, the old procedures, and most importantly, the old
infrastructure, are all still in place.  The model, we argue, continues to
be based on the notion that Internet technology is still a research project.


	       The Current Internet Management Structure

	Arguably the top-level management entity is the newly re-organized
Internet Activities Board (IAB).  The IAB consists of the chairs of two
task forces (one responsible for engineering, the other for research),
liaisons to other organizations, and administrative personnel.  The IAB's
current charter consists of managing the RFC and Internet standards
process, managing the engineering and research task forces, performing
strategic planning for the Internet, and providing liaison to other
organizations.  The IAB is a "closed" organization--its meetings are not
open to the public, and neither agendas nor minutes are openly available.

	The Internet Engineering Task Force (IETF) of the IAB is chartered
to solve short-term engineering problems in Internet technology.  It is a
open, voluntary organization with over 20 working groups.  Unfortunately, the
IETF has not been particularly successful in producing short-term solutions
(e.g., it has taken the IETF well over a year to not yet produce RFCs for
either a Point-To-Point Serial Line IP or Network Management enhancements).
New to the IETF is the Internet Engineering Steering Group (IESG) which is
tasked to govern the IETF.  This too is a closed organization, serving at
the pleasure of the chair of the IETF.

	The Internet Research Task Force (IRTF) is chartered to promote
research in networking that will presumably lead to producing the
technology the IETF will need.  The IRTF also has a steering group, the
IRSG.  The IRTF consists of several "research groups" which, as of this
writing, roughly correspond to the various research task forces found on
the IAB prior to re-structuring.  The IRTF is a closed organization.

	It should be noted that although the recent IAB reorganization
changed some titles, the personnel and infrastructure are largely unchanged.
The same relatively small closed community continues to make decisions.


	   Paradoxes in the Current Internet Management Structure

Consider three possible scenarios:

Scenario 1: The IAB

	The IAB, by declaration, determines the direction of Internet
technology, primarily through the form of managing the Internet standards
process, and thus has a wide impact on uses and development of Internet
technology in government, academia and industry, along with significant
commercial impact.  However, the IAB is, as noted earlier, a closed
organization.  Further, there is no disclosure of interests among the 
IAB membership.

	For the sake of argument, suppose one or more members of the IAB
were employees of a commercial concern.  Further, suppose that these
commercial concerns were directly tied to Internet technology.  For
example, one of of these concerns might be a vendor of Internet technology,
or another might be a supplier to such vendors.  As such, the actions of
these IAB members, who are also commercial employees, must be under careful
scrutiny.

	Let us carry this hypothetical scenario a bit further: suppose the
IAB were considering elevating a new protocol to some official standing.
If parts of the vendor community (and in particular one of the commercial
concerns sponsoring an IAB member) stood to benefit competitively or
fiscally from such a decision, then this situation might well place the
interests of those IAB members at cross-purposes with the interests of the
Internet community as a whole, and the U.S. Government in particular.  At
the very least, the IAB membership and presumably the sponsors of that
membership, have access to "inside information" which has significant
commercial impact.

	To impress upon the reader that this is not a theoretical
observation, it must be emphasized that this scenario is drawn on fact--it 
is disturbingly parallel to an actual incident which recently occurred.  It 
is not, however, the purpose of this contribution to prejudge the motives of
the individuals involved.  Rather, we note that appearance of interests is 
at cross-purposes, regardless of the intent.

Scenario 2: The IESG

	The newly-formed IESG, by declaration, develops technology for use
in the Internet.  The same hypothetical argument can be used here: the
members of the IESG are each responsible for a particular area of Internet
technology, e.g., network management.  Again, if the person filling this
role is employed by a commercial concern, then interests may be at
cross-purposes.  In particular, since theoretically the IESG deals with
technology at a less mature stage, entire engineering approaches might be
unduly advanced or hindered depending on potential market advantage or threat.

	Fortunately, the IETF is an open organization, as such interests at
cross-purposes can be more readily identified.  However, since the IETF
takes its direction from the IESG, the engineering aspect of Internet
technology remains subject to being directed by IESG members who could be
at cross-purposes with the interests of the Internet community.

Scenario 3: The IRSG

	The IRSG, by declaration, advances research into Internet
technology.  For the sake of argument, suppose that the IRSG proposes that
particular areas of research be funded.  Since the IRSG is a closed by-
invitation-only organization, its membership is likely to appear to be
acting in a self-interested fashion, since they receive early information
about areas which might be funded, and are in a substantially advantaged
position to write proposals for that funding.  Although the IRSG might
appear to be acting as an advisory board, an outsider might make the
argument that the IRSG both directs the government in funding direction,
and then makes proposals to take advantage of the advice given.

	Consider the effort to produce a White Pages service for the
Internet (a service that, among other things, is used to obtain mailbox and
other information about users in the network).  In February of 1989 the
Federal Research Internet Coordinating Committee (FRICC), an informal group
of government program managers who fund portions of the Internet, tasked an
ad hoc committee to produce a plan for a White Pages service in the
Internet.  Although the meeting was held in February, RFC1107, a report
describing the outcome of the meeting was delayed until July.  The reason
was that the organization tasked with writing the report might also receive
funds as a result of the report being implemented.  A compromise of
conscience was finally reached by significantly "watering down" the report
so that it contained no concrete proposals, instead making a reference that
"strong funding and encouragement" was needed.  (Footnote: this post-meeting
editing is not as sinister as it might seem--the meeting called was
"politically" correct in the sense that it invited over 30 attendees each
with a differing agenda.  In essence, the two-day meeting rapidly devolved
into scantily-clad turf wars rather than practical technical discussion.)

	In general, all the scenarios arise from a simple premise: people
who give advice to the government should not profit from giving that advice
unless the collection of those people is operating under a "balanced
conflict of interest".

	It must be emphasized that the purpose of this contribution is NOT
to suggest that the current structure is corrupt or unduly self-interested.
Rather, our purpose is to demonstrate that the current structure is fraught
with peril--unintentional effects, smacking of impropriety, may result from
careless action or inaction.  Further, it should be noted that we live in a
world of "retroactive ethics"--actions are not judged by standards of
ethics which existed when the actions took place, rather any action is
judged by whatever standards of ethics are currently in place.  Recent
history indicates that any appearance of impropriety is effectively proof 
of that impropriety.


			    Recommendations

	The early DARPA-sponsored work in networking, by separating program
manager from researcher, was able to avoid these dilemmas.  Unfortunately,
Internet technology and the Internet are now larger than any single
organization.  What then, can be done to avoid this problem and still
further Internet technology?  This contribution suggests that Internet
technology has matured to the point where it is time to provide management
and guidance from a wide-cross section of users and providers.  That is,
while the seminal contributions of researchers made Internet technology
possible, a different paradigm is necessary to manage today's Internet
technology, which has effectively become an engineering and maintenance
problem.

	This contribution suggests the formation of an Internet Policy
Forum (IPF) appointed by the Federal Coordinating Council on Science,
Engineering and Technology (FCCSET).  The FCCSET is a formal body of the
Federal Government chartered with getting the various agencies with large
investments in information technology to coordinate with each other.  The
IPF charter, under the auspices of the FCCSET, would be to represent the
users of Internet technology from all segments of society.  

	Although the structure of the IPF would be largely similar to the
current IAB, there are three fundamental differences:

	- the IPF would meet openly, with published agendas and minutes,

	- the membership of the IPF would primarily be representatives from
the various communities which *use* Internet technology (e.g., some
representatives would be from the Internet regional networks); and,

	- funding for the IPF membership would be solely from the sponsors
of each member; as such, interests, and in particular research and
commercial interests, of each member would be fully disclosed.

It should be noted that each of these tenets circumvents weaknesses in
the current IAB structure:

	- because the IPF would meet openly, it would be open to scrutiny
from a large, vibrant community.  Under this model, the IPF would avoid
any appearance of being "an old boy's club".  

	- because the IPF membership would be drawn from the user and
vendor communities, the members would be motivated towards solving problems
in Internet technology.  Some members of the IPF might be "industry
experts" in Internet technology, others might be from the regional network
management, and so on.  A weakness in the current IAB structure is that the
majority of its members are sponsored by a small group of concerns; if
members are inclined to think about only those things they are funded to
think about, then it is important to draw from a larger collection of
concerns.  IPF members would be, by definition, accountable for and
responsible to, the users of the network.  In contrast, it can be
persuasively argued that the IAB membership is, by and large, not
responsible to, nor do they provide representation for *any* operational
community.  Unlike the IAB, the IPF membership would have a stake in the
continued success of Internet technology.

	- because the IPF membership would be required to fully disclose
their interests, it would be possible to achieve a "balanced conflict of
interest".  Since it is naive to think that interests at cross-purposes do
not exist, it is important to foster an environment in which such things
are entirely above board.  It should be noted that all professionals, be
they sponsored directly by the U.S. Government, by government-sponsored
research organizations, by academia, or by commercial concerns, all have
self-interests.  All such interests must be openly declared.

	The IPF itself would consist of two kinds of members: voting
members who would set policy, and non-voting members who would provide
liaison to other accredited organizations.  The IPF would be responsible
for managing the RFC process and the voting membership would be responsible
for advancing standards in Internet technology.

	Unlike the IAB, there would be no task forces or steering groups
under the IPF.  However, one IPF position would be that of "Internet
Technical Director".  This position would be responsible for coordinating
"development" of Internet technology.  "Research" efforts in networking
technology would be independent of the IPF.  The existing IETF would be
restructured to report to the Technical Manager under the name of "Internet
Technical Directorate".  Like the IETF, the ITD would be an open
organization, although there would be no IESG.

	It should be noted that the coupling of IPF and ITD is different
than the IAB/IETF relationship.  In particular, the membership of both the
IPF and the ITD would come from the same wide range of sponsors.  As such,
the ITD membership would be responsible for producing workable technology
for the IPF, and the IPF would be responsible for providing effective
direction to the ITD.  This is entirely unlike the current IAB/IETF
relationship in which there is little, if any, relationship between the
membership of the two bodies.  Realistically, the IETF has little
accountability to the IAB, as evidenced by the lack of IETF progress in
producing technical solutions to short-term problems over the last few
years.  Similarly, the IAB has little credibility with the IETF membership
at large given the lack of IAB direction, and the somewhat inconsistent
policy actions of the IAB (e.g., network management protocol standardization).


			      Conclusions

	In retrospect, it should be noted that Internet technology and
management have been relatively stagnant over the last three years.  In
particular, the Internet Management structure has been ineffective at
solving problems.  The first issue of THE INTERNET CRUCIBLE analyzed many
of the technical and economic failures in this regard.

	While there are many possible causes for this stasis, it does seem
likely that the continued orientation towards concentrating power in a
small, inbred, cloistered community has been a significant factor in this
regard.  The recent re-organization of the IAB is little more than
repackaging--no fundamental change in direction or motivation has occurred.
This issue of THE INTERNET CRUCIBLE has argued that the existing management
structure is the cause for many of the problems currently being seen on
the Internet today.

But, we have also seen how Internet technology has matured in the market
and is providing useful service.  It is now time to engage a broad
community to openly guide Internet technology in its stable phase.

∂20-Sep-89  1052	clm@prefect.es.llnl.gov 	greetins  
Received: from prefect.es.llnl.gov by SAIL.Stanford.EDU with TCP; 20 Sep 89  10:52:41 PDT
Received: by prefect.es.llnl.gov (3.2/1.26)
	id AA11502; Wed, 20 Sep 89 10:57:21 PDT
Date: Wed, 20 Sep 89 10:57:21 PDT
From: clm@prefect.es.llnl.gov (Cindy L. Mason)
Message-Id: <8909201757.AA11502@prefect.es.llnl.gov>
To: mccarthy@sail.stanford.edu
Subject: greetins

howdy o east bay wanderer!

just wanted to check in with you, and let ypou know i'm still
alive and kicking.  sorry, this terminal is weirdness for me,
and there's no clear way to do a silly backspace!@  in any case,
the dai workshop was great!  the worked received a prize, and
was well received etc, however, i have spent the last two days
feeling rather sick...i am now in la (rich has been taking care
of me) so it seems i won't be back in town before you take off..

i'm sorry these schedules leave us kaput about lasagne...we shall
try again when you return...have a safe journey, and carry a towel.

signing off,

trillion

∂20-Sep-89  1324	VAL 	Coming to Austin    
 ∂20-Sep-89  1223	bledsoe@cs.utexas.edu 	Coming to Austin 
Received: from cs.utexas.edu by SAIL.Stanford.EDU with TCP; 20 Sep 89  12:22:54 PDT
Received: from alfalfa.cs.utexas.edu by cs.utexas.edu (5.59/1.42)
	id AA20235; Wed, 20 Sep 89 14:23:34 CDT
Posted-Date: Wed, 20 Sep 1989 14:18:39 CDT
Received: by alfalfa.cs.utexas.edu (5.59/1.4-Client)
	id AA28412; Wed, 20 Sep 89 14:18:43 CDT
Date: Wed, 20 Sep 1989 14:18:39 CDT
From: Woody Bledsoe <bledsoe@cs.utexas.edu>
To: val@sail.stanford.edu
Cc: bledsoe@cs.utexas.edu
Subject: Coming to Austin
Message-Id: <CMM.0.86.622322319.bledsoe@alfalfa.cs.utexas.edu>

Vladimir,

This is a follow up on the conversation that we had at IJCAI-89 in Detroit,
that you might consider an offer from UT Austin, if offered.  

There is certainly a lot of support for that idea here, expecially with
people like Bob Boyer and myself.  But that is not all.  There is also
a general feeling that such a thing might be supported when they get 
better acquainted with you and your work.  

So a possibility would be for you to visit Austin in the next couple of 
months, to give a talk and get better acquainted with the faculty here.  
Also, as I mentioned in Detroit, we would like an updated version of your
vita.  

If this sounds good to you please let me know when you might come (at our
expense of course).  And eventually, the title of a talk.  I think that
the talk should be slanted to give an idea of your main interests, and not
so much to emphacize you latest research results (or both).  

Best regards,

Woody

∂21-Sep-89  0053	weening@Gang-of-Four.Stanford.EDU 	ClariNet  
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Sep 89  00:53:36 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.61/25-eef) id AA23977; Thu, 21 Sep 89 00:53:18 -0700
Date: Thu, 21 Sep 89 00:53:18 -0700
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8909210753.AA23977@Gang-of-Four.Stanford.EDU>
To: me@sail, jmc@sail, les@sail, kolk@smiley
Subject: ClariNet

From: brad@looking.on.ca (Brad Templeton)
Newsgroups: biz.clarinet.sample
Subject: What biz.clarinet.sample is all about
Date: 21 Sep 89 04:46:39 GMT
Reply-To: info@clarinet.com
Organization: Looking Glass Software Limited, Waterloo ON
Class: news, original

ClariNet, for those who don't read comp.newprod and news.admin, is a new
venture that does Electronic Publishing to USENET sites.  One of the
things we publish is a section of the UPI newswire.  We feed live
news out to UUNET and other sites.

We have created this group to give out certain kinds of free samples that
aren't appropriate for the main USENET hierarchies.  In this case, since
our service started only recently and the baseball season is almost over,
we have decided to give away our baseball newsgroup in this group until
the end of the pennant races.

So enjoy, and if you like, tell your friends and others on USENET.

The baseball group is only 1 of 100 groups of UPI information.  We have
a pretty comprehensive line of general news, business news, sports and
news on technological industries.  We also have computer industry news.

For full details write to info@clarinet.com or phone 519-884-7473 or
800-265-2782.  You can get a free 2 week trial subscription if you like,
and it can even be fed over the internet, at least for now.

If you are not a ball fan, watch this group for future samples of other
kinds of news.
-- 
Brad Templeton, Looking Glass Software Ltd.  --  Waterloo, Ontario 519/884-7473

∂21-Sep-89  0700	JMC  
Bunzel

∂21-Sep-89  0700	JMC  
Leora

∂21-Sep-89  0730	harnad@clarity.Princeton.EDU 	APS & BBS: Responses
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 21 Sep 89  07:30:46 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA09227; Tue, 19 Sep 89 23:07:38 EDT
Received: by psycho.Princeton.EDU (4.0/1.85)
	id AA13318; Tue, 19 Sep 89 22:48:29 EDT
Date: Tue, 19 Sep 89 22:48:29 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8909200248.AA13318@psycho.Princeton.EDU>
To: harnad@elbereth.rutgers.edu
Subject: APS & BBS: Responses

To: BBS Associates

Here is further discussion from APS and the first 4 responses from BBS
Associates. Please continue to forward your views.
Stevan Harnad
----
From: Sandra Scarr <ss9v@mendel.acc.virginia.edu>

Dear Stevan:

I can't imagine that APS would want to infringe on your associateship
in any way. This is a great strength of the journal that you have
developed over the years. The reduced fee to the associates could be
honored in any arrangement with APS. The number of associates is a
great feature of the journal you have built. You have many, many
admirers for both the character of the journal and for the wide
scientific following you have built. APS would want to strengthen, not
interfere with the journal's success.

The interdiscipliary focus of the journal is a strength for APS, which
seeks to build bridges to adjacent disciplines (see promotional
materials and talk with Jim McGaugh, President). The promotional
materials for Psychological Science stress its connections to adjacent
behavioral and brain sciences. (please forgive my typing -- I am not a
bad speller, just a bad typist). Psychological Science will be a
mainstream journal across all areas of psychology and related
disciplines. I see this as a major statement of the inentions of APS to
serve the publishing needs of scientific psychology, just as you serve
such needs.

The international character of BBS is a great plus for APS, as for any
North American journal. I cannot see a problem for APS, which will be a
major sponsor of international psychology.

Really, the major issues will come down to ownership of the journal,
guarantees of editorial boards and quality. It is in your interest to
state what terms you would like to have as Editor. Do you want to stay
for 5 yars, for 6, for 7, for 10, forever? For Cambridge University
Press there is the comfort that APS will not start a similar journal
and that the large membership of APS (>7,000) will be available for
subscribership. This is also a membership with access to institutional
subscribeships from which publishers make most of their profits (sorry,
CUP does not make profits; they have "returns".)

Please let me know what the BBS e-mail list thinks of this proposal.
It is awesome to put such a proposal to vote. By the way, could you tell them
that they can join APS for $75 and receive Psychological Science, the Observer,
and the Employment Bulletin. Applications for membership can be obtained from 
Dr. Alan Kraut, Executive Officer, APS, 100 North Carolin Ave., Suite 1, 
Washington, DC 200003. Or contact me by e-mail. I will recruit, too.

If you think that we can talk further (I hope so!) please contact me by
e-mail, or call next week (804-979-2047 or 804-924-3374). APS is
looking to adopt certain high quality journals that are not currently
afiliated with other societies. BBS is the premier example of such
journals. I think you are the tops!

Your loyal associate,

  *  Sandra Scarr                  Department of Psychology *
  *  ss9v@virginia.edu             University of Virginia *
  *  Office: (804)-924-3374        Charlottesville, VA 22903 *
  *  Home: (804)-979-2047          FAX: (804)-924-7185 *

-----
(1) Forget the APS -- BBS is far too broad a journal to be brought under the
wing of a single parochial professional society. It's a deep mystery
what benefits will accrue to BBS; less so what to the APS.
---
(2) My initial reaction is to take a wait and see attitude.  Will APS
really be as successful/powerful as the old (or even current) APA?
And, as you observe, BBS's readership is not limited to cog psych,
but the full range of cog sci.
---
(3) My reaction is: At the moment, BBS looks more stable than APS.
Unless you get something very valuable from them, I would stall
for time if I were you.
---
(4) While I am a member of APS, and I support their agenda, I'm not sure I can
see any advantages to BBS in Scarr's proposal.  The problems you raise are
real ones, and they would need to be resolved before BBS should allow itself
to be "adopted" by APS.  There is, as you know, a future confrontation
brewing concerning the fate of the APA journals -- they are money-makers
for APA, but mostly supported by the efforts of people whose allegiance
is transferring to APS.  Does BBS need to get embroiled in this kind of
madness??
I really appreciate, by the way, your open approach to the running of BBS.
---

∂21-Sep-89  1223	njacobs@vax.darpa.mil 	Technical Reports
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Sep 89  12:23:37 PDT
Received: from sun41.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA09215>; Thu, 21 Sep 89 15:18:42 -0400
Posted-Date: Thu 21 Sep 89 15:25:30-EDT
Received: by sun41.darpa.mil (4.1/5.51)
	id AA11522; Thu, 21 Sep 89 15:25:32 EDT
Date: Thu 21 Sep 89 15:25:30-EDT
From: Nicole L. Fields <NJACOBS@DARPA.MIL>
Subject: Technical Reports
To: jmc@sail.stanford.edu, mps@sail.stanford.edu
Cc: njacobs@vax.darpa.mil
Message-Id: <622409130.0.NJACOBS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>

Gentlemen,

This message is to remind you that your project summaries are needed.
I sent out a message quite some time ago requesting this information
and got very little response.  If these reports are not received,
Brian cannot prepare your incremental funding for FY90.  If you need 
any further information or have any questions, please contact Nicole
Jacobs at 202/694-5800 or njacobs@darpa.mil.

To all DARPA/ISTO Distributed Parallel Systems and Secure/Survivable Nets PI's:

	ANNUAL TECHNICAL REPORTS

This is the time of year when we require information from contractors
concerning accomplishments for the current fiscal year (FY) and
objectives for the next FY.  This information is used to assist us in
planning incremental funding for those efforts that are expected to
continue into the next FY (which begins 1 Oct 89).  This information
will also be used to brief the incoming Director of ISTO.

WHAT IS REQUIRED.  This year, we expect that only two email reports
will be solicited:
	(1) A single annual technical summary.
	(2) Financial summaries.

NOTE: These email reports are in addition to the reports required as
contractual deliverables, and this informal request does not waive any
contractual requirement.

TECHNICAL SUMMARIES.  The technical summaries provide us with an
up-to-date view of the state of activity in our community.  The
challenge is to be concise yet substantive.  Responses are needed by
the morning of 25 September.

PLEASE ADHERE to the format below.  Specifically, (1) do not include
markup commands from a text processor (except in formulas that are
especially complex), and (2) do not use leading indentation or any
other extra embedded horizontal whitespace.  I suggest grabbing this
text with a text editor and filling in the blanks.

Here is what is required:
================================================================

(1) BASIC DATA.

(1.a) DARPA/ISTO project code, ARPA Order number, agent, contract number.

Example: AA, 1111, SPAWAR, N0037-C-0004
The first of these is the "task code" you were just sent.  (Call
Nicole Jacobs if you don't receive this.)

(1.b) Institution.

The institution contracting with the government.

(1.c) Project title. 

One line project title.

(1.d) Project mailbox.  

An internet address for official project email.  This should be active
even when PIs are travelling.  At worst, we will accept a list of
addresses. 

(1.e) PIs. 

For each: name, phone numbers, email address.

(2) PROJECT SUMMARY.

A short paragraph (approximately four sentences) outlining the
specific need for your effort, the opportunity it represents, and the
results to be delivered.  Your summary should indicate the expected
impact of the project, i.e., how future technologies will be different
because of this investment by DARPA.

(3) APPROACH.

A short paragraph describing the overall approach taken to research.
That is, what actual steps will be taken to achieve research goals?

(4) ACCOMPLISHMENTS. 

Two to four concrete accomplishments for fy89.  Emphasize technical
results with externally recognizable impact.  If you achieved a major
milestone of broad community impact, indicate so and describe it.
(There should be such an event every year or two.)

(5) OBJECTIVES.

Two to four for fy90, each as a short (two to three sentence)
paragraph.  Emphasize both technical results and impact.  Be specific.

(6) TECHNOLOGY TRANSFER progress and plans.  

(6.a) Customers. 

Customers for your results, whether the results be theorems or
software components or software interface definitions.  (Include
potential users outside of your research group with whom you have had
discussions as well as actual current users.  Include both research
groups and development groups.)  This and the other responses in this
section should be only two or three sentences each.

(6.b) Interfaces and consensus.

Opportunities pursued for accelerating community consensus, where
needed and where possible, for systems interface definitions that your
work creates, contributes to, and/or depends on.  (Any producer of a
large system prototype should address this issue.  At the research
prototype stage, efforts involving collaboration among research groups
to agree on component interfaces should be described.)

(6.c) Sources of technology.

Major producers of technology that you rely upon, including technical
results, interface definitions, and systems components.  (Do not
mention common commercial components.  You might mention Mach or the
Boyer-Moore theorem prover, but you wouldn't mention DEC or Sun
workstations.)

(7) OTHER INFORMATION.

(7.a) Major personnel changes.   

New hires, departures, etc.

(7.b) Major recent publications, honors, etc.

Send us copies of important publications once in a while.

================================================================

WHAT TO DO.

One report should be sent for each independently funded effort (i.e.,
each project code).  If you receive multiple copies of this message, it
is probably because you are involved with multiple funded efforts.  If
there is any question about what constitutes an "independently funded
effort," please call Nicole or me.

Send reports by email directly to Nicole Jacobs at DARPA:
	njacobs@vax.darpa.mil		(202)694-5800

If you do not receive an acknowledgement within a week, call Nicole to
ensure she has received your message.

Follow the format guidance for your responses.

MAILING ADDRESSES.  Help us keep our mailing list current.  We often
send official correspondence by email.  Again, if there is a question,
call Nicole and check to see what address we have on file.

Deadline: Please respond by the morning of 25 September.  Early responses
are requested.  If you MUST respond late, call us first.  Thanks.

Finally: Many of your contracts require regular reports to be sent to
the contracting agent and to certain parts of DARPA.  Please also send
one copy of the reports directly to me at DARPA.

Thanks,
				Brian

-------

-------

∂21-Sep-89  1350	VAL 	re: Technical Reports    
[In reply to message sent Thu 21 Sep 89 15:25:30-EDT.]

This is addressed "to all DARPA/ISTO Distributed Parallel Systems and
Secure/Survivable Nets PI's", so I think it was sent to us by mistake.
A few weeks ago we responded to a similar request from Steve Cross.

∂21-Sep-89  1547	weening@Gang-of-Four.Stanford.EDU 	Comparison of Sun and DEC machines 
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Sep 89  15:47:46 PDT
Received:  by Gang-of-Four.Stanford.EDU (5.61/25-eef) id AA26635; Thu, 21 Sep 89 15:47:27 -0700
Date: Thu, 21 Sep 89 15:47:27 -0700
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8909212247.AA26635@Gang-of-Four.Stanford.EDU>
To: clt@sail, jmc@sail, tom@polya
Subject: Comparison of Sun and DEC machines

Below is a chart I made up based on the information we have gotten.
The bottom line shows list prices, not counting maintenance, because
the discounts and maintenance options are too confusing for me to work
out right now.  The discounts seem roughly the same (DEC 35%, Sun 38%,
both with various additional discounts).

My own impressions are:

- The systems are roughly comparable in performance, with DEC slightly
  more powerful and more expandable.

- Sun seems better in software and in using industry standards for both
  software and hardware.

- DEC's maintenance is probably better, and they have a longer warranty
  period.

If we have to buy the machine, at this point I'd vote for the
Sun-4/390.  But if the DEC gift comes through, it would certainly be
worth taking.  I'm not sure which system (the 5400 or 5800) we should
propose for the gift.  Our current needs could be met by a 5400 and it
costs less, but it is less expandable.

		Sun-4/370	Sun-4/390	DEC 5400	DEC 5800
		---------	---------	--------	--------

Processor	SPARC		SPARC		MIPS R3000	MIPS R3000
								(up to 2)
Main memory	8-56 MB		8-56 MB		16-64 MB	32-256 MB

Cache		128 KB		128 KB		128 KB		128 KB + 256 KB

System bus	VME		VME		Q		BI

Disk controller	SMD		IPI		KDA50		KDB50

Disk type	?		?		RA90		RA90

Disk capacity	688 MB		1.0 GB		1.2 GB		1.2 GB
  (per drive)
Max disk capac	5.5 GB		32 GB		9.7 GB		58 GB

Disk rate	2.4 MB/sec	3.0 MB/sec	2.8 MB/sec	2.8 MB/sec

Tape options	1/4" cartridge	1/4" cartridge	TK70 cartridge	TK70 cartridge
		1/2" 6250 bpi	1/2" 6250 bpi	1/2" 6250 bpi	1/2" 6250 bpi
		8mm video	8mm video

Serial ports	up to 68	up to 68	? (many)	? (many)

Operating sys	Sun OS		Sun OS		Ultrix		Ultrix

Warranty	3 months	3 months	12 months	12 months


			Prices for sample configuration

Memory		32 MB		32 MB		32 MB		32 MB
Disk		3.4 GB		4.0 GB		3.6 GB		3.6 GB
Tape		1/4"		1/4"		TK70		TK70
Serial lines	16		16		8		8

List price	$123,400	$140,800	$150,500	$206,600

All of the above configurations show cartridge tape drives.  We might
want to get one of those for installing system software and sending
tapes to other sites, but for backups we will want to use 8mm video
tape (or something better if it comes along).  Sun sells those fairly
cheap; DEC doesn't, but we can use one connected to another machine on
the network.  6250 bpi tape drives are available for all of these
systems, but are expensive and probably not worth getting.

∂21-Sep-89  1609	kolk@smiley.Stanford.EDU 	alternative to ap news wire  
Received: from smiley.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Sep 89  16:09:11 PDT
Received: by smiley.Stanford.EDU (4.0/inc-1.0)
	id AA16782; Thu, 21 Sep 89 16:06:32 PDT
Date: Thu, 21 Sep 89 16:06:32 PDT
From: kolk@smiley.Stanford.EDU (Dan Kolkowitz)
Message-Id: <8909212306.AA16782@smiley.Stanford.EDU>
To: jmc@sail.stanford.edu, les@sail.stanford.edu, me@sail.stanford.edu
Subject: alternative to ap news wire

Here is a description of clarinet news services.  They primarily use
the UPI news wire but are incorporating other sources.  They break it
down categorically according to the news wire instructions.  I'm including
the descriptions that they give of the service below.  I think that we can
get this for a modest and reasonable price.

Dan
----
>From looking!brad@watmath.waterloo.edu Thu Sep 21 14:16:03 1989
Return-Path: <looking!brad@watmath.waterloo.edu>
Received: from watmath.waterloo.edu by smiley.Stanford.EDU (4.0/inc-1.0)
	id AA16713; Thu, 21 Sep 89 14:15:33 PDT
Received: from looking.uucp by watmath.waterloo.edu with uucp
	id <AA20530>; Thu, 21 Sep 89 17:17:59 EDT
>From: brad@looking.on.ca (Brad Templeton)
X-Mailer: SCO System V Mail (version 3.2)
To: kolk@smiley.stanford.edu
Subject: Main ClariNet Blurb
Date: Thu, 21 Sep 89 16:54:42 EDT
Message-Id:  <8909212054.aa28074@looking.on.ca>
Status: R

[no save]

.xgw-vgx, -gg                     xx   -qgb. -qg-          xx
dM'   !#I  MM    ....    ....... ...    |MM#. |M    ....  _MM__
MM         MM   !#"'q#,   MMP"##  MM    |M'MM,|M  dMP"`Mb  MM
MM.   ...  MM   dgP'TM|   MM`     MM    |M 'qM+M  MM---""  MM
'M#,.,+P   MM   MM..dM|.  MM      MM    |M   YMM  vMb..gy  qMLjg
   ""`    """"`  '""`'"" """"`   """"` """"   '"    '""     '""
 -------------------------------------------------------------
 C  o  m  m  u  n  i  c  a  t  i  o  n  s          C  o  r  p.

 124 King St. N., Waterloo, Ontario, N2J 2X8      519/884-7473



What is ClariNet?

	ClariNet is a new electronic publishing network service that
	provides professional news and information, including live UPI
	wireservice news, in the USENET file format.

	ClariNet lets you read an "electronic newspaper" right on your
	own computer.  You can also get timely computer industry news,
	technology related wirestories, syndicated columns and features,
	financial information, stock quotes and more.

	ClariNet gives you scores of professionally produced newsgroups
	for about the same price as the analogous paper publications --
	but without advertising.   Instead of dialing up to an online
	service with a slow modem to read online news, you get it delivered
	to your own machine, to read and search at your own convenience,
	with your own favourite tools, on your own high speed terminal --
	and there's no meter running while you read.

	ClariNet news is provided using the USENET message interchange
	format, and available via UUCP and other delivery protocols.
	These formats are supported by a variety of packages for Unix
	and other systems.  If you're already on USENET, you're ready to
	receive ClariNet with no extra effort.  If you're not on USENET,
	we can help you get newsreading software, usually for free.

How does it work?

	At ClariNet, we licence publications of all sorts for electronic
	distribution.  We collect news directly off the UPI wire, classify
	it and convert it into USENET format.  In turn, generated articles
	are fed out on a regular basis to the UUNET hub machine, Anterior's
	Bay Area USENET hub and other distribution points.  You pick them up
	in USENET format and pay a subscription fee based on the number of
	readers at your site.

	(UUNET is a non-profit firm that exists to provide communications
	services to the Unix/USENET community.  They have the highest-capacity
	system of this type, which is why we buy their services.  Anterior
	runs a similar for-profit service in the SF Bay area.)

	You read your news using USENET newsreading tools.  Small
	volume publications can also be sent via electronic mail.  To
	ClariNews subscribers, we provide free use of the NewsClip
	programming language.  This lets you filter your newsreading as
	finely as you desire.

	With special permission, you can even feed ClariNet publications
	to other sites -- in fact, we'll pay you to do this.  Feed
	enough sites and your subscriptions will be free.

What are the publications?

   ClariNews -- our "electronic newspaper," gathered live from the wire
	services of United Press International (UPI).  ClariNews articles
	are distributed in one hundred newsgroups based on their subject
	matter, and are keyworded for additional topics and the geographical
	location of the story.  ClariNews has everything from headlines
	to industry news to box scores to network TV schedules.

	ClariNews is divided into three main products:

    ClariNews-N/S/F -- our general news"paper" with news, sports and
	features.  All major news, sports, weather and UPI feature articles,
	plus limited business and science/technology coverage.  Coverage
	of US and international news, Washington and more.  Around 400 stories
	per day, but categorized and headlined to help you read only what
	you want to read.  Get your news and sports scores long before they
	show up in your daily newspaper.

    TechWire -- special groups for stories on science and technology and
	the industries around them.  Computers, Electronics, Telecom,
	Space, Aerospace, Defence, Biotechnology, Research, Education
	and more.  Around 30-40 stories/day.   Coming soon: A daily report
	of the stock prices of computer industry stocks.

    ClariNews-Biz -- Business and financial stories.  Economic indicators,
	corporate news, regular stock market reports, legal issues,
	government info, commodities and much more.  Around 200 stories/day


    Newsbytes -- a weekly computer industry newsmagazine, on your computer
	well before InfoWorld or PC Week can be on your desk.  Gathered from
	11 bureaus around the world.  Newsbytes covers the IBM, Apple, Unix,
	Telecom and general computing worlds, along with coverage of
	trends, legal issues, industry news and reviews of new products.

    Networker's Journal -- a weekly review of what's happening in the
	wide area computer networking world, including online services,
	databases and BBSs.

    Syndicated Columns -- we're working to bring you major syndicated
	columns and features from syndicates willing to do electronic
	publishing.   We can bring supplemental feature services,
	opinion columnists, humour columnists, consumer advocates and
	more.  Tell us what you would like.

    Coming Soon:
	We're working on getting new publications all the time.  In
	the works are a Science Fiction publication, Hollywood and
	Entertainment news, Unix industry news, academic journals and
	our own publications.



	We have sheets with more info on these individual products.
	Contact us and list the products that you are interested
	in if you want more info.

	NewsBytes and ClariNews are operational now.  Additional
	products will come on line shortly.


What does it cost?

	ClariNet publications can be purchased either with single
	subscriptions (for 1 or 2 people) or with a "site" subscription
	that covers your computer, LAN or facility.  Each product has
	a "base price" that is the cost for 1 or 2 readers.

	For site subscriptions, we calculate a subscriber count based
	on the newsreading population of your site.   Send us your best
	estimate of the number of people who read any kind of electronic
	news on your computer, including local discussions, and we can
	provide a quote.   The more subscribers you have, the less you
	pay per subscriber -- as much as 88% less.

	If you're willing to use the Unix "group" facility or NNTP access
	lists to restrict readership on your machine, we can also bill
	individual subscribers for you, and give you (the sysadmin) a free
	personal subscription for your trouble.

	Here are the base prices for the products, along with the
	percentages we use in calculating subscriber counts.  If you
	don't currently use any form of electronic news, tell us the
	general user population of your systems, and we will work from
	that.

	Product			Base Price	Site Licence Percentage

	Full ClariNews		$25/month	20%	(Special Discount Ramp)
	   News/Sports/Features	$10/month	20%
	   TechWire		$10/month	25%
	   Business News	$15/month	15%
	   N/S/F + TechWire	$18/month	25%

	Newsbytes		$25/6 months	20%
	Networker's Journal	$20/year	12%

		Trial Products

	Typical Synd Column	$20/year	12%
	Infomat			$30/year	15%

		Packages

	ALL Products		$30/month	25%	(Special Discount Ramp)

	All but TechWire &
	  Business News		$17/month	20%

	As an example, here are the "site subscription" prices on these
	publications for a system with 100 netreaders.  All prices are 
	in US dollars per month:

	Product			Price/Month	Price/Netreader/Month

	ClariNews-N/S/F		$83		$0.83
	TechWire		$93		$0.93
	NewsBytes		$34.50		$0.35
	Networker's Jrnl	$9.14		$0.09
	TechWire+NewsBytes	$111.60		$1.12
	Typical Synd. Column	$9.14		$0.09
	ClariNews-ALL		$145		$1.45
	Everything!		$174		$1.74

	Prices are subject to change without notice to non-subscribers
	Write to info@clarinet.com for an exact quote.  Prices are in
	U.S. Dollars.

	We are currently considering a special educational discount
	program where educational institutions that purchase a site
	licence would be allowed to not count undergrads in their
	readership figures.

Can I see what it looks like?

	We'll be glad to give you a free two-week trial of any
	product, with no obligation.  If you already connect to
	UUNET or another authorized feed site, we can set a trial
	up for you almost immediately, with no effort on your part.

	If you're not satisfied with the products at the end of
	the trial, send back the invoice (even by e-mail) and you
	will be charged nothing.  "Keep those first issues as our gift
	to you."  :-)


	Some sample articles are included at the end of this article.
	You can contact us for more.

What about feeding other sites?

	USENET works by having one site in an area feed other sites.
	ClariNet can work that way too.  If you apply for permission,
	and you have the capacity to do it reliably, we will let you
	feed other sites.  In fact, we'll pay you to do this by
	reducing your subscription fees by 10% of the fees we charge
	the sites you feed directly.  You can either register with us
	to feed sites, or you can go out and find possible subscribers
	amongst the sites you already know.  If you do that, we'll discount
	your price even further.

	Feed 7-10 similar sized sites in this manner and your subscription
	will be free.

How do I subscribe?

	If you need more information, mail to info@clarinet.com.  If you
	want an electronic order form and the full sheet of terms and
	conditions, mail to order@clarinet.com.   You can also phone
	800-265-2782 or 519-884-7473.  E-mail is easiest, but you may
	wish to use the phone to talk in person or give confidential
	information like UUCP connect info or credit card numbers.

	Internet users should be aware that certain commercial messages
	may not be appropriate for transmission over the internet.

	You can pay by check or credit card, or we can invoice site
	subscribers.  There is a $1 discount per invoice if you allow
	electronic mail invoicing and pay promptly.  We take Visa, MasterCard
	and Amex.

	You can also write to us at:

		ClariNet Communications Corp.
		124 King St. N.
		Waterloo, ON
		N2J 2X8

How do I get a feed?
	
	Two easy places are UUNET and Anterior (Fernwood).  Write to
	postmaster@uunet.uu.net for details on how to become a customer
	of theirs.  They provide cheap communications, but do charge
	a $35/month administrative fee on top of their telecom charges.
	After that, they can feed you for as little as $5/hour.
	Fernwood has a $18/month administration fee and slightly higher
	hourly rates.  Write to postmaster@fernwood.mpk.ca.us.

	You can also get a feed from one of our existing customers in
	your area.  Our net is growing.  Contact us to see if we have
	a feed site in an area that you can call.

	At ClariNet itself, we are restricting ourselves to feeding
	customers within Canada.

	You can also get a live feed over the Internet, if you are an
	authorized internet user.  Currently there is no NSFnet policy
	against this.  They have the ClariNet proposal under review, and
	we will let you know what they decide for the long term.

	If there is no feed point suitable for you, we will keep you on
	file and inform you when one shows up.

	Mail feeds are possible for very small publications like syndicated
	columns.  On the other hand, the individual sending requirements
	can make this sort of arrangement expensive.  In addition, you should
	be sure you have the permission of the sites that will pass the
	mail on to you.

Do I really need more stuff to read?

	No.  But we do think you can use your time better when you read
	professionally written and classified information -- information
	worth paying for.   With NewsClip you can arrange so that each
	article you see is important to you and worth reading.  Compare
	that to the time you spend on USENET or online services -- it
	might actually save you time and money.

	In fact, ClariNet could be the best reason to run USENET software
	on your machine -- or something that makes it even more worthwhile
	to all levels of your company.

What are the terms and conditions?

	ClariNet products are offered under special terms.  If you wish
	to subscribe, we'll send you the full sheet of terms and conditions.
	In general, the terms state that:

		You must not interfere with the operation of the ClariNet
		network (no forged control messages etc.)
		You must respect copyright, and not allow unauthorized access.
		You must pay your subscription fees on time.
		Some products may not be archived forever.
		The rules are subject to change with 30 days notice.
		Customers may not re-sell or re-publish ClariNet products.
		 (Some limited use on fee paying BBSs is allowed.)
		Articles must be kept intact.
		ClariNet is not responsible for information provided by
		  outside information providers.  Most products are offered
		  on an as-is basis.  (UPI is as-is)
		Customers are responsible for the cost of communication from
		  a ClariNet feedpoint to their computer.

What are the advantages of ClariNet?

	Many online services offer you electronic publications, including
	live newswires.  The difference with ClariNet is that we deliver
	it to your computer.  You don't have to dial up and read it with
	a 2400 bps modem, with a meter running.   You get to read your
	news on your own machine, on your own time, and on your own
	high speed terminal.

	The subscription fee is quite low.  General ClariNews at $10 a
	month is a bargain compared to many online services, where
	that $10 would buy less than an hour of newsreading per month.

How can I handle 500 new articles of news each day?

	You can search the news on your own machine with your own tools.
	ClariNews is quite large, and you would not be expected to read
	all of it, just as you don't read all of a daily paper.

	Because ClariNews is headlined by professional editors, all it
	takes is a simple "=' (in RN) to make a quick scan of the headlines.
	Then pick and choose the stories you like.

	Or use a NewsClip filter program.  ClariNews is highly keyworded.
	You can arrange to see only articles that describe topics, people
	places or companies that you are interested.   Want to see only
	the computer stories that mention Sun Microsystems? -- no problem.

	There are hundreds of different keywords that the editors place on
	each article.  Topics from AIDS to child abuse to televangelism to
	inflation to gun control can be tracked.

How fast can I get the news?

	We get the news right as it comes off the wire, and deliver it
	to UUNET within two hours.  If you call UUNET regularly, or have
	an internet connection, you can get the news as it happens.
	Most evening papers close out just after noon.  Many morning papers
	close out at midnight.   ClariNews stories will beat the papers for
	you every time.

	NewsBytes closes after PC Week and InfoWorld and comes out before
	they are printed.  That's the advantage of electronic News.

How will you stop information thieves?

	We'll reward people who report piracy by giving them free
	subscriptions.  Many people steal software and information, we're
	sad to say, but people would have to be crazy to steal from a news
	service.   If they are caught and it is reported in the news, they
	would be likely to lose a lot more than the pirate feed.

Can I buy NewsClip without ClariNet?

	Currently it is available only with ClariNet.  We will, however
	make special arrangements for sites that can't buy ClariNet.
	The licence fee would depend on site size.

What if my site can't join but users want to subscribe?

	If your sysadmin is willing to put the "clari" news articles
	in a special "group" that can only be read by people we
	authorize, we will be glad to bill individual users or
	groups of users.   The discounts apply to each billing unit.
	(So if everybody wants to be billed individually, each pair pays
	the base price, but you can combine for a discount.)

	On small sites, even this security measure is not required.

	Check with your library to see if they have a standard policy
	on ordering publications.

	Finally, if due to budget bureaucracy you find that you can buy
	software but not electronic news, we can arrange to sell you
	NewsClip and give you ClariNet free.... (sigh).

What if I don't have a Unix system?

	There are software packages available that handle USENET format
	news for many different operating systems.  There are even
	a couple of free/shareware packages for the IBM-PC.  You can
	arrange to ship your news to your IBM-PC at home or on the
	road, if desired.

What more information can I get?

	Write to info@clarinet.com for more information.  Here are some
	of the files of information that we have.  Many of these files
	duplicate what's said in this file, however.

feed		- How to become a site that feeds other sites
netjournal	- Description of "Networker's Journal"
newsclip	- Details on the NewsClip language
order		- How to place an order
prices		- Prices and user counts for ClariNet products
techwire	- Description of TechWire (posted to USENET)
terms		- Terms, conditions and network rules
trial		- Description of trial products (Informat, Horowitz, Hollywood)
clarinews	- Description of ClariNews (posted to USENET)
newsbytes	- Description of NEWSBYTES computer news (posted to USENET)


	Here's a sample UPI wire article:

Path: clarinews
>From: clarinews@clarinet.com
Newsgroups: clari.news.politics,clari.news.asia
Subject: China now banning books, magazines
Keywords: government, censorship, politics, china
Date: 18 Jul 89 09:46:00 GMT
Lines: 70

BEIJING (JULY 18) UPI - China, widening a crackdown on publications in 
response to pro-democracy protests, has banned more than 60 categories of
books and periodicals which it said are ``pornographic, superstitious or
violent,'' Chinese news reports said Tuesday. 

Authorities warned those selling the banned publications would be 
``severely punished,'' the Beijing Evening News reported. 

In Tokyo, meanwhile, Japan Air Lines (JAL) said it has received a letter
threatening to kill two Japanese every month because of Japan's economic
cooperation with China. 

The official Beijing Daily said the capital's Press and Publishing Bureau
on Monday ``identified more than 60 categories of books and periodicals
with pornographic, superstitious or violent contents.'' 

The bureau ordered that all bookshops and and publishing houses 
immediately stop selling these publications, the Beijing Daily said. 

It said other categories of books and periodicals advocating ``bourgeois
liberalism,'' a catchphrase for Western political systems and private
ownership, would also later be identified and included in the ban. 

The move is the latest by Chinese authorities in a campaign to stamp out
Western influences, blamed for the recent student-led pro-democracy
protests in April and May. 

The Chinese army brutally suppressed the Beijing protesters June 3 and 4,
killing hundreds or perhaps thousands and igniting a citywide uprising the
government has branded a ``counterrevolutionary rebellion.'' Almost 3,000
people are confirmed to have been arrested in a nationwide sweep for
supporters of the movement. 

Officials two weeks ago raided private bookstalls, confiscating works
authored by leading dissidents and liberal intellectuals. Many of the
writers are now under arrest, in hiding or have fled the country. 

Last week, authorities cleared all foreign newspapers and periodicals from
the lobby bookshops of tourist hotels in the capital, and ordered that
writings by pro-democracy activists be ``confiscated or destroyed.'' 

In Tokyo, the Japanese Foreign Ministry said it was taking seriously the
letter signed by the previously unknown ``Supreme Command for Death
Squads'' and received by JAL's Beijing office Monday. It said it has asked
the Chinese government ``to keep a close watch to ensure the safety of the
Japanese in China.'' 

A JAL spokesman in Tokyo said the letter, written in Chinese, ``is 
believed to have come from elements opposed to the Chinese communist 
regime.'' 

It threatened to kill one Japanese person a month starting Aug. 15 and
then one per week from next year unless Tokyo halts its ``its economic
aggression in conspiracy with the Chinese Communist Party.'' 

The letter said targets of the attacks would include businessmen, tourists
 and ``friends of the Chinese Communist Party,'' but said Japanese Embassy
staff, students and journalists would be left unharmed, the spokesman said.

He said the letter did not indicate how the group planned to carry out the
attacks. 

The Japanese Foreign Ministry said it is investigating the background of
the group. 

Japan, the largest donor of foreign aid to China, is also one of its
biggest trading partners. During the recent Paris summit of seven major
industrialized nations, Tokyo led a movement against isolating China from
the international community despite Beijing's bloody crackdown on the
pro-democracy movement. 

    Here are two sample NewsBytes articles:

=====================================================
>From: newsbytes@clarinet.com
Newsgroups: clari.nb.unix
Subject: Alliance Promotes X Terminal
Keywords: Bureau-TYO
Lines: 24

TOKYO, JAPAN, 1989 MAY 23 (NB) -- An alliance of eighteen
Japanese firms, including hardware makers, software vendors and
users, have announced plans to organize the UWS Consortium in mid-
June.  UWS stands for User Interface WorkStation.

The X terminal was developed by the Massachusetts Institute of
Technology, or MIT, of the U.S., based on X Window.  The UWS
Consortium will develop Japanese functions for X Terminal and
expand it for general purpose workstations.

The X Terminal is an intelligent terminal capable of controlling
host computers, such as supercomputers and general purpose
computers on a network.

To start, the consortium will release a monochrome terminal for
350,000 yen or $2,415 and a color terminal for 650,000 yen or
$4,482 for Unix-based workstations by this fall.

The consortium will call for leading Japanese hardware makers
for further development products that standardize X Terminal.

(Naoyuki Yazawa/19890525)

===========================================================
>From: newsbytes@clarinet.com
Newsgroups: clari.nb.business
Subject: DEC to Freeze Salaries
Keywords: Bureau-BOS
Lines: 17

MAYNARD, MASSACHUSETTS, U.S.A., 1989 MAY 20 (NB) -- Digital
Equipment Corp. will freeze the salaries of all of its U.S. employees
when the new fiscal year begins this summer, the company has
confirmed.

DEC, which is the largest private employer in Massachusetts and
New Hampshire, said the freeze would take effect on July 2, the start
of the company's 1990 fiscal year.

The freeze will affect 73,500 DEC employees across the country, and
will remain in effect indefinitely. DEC is in a pitched battle against
IBM, HP and other vendors in the mid-size computer arena.

(Jon Pepper/19890526/Press Contact: Joyce Radnor, DEC, 508-486-
5096)

------
>From looking!brad@watmath.waterloo.edu Thu Sep 21 14:16:46 1989
Return-Path: <looking!brad@watmath.waterloo.edu>
Received: from watmath.waterloo.edu by smiley.Stanford.EDU (4.0/inc-1.0)
	id AA16718; Thu, 21 Sep 89 14:16:22 PDT
Received: from looking.uucp by watmath.waterloo.edu with uucp
	id <AA20585>; Thu, 21 Sep 89 17:18:52 EDT
>From: brad@looking.on.ca (Brad Templeton)
X-Mailer: SCO System V Mail (version 3.2)
To: kolk@smiley.stanford.edu
Subject: ClariNet current group list
Date: Thu, 21 Sep 89 16:54:56 EDT
Message-Id:  <8909212054.aa28083@looking.on.ca>
Status: R


Following is a list of the newsgroups in Clarinet and a brief description 
of the typical content of articles in that group: (note that you may or may
not get all these groups depending on which Clarinet package you subscribe

clari.news.flash:
	One line news flashes of extremely important stories, such as
	"Space shuttle explodes," or "Kennedy Shot." 
	Usually only one story of this magnitude per year or two. 

clari.news.bulletin: 
	Most important breaking stories of the week.   1 or 2 per week,
	but each comes in multiple versions as details break.

clari.news.urgent: 
	Top breaking stories of the day.  1-3 per day.

"Breaking" news stories usually come in several versions as developments
occur.  In addition, as time passes, they will leave the breaking
newsgroups and move into just appropriate newsgroups for their category.

clari.net.products: 
	Product announcements of Clarinet products.

clari.net.admin:
	Information regarding the Clarinet network for Sysadmins.

clari.net.announce:
	General announcements regarding the Clarinet network.

clari.tw.aerospace: 
	The aerospace industry.   Low volume.

clari.tw.computers: 
	Computer industry news, developments, and new product info.
	4-5 articles/day

clari.tw.defense: 
	Military developments in weaponry, SDI, radar systems, etc.
	Aprrox 1 article per day.


clari.tw.education: 
	A broad topic, covering education, research and most aspects
	of higher education, including regional university stories.
	Approx. 6 articles per day.

clari.tw.electronics: 
	Latest developments in electronics and circuitry.
	Approx. 1 article per day.

clari.tw.misc: 
	Technological articles of various appeal.  Chemicals, consulting,
	underwater industry, conservation, unusual energy sources, futurism,
	general r&d etc.

clari.tw.nuclear: 
	The nuclear industry, plant leaks, plant closings, nuclear waste.
	Approx 2 articles per day.

clari.tw.science: 
	Scientific discoveries, experiments, research.
	2-3 articles per day.
	
clari.tw.space: 
	Space travel, space probes, satellite launchings, technological 
	spinoffs from the space industry.
	
clari.tw.telecom: 
	News relating to the telecommunications industry, along with
	coverage of AT&T and the Baby Bells.
	3-6 articles per day.
	
clari.news.almanac
	Almanac for the day.  Includes famous birthdays, important historical 
	events which occurred on the given day, a quote of the day.

clari.news.arts
	The stage, dance and general art. 
	
clari.news.aviation
	The airline industry, crashes, near-misses, and air controllers
	are covered here.

clari.news.books
	News form the world of books and publishing.

clari.news.briefs
	Summaries of the top 12-15 stories of the moment.
	Updated 4-5 times per day.

clari.news.canada
	News pertaining to Canada.  Unfortunately, UPI editors often
	don't properly identify the location of stories, so this is
	only a small selection of actual stories involving Canada.

clari.news.children
	Children in the news, from day-care to achievements to crime.

clari.news.consumer
	Product recalls, new legislation, consumer safety.

clari.news.demonstration
	Demonstrations around the world, both peaceful and violent.

clari.news.disaster
	Man-made and natural disasters.  Covers a range of topics from
	airline crashes to bad weather to volcanic eruptions.  The word
	disaster is a bit of an exaggeration.

clari.news.economy
	Any news impacting upon economies, American or foreign.

clari.news.election
	News of elections around the United States and around the world.

clari.news.entertain
	The latest form the entertainment industry.  Contains a daily
	"people" column.

clari.news.environment
	Covers a diverse range of environmental issues, from endangered
	species to oil spills, to clean-air bills.

clari.news.europe
	News that UPI editors coded with a European location.  Unfortunately,
	this is not all stories about Europe.

clari.news.features
	Feature news stories of the day.  Usually 1-6 stories.
	(Currently still under development)

clari.news.fighting
	Armed conflict around the world.

clari.news.goodnews
	And you thought all the news was bad!

clari.news.gov
	Misc news involving governments, both domestic, and foreign.
	3-6 articles per day.

clari.news.gov.agency
	IRS, FBI, DEA, etc. all get covered here.

clari.news.gov.budget
	Any news related to government budgets.

clari.news.gov.corrupt
	Corruptions, scandals, extortion, and various other breaches of
	trust.

clari.news.gov.international
	Government news from abroad.  Much non-US news ends up in this
	newsgroup.

clari.news.gov.officials
	People in the government.

clari.news.gov.state
	News from various state governments.

clari.news.gov.taxes
	Tax hikes, the deficit, new tax laws, etc.

clari.news.gov.usa
	Government stories pertinent to the US.  This group also contains
	a fair number of general US national stories.

clari.news.group
	Articles about special interest groups that don't have their own
	newsgroup.  (See below.)  Currently includes men, asians, general
	ethnics, border areas, hispanics, the poor, seniors, the handicapped,
	gays, rural areas and philanthropy.   Not a very connected set --
	but this group is due for expansion.

clari.news.group.blacks
	News of special significance to blacks.

clari.news.group.women
	News of special significance to women.  Includes abortion.

clari.news.health
	Latest findings and studies from the health field.

clari.news.health.aids
	Issues, research, and  politics surrounding AIDS.

clari.news.interest
	News often not of great significance, but interesting nonetheless.
	"Human interest" stories, as they are called.
	
clari.news.interest.animals
	Animals in the news!

clari.news.interest.history
	News of historical significance.  For example, gatherings to mark
	the 50th anniversary of WW II.

clari.news.interest.people
	People in the news.  Stories on celebs.

clari.news.interest.quirks
	Strange, unusual, and often humourous stories.  Includes the
	daily "quirks in the news" summary.

clari.news.issues
	Political, environmental, and financial and moral issues of our
	day.  Includes civil rights.

clari.news.issues.conflict
	Conflicts and their resolutions.

clari.news.labor
	Strikes, unions, employment figures, coalitions.

clari.news.labor.strike
	Labor strikes and their resolutions.

clari.news.law
	Courts, lawyers, lawsuits, new laws and legislation.

clari.news.law.civil
	Civil trials and lawsuits.

clari.news.law.crime
	Crimes and criminal law.

clari.news.law.crime.sex
	Sex crime, trials, and laws.

clari.news.law.crime.trial
	Famous trial proceedings and their outcomes.

clari.news.law.crime.violent
	Violent crimes, trials, and eventual sentencing.

clari.news.law.drugs
	Drug busts, trials, new laws, the "war on drugs."

clari.news.law.investigation
	Investigations and Federal Grand Juries.

clari.news.law.police
	Police and police actions in the news.

clari.news.law.prison
	Prison riots, programs, and reform.

clari.news.law.profession
	Lawyers, judges, and the likes.

clari.news.law.supreme
	Latest decisions from the Supreme Court.

clari.news.lifestyle
	Trends in living.

clari.news.military
	Military news from around the world.  Equipment, strategies, 
	military figures.

clari.news.movies
	Movies, the people who star in them, and direct them.  Regular
	box office reports.

clari.news.music
	News and reviews from the world of music.  Regular pop chart
	summaries.

clari.news.politics
	Politics and politicians.

clari.news.religion
	Organized religion, evangelism, cults, scandals.  News important
	to Jews is also covered here.

clari.news.sex
	Studies and research on human sexuality, birth control, porn,
	and related matters.

clari.news.terrorism
	Terrorist actions and demands around the world.

clari.news.trends
	Changes in consumer spending, employment figures etc.  Survey
	and poll results.

clari.news.trouble
	General reports of low level "bad news."  Accidents, mishaps etc.

clari.news.tv
	TV highlights, people, and special shows.  Low-detail listings,
	plus ratings.

clari.news.weather
	Temperatures around the US and around the world, brief synopses of
	weather arounds the US.

clari.sports.baseball: 
	Latest box scores, roundups, and standings, as well as a host of 
	other news related to baseball.  Includes a story for each game
	in the majors, and scouting reports as well.  See also sports.misc
	for the daily "sports digest" and "sports transactions" articles.
	
clari.sports.basketball: 
	Game scores, standings, and other news from the world of
	basketball.
	
clari.sports.football: 
	Scores, standings, statistics, injuries, and general football news.

clari.sports.hockey: 
	Scores, standings, general hockey news.

clari.sports.misc: 
	Coverage of other sports, especially tennis, golf, auto racing 
	and the olympics.  Daily sports transactions and highlights.

clari.biz.commodity: 
	Commodity prices.  Metals and popular commodities.  Currently
	livestock and some farm commodities are not being forwarded.
	
clari.biz.courts: 
	Business and the legal system.
	
clari.biz.economy: 
	Economics slowdowns, upturns, and "soft landings."   Other news
	which might effect the economy including interest rates, new
	budgets, etc.
	
clari.biz.finance: 
	Money reports, foreign exchange.
	
clari.biz.finance.earnings: 
	Earnings of various companies.   Dividend reports, too.
	
clari.biz.finance.personal: 
	Personal finance and investing.
	
clari.biz.finance.services: 
	News from the financial service industry.
	
clari.biz.index: 
	Stock market index reports.
	
clari.biz.invest: 
	General investment news.  Bonds etc.
	
clari.biz.labor: 
	Labor relations.
	
clari.biz.mergers: 
	Mergers and Acquisitions.
	
clari.biz.misc: 
	Other general business news.
	
clari.biz.market.dow: 
	Dow Jones reports.
	
clari.biz.market.amex: 
	American Stock Exchange reports.
	
clari.biz.market.otc: 
	Nasdaq stock reports.

clari.biz.market.ny: 
	New York Stock Exchange reports
	
clari.biz.market.report: 
	General market reports.  Standard & Poor's info.


======================================================================

	Newsbytes Computer Industry News Service Newsgroups

clari.nb.govt: 
	Computer stories involving government contracts, grants, research
	programs, laws lawsuits, and foreign governments.
	Approx. 15 articles per week.

clari.nb.telecom: 
	Latest news from the telecommunications industry.
	Approx. 25 articles per week..

clari.nb.trends: 
	Research and discoveries which may impact upon current technologies
	and research.  Burgeoning and declining markets.  Software and
	hardware sales trends.
	Approx. 12 articles per week.

clari.nb.business: 
	Latest business news form the computer industry.
	Approx. 30 articles per week.

clari.nb.unix: 
	News on UNIX and UNIX-ike operating systems.
	Approx. 10 articles per week.

clari.nb.general:  
	General interest stories of a technical nature.
	Approx. 20 articles per week.

clari.nb.apple: 
	Latest Apple & Macintosh news
	Approx. 15 articles per week.

clari.nb.ibm: 
	Latest news on IBM products and developments.
	Approx. 15 articles per week.

clari.nb.review: 
	Reviews of newly released hardware & software products.
	2 - 3 reviews per week.

∂21-Sep-89  1651	pamela@cs.columbia.edu 	re: WGBH/BBC project 
Received: from cs.columbia.edu by SAIL.Stanford.EDU with TCP; 21 Sep 89  16:47:27 PDT
Received: by cs.columbia.edu (5.59++/0.1) 
	id AA00759; Thu, 21 Sep 89 19:48:12 EDT
Date: Thu, 21 Sep 1989 19:48:12 EDT
From: Pamela McCorduck <pamela@cs.columbia.edu>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: WGBH/BBC project
In-Reply-To: Your message of 19 Sep 89 1316 PDT
Message-Id: <CMM.0.88.622424892.pamela@cs.columbia.edu>

We're talking about Friday, Sept. 29.  Come for drinks, come for dinner.
A pleasure to see youwhenever...Let me know.

Pamela

∂21-Sep-89  1747	tom@Polya.Stanford.EDU 	Comparison of Sun and DEC machines  
Received: from Polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Sep 89  17:47:00 PDT
Received:  by Polya.Stanford.EDU (5.61/25-eef) id AA06425; Thu, 21 Sep 89 17:47:22 -0700
Date: Thu, 21 Sep 89 17:47:22 -0700
From: Tom Dienstbier <tom@Polya.Stanford.EDU>
Message-Id: <8909220047.AA06425@Polya.Stanford.EDU>
To: weening@Gang-of-Four.Stanford.EDU
Cc: clt@sail, jmc@sail
In-Reply-To: Joe Weening's message of Thu, 21 Sep 89 15:47:27 -0700 <8909212247.AA26635@Gang-of-Four.Stanford.EDU>
Subject: Comparison of Sun and DEC machines

To add a little here,,There are 8MM's for the dec boxes by third party
people. Cheriton currently has one. Also, the BI bus on the 5800 is
around 50MB.


tom

∂21-Sep-89  1759	VAL 	reply to message    
[In reply to message rcvd 21-Sep-89 17:59-PT.]

ok

∂21-Sep-89  1953	VAL 	The scope paper

It seems reasonable, and the Tweety example is done very nicely.
But "scoped circ'n" isn't really a new form of circ'n, I think. If
we ignore "NS", this is the usual formula circ'n applied to a
different formula.

I think I can prove that, for a large class of examples including
theirs, the effect of varying Scope is the same as the effect of
minimizing it.

∂21-Sep-89  2120	perlis@yoohoo.cs.umd.edu 	your visit    
Received: from mimsy.umd.edu by SAIL.Stanford.EDU with TCP; 21 Sep 89  21:20:12 PDT
Received: from yoohoo.cs.umd.edu by mimsy.umd.edu (5.61/4.7)
	id AA02197; Fri, 22 Sep 89 00:18:40 -0400
Received: by yoohoo.cs.umd.edu (5.61/3.14)
	id AA14369; Thu, 21 Sep 89 21:35:52 -0400
Date: Thu, 21 Sep 89 21:35:52 -0400
From: perlis@yoohoo.cs.umd.edu (Don Perlis)
Return-Path: <perlis>
Message-Id: <8909220135.AA14369@yoohoo.cs.umd.edu>
To: jmc@sail.stanford.edu
Subject: your visit
Cc: perlis@yoohoo.cs.umd.edu, sarit@grinch.umiacs.umd.edu

John, it would be great to talk to you when you are in the 
DC area. My schedule is as follows:

Tues (Oct 3)
	11am - 12:15pm class
	12:30 - 1:45pm Cog Sci seminar (Intro to the brain)
	2 - 3pm office hours
	4 - 5pm Minker's Logic Programming seminar

Wed (Oct 4)
	11am - noon class

Thurs (Oct 5)
	11am - 12:15pm class
	12:30 - 1:45pm Cog Sci seminar (Cherniak on minimal rationality)
	2 - 3pm office hours
	3 - 4pm meeting

Any other times are free, and also we can probably meet during my
office hours (students rarely show up!). Moreover, you are of
course welcome to attend the seminars mentioned.

My phone numbers are:

home: (301 927-2116
office:	(301) 454-7931

-Don

P.S. I will try to have Sarit Kraus meet with us as well. I am
pleased that the "scope" notion interests you.

∂22-Sep-89  0945	rabin@harvard.harvard.edu     
Received: from harvard.harvard.edu by SAIL.Stanford.EDU with TCP; 22 Sep 89  09:45:13 PDT
Received: by harvard.harvard.edu (5.54/a0.25)
	(for jmc@sail.stanford.edu) id AA26312; Fri, 22 Sep 89 12:48:06 EDT
Received: by endor.HARVARD.EDU; Fri, 22 Sep 89 12:47:57 EDT
Date: Fri, 22 Sep 89 12:47:57 EDT
From: rabin@harvard.harvard.edu (Michael Rabin)
To: jmc@sail.stanford.edu

Lecture

Hi, 

We have sceduled your lecture for Monday at 4pm.
Please e-mail atitle and abstract. See you soon
Michael

∂22-Sep-89  1311	VAL 	Seminar   

I propose that we begin with two introductory lectures: on nonmonotonic formalisms
(or at least circumscription) on Oct. 16 and on logic programming on Oct. 30.
Arima will speak on Oct. 23. Do you agree with this plan, and would you like to
give any or both of the lectures?

∂22-Sep-89  1353	VAL 	re: Seminar    
[In reply to message rcvd 22-Sep-89 13:39-PT.]

Fine.

∂22-Sep-89  1546	CLT 	hardware  
To:   pullen@VAX.DARPA.MIL, scherlis@VAX.DARPA.MIL,
      cross@VAX.DARPA.MIL
CC:   JMC@SAIL.Stanford.EDU, weening@Gang-of-Four.Stanford.EDU,
      tom@polay   

Our ancient Dec KL10, SAIL, will be retired (too old and expensive to
maintain) soon (hopefully in February).  Darpa funded the purchase
of the original machine in 1965 and we have used this equipment with
occasional upgrades continuously since that time.  SAIL together with
the Alliant provides the main computing facilities used by our group.  

The Alliant is not a viable replacement for SAIL for a number of reasons.
It is a special purpose computer and the maintainence cost is
excessively high (approx 70k/yr including overhead).  Sharing software with
non-Alliant users is difficult (big systems do not port readily), and the
Alliant itself is almost 4 years old and will soon be obsolete and
ready for retirement.  Thus we need to purchase replacement hardware.

We want to continue with a multi-user time-sharing system probably shared
with a small number of additional users.  We have met with both DEC and
SUN representatives.  A summary of the information collected is appended
at the end of this message.  The bottom line list prices for processor-
memory-disk configurations of interest range from $140k to $207k.
Stanford discounts seem to be around 35% to 40%.

In addition we will need terminals for office and home.  (Sail terminals
won't work).  Here we are thinking of the NCD X terminals (about 2k each).
A minimum of 14 terminals (and 7 modems at .5-.9k each) is needed
for the existing researchers and additional terminals will be needed for
secretaries and students.  

This system would provide facilities for at least John McCarthy's group
(McCarthy, Lifschitz, Talcott, Mason, Weening, Pehoushek) and Zohar Manna's
group (Manna and Waldinger).

The yearly cost for maintaining the new system is estimated to be 75k (17k
for computer maintenance, 8k for terminal maintainence, and 50k for
systems programmer, backup, etc.)  This is to be contrasted with the SAIL
computer charges for the two groups which were running at approximately
200k/year.

The Qlisp project recently requested approval to purchase the 4 Alliant
processors currently on loan (the loan officially terminated last April
and Alliant is pushing for a decision).  Program development can be done
on 4 processors, but it is important to have more than 4 processors to get
reasonable experimental data.  We are investigating the possibility of
using 8 processor Alliants at Argonne or elsewhere for gathering final
data.  If this is feasible then it would probably be better not to
purchase the extra Alliant processors, but to use that money for the
proposed new equipment.

-----------------------------------------------------------------------------
Summary of hardware configurations and prices
-----------------------------------------------------------------------------
Date: Thu, 21 Sep 89 15:47:27 -0700
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
To: clt@sail, jmc@sail, tom@polya
Subject: Comparison of Sun and DEC machines

Below is a chart I made up based on the information we have gotten.
The bottom line shows list prices, not counting maintenance, because
the discounts and maintenance options are too confusing for me to work
out right now.  The discounts seem roughly the same (DEC 35%, Sun 38%,
both with various additional discounts).

My own impressions are:

- The systems are roughly comparable in performance, with DEC slightly
  more powerful and more expandable.

- Sun seems better in software and in using industry standards for both
  software and hardware.

- DEC's maintenance is probably better, and they have a longer warranty
  period.

If we have to buy the machine, at this point I'd vote for the
Sun-4/390.  But if the DEC gift comes through, it would certainly be
worth taking.  I'm not sure which system (the 5400 or 5800) we should
propose for the gift.  Our current needs could be met by a 5400 and it
costs less, but it is less expandable.

		Sun-4/370	Sun-4/390	DEC 5400	DEC 5800
		---------	---------	--------	--------

Processor	SPARC		SPARC		MIPS R3000	MIPS R3000
								(up to 2)
Main memory	8-56 MB		8-56 MB		16-64 MB	32-256 MB

Cache		128 KB		128 KB		128 KB		128 KB + 256 KB

System bus	VME		VME		Q		BI

Disk controller	SMD		IPI		KDA50		KDB50

Disk type	?		?		RA90		RA90

Disk capacity	688 MB		1.0 GB		1.2 GB		1.2 GB
  (per drive)
Max disk capac	5.5 GB		32 GB		9.7 GB		58 GB

Disk rate	2.4 MB/sec	3.0 MB/sec	2.8 MB/sec	2.8 MB/sec

Tape options	1/4" cartridge	1/4" cartridge	TK70 cartridge	TK70 cartridge
		1/2" 6250 bpi	1/2" 6250 bpi	1/2" 6250 bpi	1/2" 6250 bpi
		8mm video	8mm video

Serial ports	up to 68	up to 68	? (many)	? (many)

Operating sys	Sun OS		Sun OS		Ultrix		Ultrix

Warranty	3 months	3 months	12 months	12 months


			Prices for sample configuration

Memory		32 MB		32 MB		32 MB		32 MB
Disk		3.4 GB		4.0 GB		3.6 GB		3.6 GB
Tape		1/4"		1/4"		TK70		TK70
Serial lines	16		16		8		8

List price	$123,400	$140,800	$150,500	$206,600

All of the above configurations show cartridge tape drives.  We might
want to get one of those for installing system software and sending
tapes to other sites, but for backups we will want to use 8mm video
tape (or something better if it comes along).  Sun sells those fairly
cheap; DEC doesn't, but we can use one connected to another machine on
the network.  6250 bpi tape drives are available for all of these
systems, but are expensive and probably not worth getting.

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

∂22-Sep-89  1656	sf@csli.Stanford.EDU 	Matijasevitch visit    
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Sep 89  16:55:53 PDT
Received: by csli.Stanford.EDU (4.0/inc-1.0)
	id AA07619; Fri, 22 Sep 89 16:57:13 PDT
Date: Fri 22 Sep 89 16:57:13-PDT
From: Sol Feferman <SF@CSLI.Stanford.EDU>
Subject: Matijasevitch visit
To: barwise@CSLI.Stanford.EDU, jmc@sail.stanford.edu, clt@sail.stanford.edu
Message-Id: <622511833.0.SF@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>

Because of delays in receiving his visa, the dates for his visit to
Stanford have been revised to Oct. 25-Nov.9.
Next problem, an office for him while he is here.  From what I heard last,
Math is packed, but I will check again when our admistrator returns next 
week.  What about CS or CSLI  or...?
-------

∂22-Sep-89  1700	barwise@russell.Stanford.EDU 	Re: Matijasevitch visit  
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Sep 89  17:00:50 PDT
Received: from localhost by russell.Stanford.EDU (4.0/inc-1.0)
	id AA05122; Fri, 22 Sep 89 17:01:57 PDT
Message-Id: <8909230001.AA05122@russell.Stanford.EDU>
To: sf@csli.Stanford.EDU
Cc: jmc@sail.stanford.edu, clt@sail.stanford.edu
Subject: Re: Matijasevitch visit 
In-Reply-To: Your message of Fri, 22 Sep 89 16:57:13 PDT.
             <622511833.0.SF@CSLI.Stanford.EDU> 
Address: CSLI, Stanford University, Stanford, CA 94305  (415) 723-0110
Date: Fri, 22 Sep 89 17:01:51 PDT
From: Jon Barwise <barwise@russell.Stanford.EDU>

CSLI is bursting at the seams. They are putting two students in my
office. But they might be able to arrange something for so famous a
visitor for a short period.  If it comes to that, contact Ingrid. I am
sure she will help if it is physically possible.

∂22-Sep-89  2004	CLT 	copier    
Did you get reimbursed for the ITOYA copier
76k yen -> $552.73

∂23-Sep-89  0835	harnad@elbereth.rutgers.edu 	APS/BBS Poll: Late Returns
Received: from elbereth.rutgers.edu by SAIL.Stanford.EDU with TCP; 23 Sep 89  08:34:55 PDT
Received: by elbereth.rutgers.edu (5.59/SMI4.0/RU1.1/3.04) 
	id AA29849; Sat, 23 Sep 89 10:46:55 EDT
Date: Sat, 23 Sep 89 10:46:55 EDT
From: harnad@elbereth.rutgers.edu (Stevan Harnad)
Message-Id: <8909231446.AA29849@elbereth.rutgers.edu>
To: harnad@clarity.princeton.edu
Subject: APS/BBS Poll: Late Returns


To: BBS Associates
Re: APS/BBS Poll

Although comments are still coming in, I think we have now converged on
the prevailing view of the Associateship, which can be summarized as
follows:

  (a) Do not continue to consider sponsorship unless there can be
  multiple sponsorship (from different disciplines, and international)
  and only if the editorial independence of BBS is guaranteed.

  (b) Clarify whether sponsorship would provide advantages to BBS and its
  Associateship or only to the sponsoring organizations.

I will convey this verdict to the APS.

Many thanks for your thoughtful contributions. I also note that email
is a medium with enormous potential, not only for this sort of
administrative discussion, but for the kind of scholarly interaction
that BBS is normally devoted to. I plan to pursue developing the medium
in the future. For now, it would be very helpful if Associates would:

(i) send me the email addresses of other Associates (I have only 750
out of 5000 so far); local email directories would be welcome;
(ii) urge colleagues who do not have email addresses to obtain and use
them (and to send them to BBS); posting a call for BBS Associates' email
addresses to your local B-Board or Message facility may help;
(iii) inform me of any nascent plans anywhere to implement deeper and
more powerful uses of the electronic networks for scholarly
communication.

Stevan Harnad
-----------------------------------------------------------
(53) ARTIFICIAL INTELLIGENCE/US

I am concerned about the effect joining APS would have on BBS. In
particular, your Question 2 raises a critical issue: what effect would
sponsorship by APS have on BBS's interdisciplinary status? APS is
likely to have a large membership relative to the number of active BBS
associates, and they are not likely to share BBS's interdisciplinary
interests. There are few highly-respected interdisciplinary journals,
and, I believe, that is the chief reason for BBS's popularity. It would
be wrong to jeopardize BBS on that basis.

I don't remember how Psychonomics Society fared after splitting from APA.
Is there a lesson there for how APS might posture itself after the split?
Is there a lesson there for how an affiliated journal like BBS would do?

(24b) PSYCHOLOGY/US

Thanks for the feedback on responses. I really take exception to the
implied threat in Scarr's response that APS will start a competing
journal if BBS does not accept the APS offer and her implicit
assumption that you should be replaced as editor after a few years. All
the more reason to reject the offer. BBS will gain nothing.

(54) PSYCHOLOGY/US

I can see why APS would want to sponsor BBS -- It's a marvelous journal
and well-established, well-edited. It would get them another journal
without the start-up hassles. But what does BBS get out of the deal? I
don't mean this as a hostile question. There must be advantages or you
wouldn't be considering it. But the advantages have not become clear
in the discussion so far.

(55)

My response to the APS proposal would be just about the same as the
response numbered (4) in your E-Mail message (including the final
sentence).

(56) COGNITIVE SCIENCE/US

The other responses have convinced me that it would be a bad idea to
accept the offer of APS sponsorship unless APS is going to be just one
of several single-discipline societies sponsoring BBS.

(57) COMPUTER SCIENCE/US

My first reaction is that if sponsorship of BBS by professional
associations is thought to be desirable for whatever reasons, it is
probably a good idea to avoid any bias (real or perceived) by ensuring
sponsorship by international associations in all the major
subdisciplines of cognitive and brain sciences (psychology, artificial
intelligence, philosophy, biology, neuroscience, information sciences,
cybernetics, etc.). If this is not feasible, the next best alternative
is to have BBS sponsored by a single truly interdisciplinary
international society (if there is one). If both these fail, it is
probably best to avoid sponsorship altogether (i.e., retain the current
status of BBS).

Thank you for taking the associates into confidence in this matter.

(58) PSYCHOLOGY/US

I would underscore your "question 2." What makes BBS unusual is that it
is NOT identified with a particular "discipline"--and especially
psychology. One ought to worry that if BBS became seen as a psychology
journal, less and less input from those other disciplines would be
evident--on analogy with Gresham's law!

(59) PHILOSOPHY/US

I vote no. I cannot see how BBS will be helped, and I can easily see
how it could be hurt. Once under the thumb of APS, BBS's editorial
policy could be changed, even if there are currently good intentions
not to do so. BBS is not hurting for subscribers, so the fact that APS
could supply 7000 more is not important. Furthermore, why should BBS
hook up with APS rather than say, AAAI, or the Cognitive Science
Society, or some neuroscience society or biological society? BBS is a
strong, independent journal; joining up with APS would destroy this.

(60) PSYCHOLOGY/US

I am a member of APS and hope they will be successful where APA has
failed. I would hope that their scientific areas of inquiry would
be as broad as those of BBS. However, I do not understand that
sponsorship of BBS by APS helps BBS. I do see that it gives added
credence to APS and relieves them of attempting to begin a new journal.
The danger, I worry, is that APS might then feel as if it has some
editorial control over BBS. That could be fine; but if the situation
among psychologists remains as political as is now the case, it could
be disaster for BBS.
I would offer appreciation for the offer of sponsorship and for any
advertising of BBS that APS might offer. However, I would be very
careful about control.

(44b) PSYCHOLOGY/US

Steve, This follow-up letter by Sandra Scarr actually reinforces my
initial dubiety and concern. I scarcely think Cambridge has any reason
to worry that, if BBS is not handed over to APS, the latter will start
a competing journal. Indeed, there is a blackmail tone to the very
utterance of this sort of consideration. Secondly, why should anyone
assume that the 7000 members of APS will refuse to subscribe to BBS
unless it is published under the auspices of APS? I can find absolutely
no advantage attaching to the APS affiliation and a host of
disadvantages. Indeed, it is my own sense that the core activities of
the APS luminaries and hangers-on will have a half-life far shorter
than what can be reasonably anticipated for BBS, even if the latter
continues to show "MIT-think" more respect than is warranted.

(61)

The inquiry from APS about sponsoring BBS sounds intriguing. I support
your interest in looking further into it.

(62) ARTIFICIAL INTELLIGENCE/US

Do you really want responses? (I hope you don't get inundated with
e-mail!)

BSS is a great journal because of the vision and leadership of its
editor and editorial staff. I see no reason to tamper with a winning
horse.

(63) PSYCHOLOGY/CANADA

It is too early to judge the advantages to both parties; if there are
clear advantages to both parties, they will not vanish overnight.

(64) ARTIFICIAL INTELLIGENCE/US

I agree with the first answer: What utility would this be for BBS?
(Little ) What harm could it be ? ( Considerable. ) So, forget it.

(65) PHILOSOPHY/US

BBS works pretty well now. If it ain't broke, don't fix it.

(66) PSYCHOLOGY/US

After reading your original mailing, and the first returns, I see
relatively little to justify the risks of being adopted by APS (even
though I am a member and believe that they are satisfying a significant
need in psychology today). If elaboration of their offer were to
provide significant incentives, then I can imagine being persuaded to
reconsider, but thus far the risks which were enumerated in the inital
responses to your memo seem to overwhelm the potential gains.

(67) NEUROSCIENCE/CANADA

On ne change pas une equipe qui gagne!

(68) ARTIFICIAL INTELLIGENCE/AI

I agree with all of the reservations and misgivings.

(70) NEUROSCIENCE/US

I thank you for the open exchange. It indicates a great view of your
role. I also think that BBS should avoid any connections. It will only
turn someone away. It works because it is so independent.

(71) NEUROSCIENCE/CANADA

Thank you very much for sending the APS/BBS correspondence. I fully
agree with your preliminary summary (September 21) of the 'straw poll'
and I'd simply like to add some weight to the Non- US and
Non-Psychology camp:

(i) Why change a winning team?

(ii) If there is to be BBS sponsorship, then there have to be definite
advantages. These are by no means evident in Scarr's offer.

(iii) Sponsorship, if considered desirable, should be international
and widely interdisciplinary, broadly reflecting BBS thrust and
associateship. A narrow psychological base could easily put off some of
its current readers and contributors. Multiple sponsorship: yes; APS
only: no.

(iv) As for the 'veiled threat' (10) of APS starting their own similar
journal: let them try it. It will be uphill, and they would have to
compete with quality.

(72) PSYCHOLOGY/UK

On the BBS proposal: From across the Atlantic the proposal doesn't make
any sense. It is rather an acknowledgement of the success of BBS, since
APS members must be only a fraction of BBS associates - even if one
allowed potential members - which we would not be. BBS started off as
an international journal. There are few enouigh of these. The issues
APS supports are important enough, but they are the USA's. By all means
keep politics in science, but make sure it is the right kind.

(73) LINGUISTICS/US

I just got all the interchanges concerning the bbs/aps potential
association. As a member of APS (and not a war time behavioral
experimental psychologist as we members were purported by someone to
be) I would see APS as a reasonable sponsor for BBS but clearly not a
more appropriate sponsor than it already has in CUP. I am not sure what
is to be gained for BBS by such sponsorship. What is to be gained is
obviously for APS to establish further their theoretical and
non-clinically based focus.

(74) COMPUTER VISION/CANADA

Seems to me there are very few pro's and very many con's to this proposal.
As a non-American, non-psychologist your points (2) and (3) are well-taken.

(75) PSYCHOLOGY/US

My feeling is that BBS has little to gain and much to lose from the APS
offer. APS has an uncertain future. It has made no commitment to break
with the empiricism and conservatism characteristic of APA journals
(some of which it might be about to inherit). APS sponsorship poses the
additional dangers of nationalism, undue concentration on psychology,
and cutthroat APA-style politics cited by other contributors. BBS
should stay independent. Don't stall them; turn them down flat!

(76) PSYCHOLOGY/EUROPE

I think I may not have received some previous exchange of emails on the
Sandra Scarr proposal - I guessed that APS stands for American
Psychological Society (replacing? the APA or in competition with it-
anyway my feeling is that BBS should NOT be associated with a single
society - its major strength is from its interdisciplinary character
and its international connections - so I'd tend to react negatively to
the "American" and "Psychological" parts which only leaves the
amorphous "Society"

(77) PSYCHOLOGY/US

I don't disagree with the concerns that people have raised about
sponsorship of BBS by APS, but I think there are other points to be
considered.

(i)  Sponsorship by an appropriate organization would help ensure the
continuation of the journal in the event you stepped down as editor.
Any desires you have to move on or branch out could also be dealt with
in an orderly fashion.

(ii) Association with APS might well move that fledgling organization in
the direction of the interdisciplinary approach you are trying to
promote.

(iii) I suspect you are in a position to negotiate a very flexible and
favorable agreement with APS, one that would have wide-ranging
guarantees and conditional statements of continued association.

For these reasons I think it would be worth exploring the association
if only to clarify the advantages and what such an agreement might
look like.

(78) COMPUTER SCIENCE/US

I am not crazy about this idea. I'd rather see BBS being adopted
by Cognitive Science Society.

(79) PSYCHOLOGY/US

Not clear what would be gained. Suppose organization making the offer
were old and powerful (rather than new and dubious)? The inclination to
accept would be stronger, but what would you gain? Could endorsement by
one group lead to endorsement by others, until finally every science
represented in BBS was included--and you were the official journal of
"cognition" or what have you? Would this be desirable--and would the
positives outweigh the probable negatives (such as red tape,
interference, quest for special privileges, etc.)? If this
question--for the strongest imaginable case--has no clear answser, I
see no point in toying with the weak one. On the other hand, if you are
inclined toward a positive answer by the former, then you remain the
victim of the latter, and must continue to weigh it on all imaginable
scales. But I don't see that the former has an unqualifiedly positive
answer.

(9b) (follow-up) PSYCHOLOGY/US

Reading the correspondence from your Associates raised a different
concern for me. It was not that they were hesistant about the link with
APS--it was the number who implicitly or explicitly defined BBS as a
cognitive science journal. I see it as far broader, although I must
admit that I've found the trend toward a cognitive science approach
rather strong of late. It's been at the expense of broad and/or
biological theory, which is what drew me to BBS in the first place.
Reading the responses from your Associates made me wonder if this was
still the right journal for me, so to speak.

Maybe the bias derives from the fact that cognitive science types use
their e-mail more regularly (further evidence of this can be found in
the absence of typos in messages other than mine)!

Keep it broad!

  [SH: For the time being email does tilt the balance toward the
  computer-science/AI population, but I hope that will soon be
  corrected. BBS is as strongly committed to neuroscience and behavioral
  biology as ever, and I welcome suggestions of topics and authors who
  should be encouraged to contribute; several treatments in these areas
  are now in press and in preparation. The fluctuations you see are
  those that occur in the behavioral and brain sciences as a whole,
  periodically. (Your typos have been corrected)]

∂23-Sep-89  1037	perlis@yoohoo.cs.umd.edu 	your visit    
Received: from mimsy.umd.edu by SAIL.Stanford.EDU with TCP; 23 Sep 89  10:37:04 PDT
Received: from yoohoo.cs.umd.edu by mimsy.umd.edu (5.61/4.7)
	id AA06676; Sat, 23 Sep 89 13:35:42 -0400
Received: by yoohoo.cs.umd.edu (5.61/3.14)
	id AA16713; Sat, 23 Sep 89 13:43:24 -0400
Date: Sat, 23 Sep 89 13:43:24 -0400
From: perlis@yoohoo.cs.umd.edu (Don Perlis)
Return-Path: <perlis@yoohoo.cs.umd.edu>
Message-Id: <8909231743.AA16713@yoohoo.cs.umd.edu>
To: jmc@sail.stanford.edu
Subject: your visit

John, I am resending this just to be sure you get it
before you leave town. -Don

---
Received: by yoohoo.cs.umd.edu (5.61/3.14)
	id AA14369; Thu, 21 Sep 89 21:35:52 -0400
Date: Thu, 21 Sep 89 21:35:52 -0400
From: perlis (Don Perlis)
Message-Id: <8909220135.AA14369@yoohoo.cs.umd.edu>
To: jmc@sail.stanford.edu
Subject: your visit
Cc: perlis, sarit@grinch.umiacs.umd.edu

John, it would be great to talk to you when you are in the 
DC area. My schedule is as follows:

Tues (Oct 3)
	11am - 12:15pm class
	12:30 - 1:45pm Cog Sci seminar (Intro to the brain)
	2 - 3pm office hours
	4 - 5pm Minker's Logic Programming seminar

Wed (Oct 4)
	11am - noon class

Thurs (Oct 5)
	11am - 12:15pm class
	12:30 - 1:45pm Cog Sci seminar (Cherniak on minimal rationality)
	2 - 3pm office hours
	3 - 4pm meeting

Any other times are free, and also we can probably meet during my
office hours (students rarely show up!). Moreover, you are of
course welcome to attend the seminars mentioned.

My phone numbers are:

home: (301 927-2116
office:	(301) 454-7931

-Don

P.S. I will try to have Sarit Kraus meet with us as well. I am
pleased that the "scope" notion interests you.

∂23-Sep-89  1557	srh@flash.bellcore.com 	"Skywriting"    
Received: from flash.bellcore.com by SAIL.Stanford.EDU with TCP; 23 Sep 89  15:57:30 PDT
Received: by flash.bellcore.com (5.58/1.1)
	id AA25881; Sat, 23 Sep 89 18:44:19 EDT
Date: Sat, 23 Sep 89 18:44:19 EDT
From: srh@flash.bellcore.com (stevan r harnad)
Message-Id: <8909232244.AA25881@flash.bellcore.com>
To: harnad@clarity.princeton.edu
Subject: "Skywriting"

From harnad Sat Sep 23 16:15:04 1989
To: baron@cattell.psych.upenn.edu
Subject: Re:  e-mail addresses

	Date: Sat, 23 Sep 89 13:48:20 EDT
	From: baron@cattell.psych.upenn.edu (Jonathan Baron)
	
	I think that e-mail is great for research collaboration, for
	interdepartmental discussion in lieu of meetings, for sending
	in journal reviews, for general discussions within small groups
	of people, for local news, and even for sending out abstracts
	requesting reviewers, as you do, but I'm skeptical about other
	uses that involve, for example, free-for-all discussions of some
	scholarly issue (in the format of the letters concerning BBS and
	APS).  It gets to be too much to read.  As I think you've learned
	over the years, editors have a valuable function, sifting the
	wheat from the chaff and improving the quality of the former.

Jonathan,

Thanks for sharing your reflections on limits of email discussion.
I have a vision, however, of what email discussion COULD be -- if
conducted at a sufficiently sophisticated level, with certain
constraints, by the best minds in the fields in question -- and I
hope to be able to implement the vision eventually. In my opinion
(and experience), the speed, flexibility and scale of multiple
email network communication (a fantasy I now call "skywriting") is
potentially so powerful and congenial to a creative thinker's scale
and tempo of communication and interaction that it may represent
just as revolutionary a leap in human communicative (and creative)
potential as its three momentous predecessors (the advent of
language, writing and printing).

Don't form too firm a pessimistic conclusion from what you see on
the Net now: For mere historical and technological reasons it's
currently dominated by computer scientists and students (rather as
if not just the form but the contents of the printed page had been
dominated for a century by Guttenberg and a legion of linotype
operators) and shunned by a more mature generation in all other
disciplines, whose majorities are still computer-phobic. Try instead
to imagine what it would be like if the strongest thinkers in all
the disciplines were on the Net now, and used it concertedly.

There's no great problem with sifting the sky-wheat from the
sky-chaff:  Wherever necessary a discussion group could restrict
access to read-only for all but the qualified; and one can always
protect oneself from unwanted discussion by one keystroke -- or
even pre-filters in your software. Wasted forests of paper and junk
mail choking your mail-box are not at issue. The facilities for
scanning, filtering and discarding unwanted skywriting are even
more powerful than the facilities of skywriting itself!

But what you have to sample for yourself are the remarkable rewards of
skywriting on topics that are currently of pressing scientific interest
to you. I can't say I've done this with the best minds in the field yet
(not enough of them are on the Net, to date). But then it's perhaps all
the more remarkable that even with the Net's current demography I can
report that lately my own interactions via skywriting have advanced my
ideas more than conventional reading/writing, speaking/listening
combined. Imagine getting a provocative idea or experimental finding,
posting it at 9 am, and already receiving feedback from the relevant
parts of the world scholarly community within minutes: "Here's a
logical problem; here's some contrary evidence; here's some supporting
data; here are some surprising ramifications; so-and-so has already
reported this... etc." Sure, there's a lot of potential information from
such a source, but what earnest thinker would rather turn AWAY from
it?

"Editing" is really a software matter at this "pilot" stage of inquiry,
conducted in the skywriting medium. Selecting and archiving what turns
out to be lapidary in this evolution of ideas is another matter, but
we're talking about improving the flow right now. Are you really so
satisfied with the current scale, scope and rate of generation and
communication of ideas through the conventional media of scholarly
communication? -- And might one not have hesitated at the advent
of printing too ("There's already too much to read...")!

The possibilities of skywriting are an empirical matter, but I'm going
to do my level best to make sure they are properly tested, rather than
pre-emptively discarded or passed over because of the skewed initial
conditions.

Thanks for the opportunity to formulate my position for the
Associateship!

Stevan Harnad


∂23-Sep-89  1632	perlis@yoohoo.cs.umd.edu 	Oct 5th  
Received: from mimsy.umd.edu by SAIL.Stanford.EDU with TCP; 23 Sep 89  16:31:52 PDT
Received: from yoohoo.cs.umd.edu by mimsy.umd.edu (5.61/4.7)
	id AA11826; Sat, 23 Sep 89 19:30:05 -0400
Received: by yoohoo.cs.umd.edu (5.61/3.14)
	id AA17411; Sat, 23 Sep 89 19:37:49 -0400
Date: Sat, 23 Sep 89 19:37:49 -0400
From: perlis@yoohoo.cs.umd.edu (Don Perlis)
Return-Path: <perlis@yoohoo.cs.umd.edu>
Message-Id: <8909232337.AA17411@yoohoo.cs.umd.edu>
To: jmc@sail.stanford.edu
Subject: Oct 5th
Cc: perlis@yoohoo.cs.umd.edu

The 5th would be fine. Chances are I can even give you
a ride to the airport.

Don

∂23-Sep-89  1801	harnad@clarity.Princeton.EDU 	Symbols and Connections  
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 23 Sep 89  18:01:44 PDT
Received: from reason.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA03636; Sat, 23 Sep 89 20:58:16 EDT
Received: by reason.Princeton.EDU (4.0/1.85)
	id AA03113; Sat, 23 Sep 89 20:58:14 EDT
Date: Sat, 23 Sep 89 20:58:14 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8909240058.AA03113@reason.Princeton.EDU>
To: jaf@cunyvms1.BITNET
Subject: Symbols and Connections

To:   Jerry Fodor
From: Stevan Harnad
CC:   Some interested parties
Re:   Your critique of the connectionist model of mind

Jerry, I enjoyed your talk yesterday. Having had a chance, no doubt,
to sample more of the prima facie objections people tend to come up
with, you seemed ready now to take on all comers. 

(1) For a change, I basically agree that systematicity comes naturally and
for free in symbol systems (SSs) and not in connectionist systems (CSs),
and that it seems a nice property to have if you're modeling the
mind. My first point was only that it is to invite misunderstanding, in
comparing SSs and CSs, to speak of a string of symbols on its tape as a
"state" of an SS and the on/off state of a node as a "state" of a CS.
That almost begs for misunderstandings about local/distributed states
and even what a "state" of a system is.

Nothing would be lost, and the punchline would be the same, if you
allowed that a state of a SS, the putative carrier of the meaning,
may include the entire machine state -- what's on the tape now, what's in
the machine table, the current "state" of all its flip-flops etc. This
is still implementation-independent, because the same state could have
been realized by countless machines, in countless ways (as long as
systematic semantic interpretability was preserved). But the current
state is how it's realized now. Just the "0" inscription on the tape is
not the SS's "state." In fact, in another machine, one that was in another
syntactic state, the very same local strip of tape with the "0" on it
would be part of another SS and may not even be the same symbol, let
alone the same symbolic state (or the same meaning).

Likewise with the node in the CS. The state of the CS is not just the
state of a node (even if that's the node whose activity is being
interpreted as meaning that P). The node could only mean that P (if
indeed it does mean that P) in virtue of its connectivity with he
rest of the nodes in the system, and their respective weights and
activations. Again, the state that means P may include the entire
machine state.

In this (4th?) sense, machine states, whether SS or CS, are always
"distributed" over the entire system. Put another way, the REALIZATION
of a symbol token in an implemented SS is distributed, not
restricted to the inscription under the tape-head, and so is the
realization of the putative "symbol" in a CS. But now, as you
correctly argued, the putative "symbol" state in the CS turns out not
to have the systematic properties of the symbol state of the SS; in
particular, it will not have the requisite token-state/subtoken-state
relation you singled out (P&Q vs. P), so your argument still goes
through.

(2) My second point was a bit more critical, because I don't think the
symmetry of the learnability of an A>B and a B>A discrimination really
demonstrates the kind of nontrivial systematicity that will do the work in
the case of animals that language/thought systematicity does in the
case of humans in showing that the mind (verbal and nonverbal) is
systematic. For one thing, if you took the simplest, generic,
completely-interconnected CS system -- WITHOUT TAKING ANY STEPS
WHATSOEVER TO CAPTURE "SYSTEMATICITY" -- it would be quite natural for
it to be equally able to learn A>B or B>A by default. Ineed, you'd have
to build special contraints into it in order to AVOID having this quite
natural symmetry. And this in a CS system that is virtually the
prototype of one that lacks systematicity. So this kind of simple
default symmetry cannot be the requisite evidence for the kind of
systematicity that shows that animal minds to must be SSs rather than
CSs.

There may be real examples, but they're more likely to be in the
domain of what OTHER capacities you inherit when you learn a given
discrimination (e.g., transitivity), but these tend to be difficult to
demonstrate in animals.

(3) Finally, and this is perhaps the most substantive point -- and
begins to impinge on the problem of symbol grounding, where we DO
diverge quite markedly -- I don't think that the difference between
what a CS gets by theft (hard-wiring) vs. honest toil (learning) is
anywhere near as inconsequential as you think it is. First, note that
you yourself are making a similar distinction in your argument on
behalf of the SS model of mind, but in reverse: You stress that there
are things you get with a SS for free (e.g., systematicity), whereas in
a CS they would require heroic measures (whether through hard-wiring or
learning) to get them, and that this is an important difference, one
that strongly favors SSs (and I agree).

But, by symmetry, a CS advocate should be able to say that surely
there's an important difference between a CS that must be hard-wired in
advance in order to exhibit a certain property (heroic measures) and
one that can learn them, ESPECIALLY if that property happens to be
systematicity! Note that I'm not insisting that CSs CAN be trained up
to become SSs (although some connectionists are claiming this), just
that it's an open question (until someone shows it's possible, by doing
it, or proves formally that it's impossible).

But given this possibility, it seems much too hasty to say, "Whether a
CS becomes an SS by theft or by honest toil, either way, I win, because
my claim was that only an SS could do the job (of being a mind)." For
note that, if there's any empirical content to this discussion at all
(and I, for one, don't think Turing gave us much in the way of "theory
of mind" if he gave us anything at all: Theories, after all, are
supposed to do the job, and I don't yet see Turing machines passing
the Turing Test...) then it must have to do with what CSs can or cannot
DO, qua CSs. And whereas I agree completely that a CS that must be
hard-wired in such a way as to become just another implementation of
an SS (for doing Lisp, say) in no way attests to CSs as models of
the mind (or as bearers of systematicity), it is not at all clear to me
that the same would be true of a CS that could be trained to exhibit the
systematicity of language and logic by the right regimen of experience
(and the right internal wherewithal to exhibit this learning capacity).
For that learned sysetmaticity would surely be something that the CS --
that very same CS, qua CS -- COULD do; that CS would in a real sense be
doing the job! And if so, it would no longer be so obvious that if you
want to model the mind, you'd better head straight for an SS.

But a pox on both their houses anyway. As you know, I don't believe
that a pure SS could have a mind (or intrinsic meaning) because of the
symbol grounding problem. A hard-wired CS that was really just an
implementation of an SS would be in the same boat. And although I
don't know whether or not a vanilla CS can be trained into an SS, I
suspect that that alone wouldn't be enough for it to have a mind either
-- or, at any rate, there would have to be more to this "training" then
just become a Lisp machine. I think a lot of behavioral capacity (I
dare not call it "cognitive" lest you bash me on the head with
systematicity, etc.), including discrimination and identification
(i.e., categorization), is NONsymbolic, and that the rest must be
grounded in this. The mechanisms underlying discrimination and
especially categorization are the ones that allow us to pick out the
objects that our primitive symbols then refer to. Without this link to
objects, symbols are just hanging by skyhooks, be they ever so
systematic.

If the TBA is not filled and there's interest, this is what I would
talk about on October 20th.

Cheers, Stevan

∂24-Sep-89  1345	scherlis@vax.darpa.mil 	Announcement    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 24 Sep 89  13:45:06 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA03971>; Sun, 24 Sep 89 16:09:05 -0400
Posted-Date: Sun 24 Sep 89 16:03:15-EDT
Received: by sun45.darpa.mil (4.1/5.51)
	id AA02173; Sun, 24 Sep 89 16:03:19 EDT
Date: Sun 24 Sep 89 16:03:15-EDT
From: William L. Scherlis <SCHERLIS@DARPA.MIL>
Subject: Announcement
To: SW-PI@vax.darpa.mil
Message-Id: <622670595.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>



	We expect that Barry Boehm will start his term as Director
	of ISTO in November.  In the meantime, Stephen Squires is
	Acting Director.  

	Over the next few weeks, I expect to be sending you 
	summaries of the ISTO Software Program activity and
	plans for your comments and suggestions.

-------

∂25-Sep-89  0800	menke@harvard.harvard.edu 	talk at harvard, october 2  
Received: from harvard.harvard.edu by SAIL.Stanford.EDU with TCP; 25 Sep 89  08:00:25 PDT
Received: by harvard.harvard.edu (5.54/a0.25)
	(for JMC@SAIL.Stanford.EDU) id AA13952; Mon, 25 Sep 89 11:03:15 EDT
Received: by endor.HARVARD.EDU; Mon, 25 Sep 89 11:03:10 EDT
Date: Mon, 25 Sep 89 11:03:10 EDT
From: menke@harvard.harvard.edu (Baiba Menke)
To: JMC@SAIL.Stanford.EDU
Subject: talk at harvard, october 2


In order to process an honoraria for you I need you
home address and social security number.  If I could
have it ASAP it would be greatly appreciated.

Thank you.

Baiba Menke

∂26-Sep-89  1629	CLT 	sail users
Since I have to do the bookkeeping,
please inform me when you want to give someone an account.
Note that added users cost in bookkeeping effort and
will in general incur printing charges which we must either eat
or charge to some other account.

∂26-Sep-89  1841	poser@crystals.STANFORD.EDU 	curious advertisement
Received: from crystals (crystals.Stanford.EDU) by SAIL.Stanford.EDU with TCP; 26 Sep 89  18:40:31 PDT
Received: by crystals ; Tue, 26 Sep 89 18:41:00 pdt
Date: Tue, 26 Sep 89 18:41:00 pdt
From: Bill Poser <poser@crystals.STANFORD.EDU>
Message-Id: <8909270141.AA05944@crystals>
To: jmc@sail
Subject: curious advertisement


	I thought you'd get a kick out of an ad that I saw in London
this morning:

	Close your eyes and think of England.
	John McCarthy has done nothing else for three years.

The first line is familiar, but I have no idea what the sign as a whole
refers to.

						Bill 

∂27-Sep-89  0923	@po2.andrew.cmu.edu:jhm+@andrew.cmu.edu 	Re: Networks Considered Harmful - For Electronic Mail 
Received: from po2.andrew.cmu.edu by SAIL.Stanford.EDU with TCP; 27 Sep 89  09:23:17 PDT
Received: by po2.andrew.cmu.edu (5.54/3.15) id <AA03429> for JMC@sail.stanford.edu; Wed, 27 Sep 89 12:23:20 EDT
Received: via switchmail; Wed, 27 Sep 89 12:23:03 -0400 (EDT)
Received: from foo.expres.cs.cmu.edu via qmail
          ID </afs/andrew.cmu.edu/service/mailqs/q002/QF.IZ8DCv600hl=E0Jp0v>;
          Wed, 27 Sep 89 12:22:26 -0400 (EDT)
Received: from foo.expres.cs.cmu.edu via qmail
          ID </afs/andrew.cmu.edu/usr12/jhm/.Outgoing/QF.cZ8DBoG00hl=EBoQ1e>;
          Wed, 27 Sep 89 12:21:08 -0400 (EDT)
Received: from Messages.7.10.N.CUILIB.3.45.SNAP.NOT.LINKED.foo.expres.cs.cmu.edu.rt.r3
          via MS.5.6.foo.expres.cs.cmu.edu.rt_r3;
          Wed, 27 Sep 89 12:20:59 -0400 (EDT)
Message-Id: <cZ8DBfy00hl=4BoPpw@andrew.cmu.edu>
Date: Wed, 27 Sep 89 12:20:59 -0400 (EDT)
From: Jim Morris <jhm+@andrew.cmu.edu>
To: telecom@eecs.nwu.edu
Subject: Re: Networks Considered Harmful - For Electronic Mail
Cc: John McCarthy <JMC@sail.stanford.edu>

I think John's message was very important -- a sort of wake-up call for
the computer community.

> Excerpts from internet.telecom: 18-Aug-89 Networks Considered Harmful..
> John McCarthy@sail.stanf (9146)

> However, unless email is freed from dependence on the networks, I
> predict it 
> will be supplanted by telefax for most uses in spite of its many
> advantages 
> over telefax.
I believe email will be supplanted by FAX -- period. We will eventually
end up with a hybrid, but it will be achieved by the FAX business
assimilating all the knowledge we have about email.

> These advantages include the fact that information is 
> transmitted more cheaply as character streams than as images.
> Group IV compression brings the image vs. ASCII ratio down to about 5. 

> Multiple addressees are readily accommodated.
FAX store and forward services like MCI will provide this.

>  Moreover, messages transmitted as character streams can be readily
> filed, searched, edited and used by computer programs.
OCR can work for the searching part.  99% character recognition rates
are common. There are already products available that scan, recognize,
and index documents for you. The key idea is that the image is saved
too, so there is no danger of the  1% missed characters causing problems
other than missed retrieval.

As for editing, very often one wants only to annotate another document.
This can be done on the image. If one really wants to edit a document,
OCR plus some hand massaging may suffice.

> The reason why telefax will supplant email unless email is separated 
> from special networks is that telefax works by using the existing
> telephone 
> network directly.
Yes!!!

> Fax has another advantage that needs to be matched and can be 
> overmatched.  Since fax transmits images, fully formatted documents can
> be 
> transmitted.  However, this loses the ability to edit the document. 
> This can 
> be beaten by email, provided there arises a widely used standard for 
> representing documents that preserves editability.
This is a very big proviso. There is great chaos in this area right now.
The standard proposed by CCITT, called Office Document Architecture
(ODA), is getting very little support in the US where the DoD seems to
be promoting SGML and the commercial document editor vendors are
promoting their own proprietary standards. MicroSoft's Rich Text Format 
(RTF)  seems most promising since it is used by more than one document
processor. Another hope is that a single vendor, e.g. DEC with it's
ODA-related DDIF and DECWrite (=Framemaker),  will become the market
leader and establish a de facto standard, as Lotus did for spread sheets.

A much more likely development is that PostScript becomes the exchange
standard. It is there. All document processors will produce it. It looks
a little nicer than FAX, and there is at least a fighting chance that
one can translate it back into a particular document processor's
internal format.

Another advantage of FAX you failed to emphasize is simply that it deals
with pictures effortlessly. Even if you and I have precisely the same
computing equipment and are on the ArpaNet, the fastest way for me to
get a picture to you is FAX. This is true even if the picture is hand
drawn -- drawing it on paper is faster than any drawing editor I've ever
used.


> Fortunately, there is free enterprise. Therefore, the most likely way 
> of getting direct electronic mail is for some company to offer a piece
> of 
> hardware as an electronic mail terminal including the facilities for
> connecting
> to the current variety of local area networks (LANs). The most likely
> way for 
> this to be accomplished is for the makers of fax machines to offer ASCII 
> service as well. 

An AppleFAX modem will already do this for Apple PICT files. I would
like to see Adobe do the same for PostScript files.

> This will obviate the growing practice of some users of fax 
> of printing out their messages in an OCR font, transmitting them by fax,
> whereupon the receiver scans them with an OCR scanner to get them back
> into 
> computer form.

Why should this practice be obviated? Why not work at making OCR more
effective? In a race between clever computer hackers trying to make OCR
better and institutional politicians trying to straighten out the
standards who do you think will win? Which would you rather be?




Jim.Morris@andrew.cmu.edu
412 268-2574
FAX: 412 681-2066

[An Andrew ToolKit view (a raster image) was included here, but could
not be displayed.]

∂27-Sep-89  0926	boyland@sequoia.Berkeley.EDU 	looking for references about prototyping
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 27 Sep 89  09:26:23 PDT
Posted-Date: Wed, 27 Sep 89 09:14:19 PDT
Received: from sequoia.Berkeley.EDU by vax.darpa.mil (5.61/5.61+local)
	id <AA14396>; Wed, 27 Sep 89 12:14:57 -0400
Received: from redwood.Berkeley.EDU by sequoia.Berkeley.EDU (4.1/SMI-4.0)
	id AA16767; Wed, 27 Sep 89 09:14:19 PDT
Date: Wed, 27 Sep 89 09:14:19 PDT
From: boyland@sequoia.Berkeley.EDU (Mark Hastings)
Message-Id: <8909271614.AA16767@sequoia.Berkeley.EDU>
To: cps@vax.darpa.mil
Subject: looking for references about prototyping

I am looking for references about software prototyping and
programming languages designed for prototyping.
If you know of any such reference list (or even better,
can send such a list to me), I would be very grateful to learn about it.

John Boyland
(boyland@sequoia.Berkeley.EDU)

∂27-Sep-89  1126	boyland@sequoia.Berkeley.EDU 	looking for references about prototyping
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 27 Sep 89  11:26:04 PDT
Posted-Date: Wed, 27 Sep 89 09:14:19 PDT
Received: from sequoia.Berkeley.EDU by vax.darpa.mil (5.61/5.61+local)
	id <AA14396>; Wed, 27 Sep 89 12:14:57 -0400
Received: from redwood.Berkeley.EDU by sequoia.Berkeley.EDU (4.1/SMI-4.0)
	id AA16767; Wed, 27 Sep 89 09:14:19 PDT
Date: Wed, 27 Sep 89 09:14:19 PDT
From: boyland@sequoia.Berkeley.EDU (Mark Hastings)
Message-Id: <8909271614.AA16767@sequoia.Berkeley.EDU>
To: cps@vax.darpa.mil
Subject: looking for references about prototyping

I am looking for references about software prototyping and
programming languages designed for prototyping.
If you know of any such reference list (or even better,
can send such a list to me), I would be very grateful to learn about it.

John Boyland
(boyland@sequoia.Berkeley.EDU)

∂27-Sep-89  1511	cblpf!mark@att.att.com 	Networks Considered Harmful - For Electronic Mail  
Received: from arpa.att.com ([192.20.225.1]) by SAIL.Stanford.EDU with TCP; 27 Sep 89  15:11:23 PDT
Date: Wed, 27 Sep 89 16:20 EDT
From: mark@cblpf.att.com (Mark R Horton)
To: JMC@sail.stanford.edu
Subject: Networks Considered Harmful - For Electronic Mail

I think X.400 needs to be a part of the solution here.  While X.400 is
an awful, klunky, standard, it has key features that count:

(1) It is an international standard, and lots of email companies are
    supporting it and pushing it.
(2) It has provisions for phone numbers in the name/address structure
(3) It has provisions for more than just text, but also documents and
    is extensible to things like voice, bitstreams, etc.

If someone would

(a) define a standard for X.400 over dialup (probably involving TP4 on
    top of dialup, possibly with X.25 layers in between), including
    the initial handshake
(b) produce a cheap set of products, such as a FAX machine, a gateway,
    and a standalone terminal

then it might catch on.

I don't see the point of using the FAX modem.  Conservation of phone lines
is important, and will be a reason why neither email nor FAX will show up
in many homes.  Most people won't install a dedicated phone line.  If one phone
line can be shared for several purposes, it will catch on more.  When the
product dials, it should SPEAK a request to be attached to an email machine,
for the benefit of a human who answers, in between trying modem handshakes.
(This should be true of all modems, and the calling modem should identify
itself with speech so that accidental crank calls can be traced and stopped.
Automatic scripts containing typos can be a real problem, and if you have
users typing in phone numbers all the time there will be lots of errors.)

It's quite reasonable for the initial handshake to involve some ASCII which
might involve logging into a UNIX box.  The problem with UUCP is not that
the login isn't easy to do, but that, for security reasons, the conversation
is not standardized, but varies for each pair of machines.  If you adopt a
convention that anyone can drop mail off on your machine (without any
authentication) you can easily standardize on something as simple as
	expect	login
	send	email\r
and be dropped into a shell that will accept email and nothing else.
We have a similar approach here for UUCP, and it's 99% functional and
100% secure.  What it doesn't let you do is pick up mail queued for you
on the called machine, nor does it authenticate who sent the mail.

If you want this standarized, get IEEE to do it, not ACM.  IEEE is already
in the standards business, and ANSI and ISO listen to them.

	Mark Horton
	Mark.Horton@ATT.COM

∂27-Sep-89  1800	harnad@clarity.Princeton.EDU 	Searle's Problem vs. Fodor's Problem    
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 27 Sep 89  17:59:07 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA02432; Wed, 27 Sep 89 10:53:48 EDT
Received: by psycho.Princeton.EDU (4.0/1.85)
	id AA01530; Wed, 27 Sep 89 10:44:28 EDT
Date: Wed, 27 Sep 89 10:44:28 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8909271444.AA01530@psycho.Princeton.EDU>
To: srh@flash.bellcore.com
Subject: Searle's Problem vs. Fodor's Problem

Searle's Problem vs. Fodor's Problem

To:    Steve Stich
From:  Stevan Harnad

Both the symbolic (S) and the connectionistic (C) approaches to modeling
the mind seem to suffer from their own respective fatal handicap:

S suffers from Searle's Problem: Symbols have no intrinsic meaning,
they're ungrounded; their meanings are parasitic on the meanings in our
heads, which clearly do have intrinsic meaning.

C suffers from Fodor's Problem: Connectionist "representations" lack
systematicity, unlike the meanings in our heads, which clearly do have
systematicity.

My proposal is a very particular kind of hybrid approach in which C is
given only the limited and nonrepresentational role of feature
learning, a role to which it is naturally suited. The need for
systematicity (Fodor's Problem) never arises for C. S then enters as a
DEDICATED symbol system (one who's primitive symbol-tokens have
additional nonsymbolic constraints on them). The nonsymbolic
constraints are what GROUND S (thereby avoiding Searle's Problem)
through the connections between the primitive symbols tokens and the
feature-detectors that pick out the objects to which they refer from
their sensory projections.

I can't guess what the "3 Nativisms" for C will be that you will be
discussing this friday, but the question of learning and learnability
is clearly critical in all this. Fodor is satisfied with a radical
nativism for most of our concepts. That's not surprising, because he
accepts the "vanishing intersections" argument against the existence
of critical features (especially sensory ones) that pick out objects.
I think C may allow the first actual TEST of whether feature
intersections really vanish; I don't think that is decidable from the
armchair. In any case, whether feature-learning took place during
evolution or takes place during the lifetime of an organism does not
much matter (the answer is probably that there is some of each). What
matters is whether features are learnable at all. I'm still betting
they are, and that our sensory and conceptual categories are not just
"spandrels."

Stevan Harnad

∂27-Sep-89  1939	hayes@kanga.parc.xerox.com 	Vlad Dabija 
Received: from arisia.Xerox.COM by SAIL.Stanford.EDU with TCP; 27 Sep 89  19:39:49 PDT
Received: from kanga.parc.Xerox.COM by arisia.Xerox.COM with SMTP
	(5.61+/IDA-1.2.8/gandalf) id AA18231; Wed, 27 Sep 89 19:35:37 -0700
Received: by kanga.parc.xerox.com
	(5.61+/IDA-1.2.8/gandalf) id AA03224; Wed, 27 Sep 89 19:39:52 PDT
Message-Id: <8909280239.AA03224@kanga.parc.xerox.com>
Date: Wed, 27 Sep 89 19:39:52 PDT
From: Pat Hayes <hayes@parc.xerox.com>
To: hemenway@Polya.Stanford.EDU, JMC@sail.stanford.edu,
        gio@sumex-aim.stanford.edu, mas@sail.stanford.edu,
        eaf@sumex-aim.stanford.edu, AI.Lenat@MCC.com,
        ullman@nimbin.stanford.edu, Nilsson@Polya.Stanford.EDU
Cc: hayes@parc.xerox.com
Subject: Vlad Dabija


This Romanian guy in whom you expressed a possible interest is
planning to come up to Stanford, hoping to talk to anyone he can, on
Monday 6 November, the day before Election day.  If you can spare some
time for him then, please mark your calendars.

If you could email me the times you are likely to be free, so he can
schedule his day, Id be grateful.

Many thanks.

Pat Hayes
(hayes.pa@xerox.com)

∂28-Sep-89  0902	rh@mbunix.mitre.org 	Some References on Prototyping    
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 28 Sep 89  09:01:57 PDT
Posted-Date: Thu, 28 Sep 89 11:52:20 EDT
Received: from mbunix.mitre.org by vax.darpa.mil (5.61/5.61+local)
	id <AA00215>; Thu, 28 Sep 89 11:54:40 -0400
Date: Thu, 28 Sep 89 11:52:20 EDT
From: rh@mbunix.mitre.org (Rich Hilliard)
Message-Id: <8909281552.AA05467@mbunix.mitre.org>
Posted-From: The MITRE Corp., Bedford, MA
X-Alternate-Route: user%node@mbunix.mitre.org
To: boyland@sequoia.berkeley.edu
Cc: cps@vax.darpa.mil
In-Reply-To: Mark Hastings's message of Wed, 27 Sep 89 09:14:19 PDT <8909271614.AA16767@sequoia.Berkeley.EDU>
Subject: Some References on Prototyping


I thought this list was dead, so I'll reply, just to see if anything
happens.  This list is just things I could remember off the top of my
head -- my References STAK is on my other machine!


* Apple's HyperCard

* Goldberg and Robeson's books on Smalltalk-80, Addison-Wesley, 1984

* Interactive Programming Environments (Barstow, Shrobe and Sandewall,
eds.) McGraw-Hill, 1984

* ACM Software Engineering Notes (December 1982?), Workshop on Rapid
Prototyping.

* The three OOPSLA Proceedings (in ACM SIGPLAN 1986, 1987, 1988, soon
to be 1989, too)


IF this list isn't dead, is there anyone who wants to discuss CPS/CPL?

-- Rich 



Richard F. Hilliard II
The MITRE Corporation (M/S A156)
Bedford, MA  01730

617 271-7760
617 271-8752 (FAX)

INTERNET: rh@mbunix.mitre.org
UUCP: ...{att,decvax,genrad,ll-xn,philabs,utzoo}!linus!rh


		  "CPL is a friend function of Ada."

∂28-Sep-89  1011	smu!tanik@uunet.UU.NET 	Re:  looking for references about prototyping 
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 28 Sep 89  10:11:18 PDT
Posted-Date: Thu, 28 Sep 89 11:57:36 CST
Received: from uunet.UU.NET by vax.darpa.mil (5.61/5.61+local)
	id <AA00434>; Thu, 28 Sep 89 13:03:34 -0400
Received: from smu.UUCP by uunet.uu.net (5.61/1.14) with UUCP 
	id AA18742; Thu, 28 Sep 89 13:03:20 -0400
Received: by smu.edu (5.51/5.0)
	id AA00544; Thu, 28 Sep 89 11:57:36 CST
Date: Thu, 28 Sep 89 11:57:36 CST
From: smu!tanik@uunet.UU.NET (Murat Tanik)
Message-Id: <8909281757.AA00544@smu.edu>
To: cps@vax.darpa.mil, sequoia.Berkeley.EDU!boyland@uunet.uu.net
Subject: Re:  looking for references about prototyping

Please contact Prof. Luqi at Naval Post graduate school.
She will be able to help you.

If you explain your intentions more precisely, I may be
able to help you as well.

Regards
Tanik

∂29-Sep-89  0851	@REAGAN.AI.MIT.EDU:dam@AI.MIT.EDU 	The inverse method  
Received: from REAGAN.AI.MIT.EDU by SAIL.Stanford.EDU with TCP; 29 Sep 89  08:50:50 PDT
Received: from ELVIS.AI.MIT.EDU by REAGAN.AI.MIT.EDU via INTERNET with SMTP id 264466; 29 Sep 89 11:52:27 EDT
Date: Fri, 29 Sep 89 11:51 EDT
From: David A. McAllester <dam@AI.MIT.EDU>
Subject: The inverse method
To: val@sail.stanford.edu
cc: JMC@sail.stanford.edu, RLG@sail.stanford.edu
Message-ID: <19890929155139.9.DAM@ELVIS.AI.MIT.EDU>

I am teaching a course on AI and have been studying resolution and other
standard theorem proving techniques.  Out of curiousity, I read a paper
of yours entitled ``What is the inferse method?''.  I have a draft copy
dated August 1986.

In trying to understand this paper I came to the conclusion that the
inverse method can best be described as a transformation process that
maps a given resolution problem to a different, potentially more
tractible, resolution problem.

      ******* The Maslov Transform ******

Let Phi be Skolem normal form formula, i.e., a resolution problem.  To
construct the Maslov transform of Phi we first convert the matrix of Phi
to DNF.  Let C1 ... Cn be the conjunctions in the DNF form of the matrix
of Phi.  For each conjuction Ci we introduce a new predicate symbol Pi
where the number of arguments (arity) of Pi equals the number of
variables that appear free in Di.  Let Li be the atomic formula
consisting of Pi applied to the variables that appear free in Ci. Each
literal Li ``represents'' the conjunction Ci.  The Maslov transform
M(Phi) of the original problem Phi consists of the following clauses:

Rule A.  Consider a pair Ci, Cj of conjunctions from Phi'.  A
subsitution s will be called incompatible with Ci and Cj if there exist
literals A and B in Ci and Cj respectively such that s(A) is the
opposite literal from s(B) (implying that s(Ci ↑ Cj) is unsatisfiable).
By considering all such pairs A and B we can construct a finite set of
most general subsitutions that are incompatible with Ci and Cj.  For
each such substitution s construct the clause [-s(Li) or -s(Lj)].

Rule B.  Add the clause [L1 or L2 or ... L3].

Maslov's theorem: Phi is satisfiable if and only if M(Phi) is satisfiable.

     *****************

It is probably not obvious (and possible not true) that the inverse
method corresponds to the transform defined above.  First, note that the
inverse method addresses the validity of an existential CNF formula, or
equivalently, the satisfiability of a universal DNF
formula. The Ci's above are simply the negations of the Di's of
the inverse method. A pair (i;s) of the inverse method represents the
formula s(Di), or in the above notion, the literal -s(Li).  An
assemblage {(i;s), (j;w), ... (k;z)} is simply a representation of the
clause (-s(Li) or -w(Lj) or ... or -z(Lk)).  An assembledge is favorable
if it can be derived from clauses A and B above by hyper-resolution.

It is well known that any formula can be linear-time converted to an
equisatisfiable Skolem normal form formula.  Satisfiablity-preserving
convertion to DNF, however, is inherently exponential since such conversion
can be used to solve propositional satisfiability.  This implies that
the time required to construct the Maslov transform is exponential
in the worst case.

I will end with some questions.  Is the transform view of the inverse
method well known?  For any resolution problem S, let M(S) be the Maslov
transform of S.  Is there any advantage in computing M(M(S)), e.g., does
iterating the Maslov transform yield yet more insight into decision
problems?

I'm interested in your thoughts,

	David McAllester

∂29-Sep-89  1056	scherlis@vax.darpa.mil 	Awards
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 29 Sep 89  10:56:31 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.61/5.61+local)
	id <AA05144>; Fri, 29 Sep 89 13:44:13 -0400
Posted-Date: Fri 29 Sep 89 13:43:58-EDT
Received: by sun45.darpa.mil (4.1/5.51)
	id AA00976; Fri, 29 Sep 89 13:43:59 EDT
Date: Fri 29 Sep 89 13:43:58-EDT
From: William L. Scherlis <SCHERLIS@DARPA.MIL>
Subject: Awards
To: CPS@vax.darpa.mil
Cc: avantil@nswc-wo.arpa, squires@vax.darpa.mil
Message-Id: <623094238.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>


	We expect to be announcing shortly the results of the
	CPS/CPL solicitation of last Spring.  Annoucements
	and other information will be distributed electronically
	through this mailing list (and also available directly
	from DARPA/ISTO in hardcopy form).  

-------

∂29-Sep-89  1252	balzer@vaxa.isi.edu 	Prototyping Panel at OOPSLA  
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 29 Sep 89  12:52:24 PDT
Received: from vaxa.isi.edu by vax.darpa.mil (5.61/5.61+local)
	id <AA06175>; Fri, 29 Sep 89 15:49:40 -0400
Posted-Date: Fri, 29 Sep 89 12:49:29 PDT
Message-Id: <8909291949.AA01863@vaxa.isi.edu>
Received: from LOCALHOST by vaxa.isi.edu (5.61/5.61)
	id AA01863; Fri, 29 Sep 89 12:49:31 -0700
To: CPS@vax.darpa.mil
From: balzer@VAXA.ISI.EDU
Subject: Prototyping Panel at OOPSLA
Date: Fri, 29 Sep 89 12:49:29 PDT
Sender: balzer@vaxa.isi.edu

Starting with the criteria set for in the CPS report, this panel will
explore the adequacy, or lack thereof, of existing Object Oriented systems
in supporting Prototyping, and suggest remedies for their shortcomings. 
Of particular concern is the ability of such systems to postpone 
committments.

The panel consists of:
    Robert Balzer, ISI (moderator)
    Peter Deutsch, ParcPlace Systems
    Richard Gabriel, Lucid
    Stan Zdonik, Brown University

The panel will be held at OOPSLA in New Orleans on Thursday morning Oct. 5
at 7:00am (yes, seven o'clock in the morning!)

∂29-Sep-89  2253	ungar@self 	Re:  looking for references about prototyping   
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 29 Sep 89  22:53:35 PDT
Posted-Date: Fri, 29 Sep 89 22:52:39 PDT
Received: from self.Stanford.EDU by vax.darpa.mil (5.61/5.61+local)
	id <AA07521>; Sat, 30 Sep 89 01:52:35 -0400
Received: by self (4.0/inc-1.0)
	id AA00438; Fri, 29 Sep 89 22:52:39 PDT
Date: Fri, 29 Sep 89 22:52:39 PDT
From: ungar@self.stanford.edu (David Ungar)
Message-Id: <8909300552.AA00438@self>
To: boyland@sequoia.Berkeley.EDU, cps@vax.darpa.mil
Subject: Re:  looking for references about prototyping

Another language designed for prototyping is SELF (Ungar & Smith--OOPSLA '87).
We will be giving a talk on the implementation at OOPSLA next week.
It's a successor to Smalltalk, except it has no classes and
merges state and behavior.
It also runs twice as fast.

David Ungar

∂30-Sep-89  1041	harnad@clarity.Princeton.EDU 	Re: Skywriting 
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 30 Sep 89  10:41:42 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA29749; Thu, 28 Sep 89 16:40:56 EDT
Received: by psycho.Princeton.EDU (4.0/1.85)
	id AA06859; Thu, 28 Sep 89 16:07:37 EDT
Date: Thu, 28 Sep 89 16:07:37 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8909282007.AA06859@psycho.Princeton.EDU>
To: srh@flash.bellcore.com
Subject: Re: Skywriting

To: BBS Associates

This is the last of the postings on "Skywriting," a spin-off
from the APS/BBS Poll. After this the BBS-list returns to
business as usual: the posting of abstracts of forthcoming
target articles. (Your reactions are still welcome, though.)

Stevan Harnad
-----
Jonathan Baron (baron@cattell.psych.upenn.edu), Psychology
Department, University of Pennsylvania, wrote the following
in response to the remarks about "Skywriting" that followed
the recent BBS/APS poll:

	"I think that e-mail is great for research collaboration, for
	interdepartmental discussion in lieu of meetings, for sending
	in journal reviews, for general discussions within small groups
	of people, for local news, and even for sending out abstracts
	requesting reviewers, as you do, but I'm skeptical about other
	uses that involve, for example, free-for-all discussions of some
	It gets to be too much to read. As I think you've learned over
	the years, editors have a valuable function, sifting the wheat
	from the chaff and improving the quality of the former."

In response to my request for permission to identify him by name,
Dr. Baron responded:

	"Sure, you can identify me by name and quote me. But I would
	also appreciate it if you would mention that I have CHANGED my
	views as a result of reading your reply. They were more in the
	spirit of questions. (I sometimes ask questions by making
	assertions.)

[The reply in question follows:]

Thanks for sharing your reflections on limits of email discussion. I
have a vision, however, of what email net discussion COULD be -- if
conducted at a sufficiently sophisticated level, with certain
constraints, by the best minds in the fields in question -- and I hope
to be able to implement the vision eventually. In my opinion (and
experience), the speed, flexibility and scale of multiple email network
communication (a medium I've dubbed "Skywriting") is potentially so
powerful and congenial to a creative thinker's scale and tempo of
communication and interaction that it may represent just as
revolutionary a leap in human communicative (and creative) potential as
its three momentous predecessors (the advent of language, writing and
printing).

Don't form too firm a pessimistic conclusion from what you see on the Net
now: For mere historical and technological reasons it's currently
dominated by computer scientists and students (rather as if not just
the form but the contents of the printed page had been dominated for a
century by Guttenberg and a legion of linotype operators) and shunned
by a more mature generation in all other disciplines, whose majorities
are still computer-phobic. Try instead to imagine what it would be like
if the strongest thinkers in all the disciplines were on the Net now,
and used it concertedly.

There's no great problem with sifting the wheat from the chaff:
Wherever necessary a discussion group could restrict access to
read-only for all but the qualified; and one can always protect oneself
from unwanted discussion by one keystroke -- or even pre-filters in
mail one's software. Wasted forests of paper and junk mail choking your
mail-box are not at issue. The facilities for scanning, filtering and
discarding unwanted Skywriting are even more powerful than the
facilities of Skywriting itself!

But what you have to sample for yourself are the remarkable rewards of
Skywriting on topics that are currently of pressing scientific interest
to you. I can't say I've done this with the best minds in the field yet
(not enough of them are on the Net, to date). But then it's perhaps all
the more remarkable that even with the Net's current demography I can
report that lately my own interactions via Skywriting have advanced my
ideas more than conventional reading/writing, speaking/listening
combined. Imagine getting a provocative idea or experimental finding,
posting it at 9 am, and already receiving feedback from the relevant
parts of the world scholarly community within minutes: "Here's a
logical problem; here's some contrary evidence; here's some supporting
data; here are some surprising ramifications; so-and-so has already
reported this... etc." Sure, there's a lot of potential information from
such a source, but what earnest thinker would rather turn AWAY from
it?

"Editing" is really a software matter at this "pilot" stage of inquiry,
conducted in the Skywriting medium. Selecting and archiving what turns
out to be lapidary in this evolution of ideas is another matter, but
we're talking about improving the flow right now. Are you really so
satisfied with the current scale, scope and rate of communication of
ideas in the conventional media of scholarly communication? Might one
not have hesitated at the advent of printing too ("There's already too
much to read...")!

The possibilities of Skywriting are an empirical matter, but I'm going
to do my level best to make sure they are properly tested, rather than
pre-emptively discarded or passed over because of the skewed initial
conditions.
-----

William P. Gardner <wpg@mendel.acc.virginia.edu>, Psychology
Department, University of Virginia, (also identified with permission)
responded:

> I worry about your view that `` "Editing" is really a software matter
> at this "pilot" stage of inquiry''. Maybe I don't understand you --
> what will change when we are beyond the pilot stage?

You misunderstood me. It was the pilot stage of the scholarly thought
process that I was referring to. I think the scope and pace of
skywriting is exceptionally well suited to this, and I don't think it
needs any editing at all if high-quality thinkers are involved. The
refereed journals are for a later, more lapidary stage of the evolution of
ideas -- although I think even some portions of the early
brain-storming will turn out to be worth archiving in some form, if
only as a periscope on the evolution of ideas.

> The problem with electronic discussions is that there are few
> incentives for participants to edit themselves. Even great minds write
> poorly (sometimes) on the net. There are notorious lapses in the
> courtesy required by a discussion which is fiercely critical about
> substantive matters. As a result, the marginal return from time on the
> net seems lower than time spent reading material in traditional media.

I agree completely about all of this, but I believe it's because the
medium is so new that people simple haven't realized how radically
different its vast potential really is from what they're used to or
expect. Their imaginations have failed them. It's rather like those old
family movies, in which everyone is posing statically in front of the
camera as if for a still photograph.

If the best thinkers come out in force in each discipline and set a
breath-taking example of how serious ideas can evolve in the sky, then
the Net's current contents will be put into context once and for all
for the earth-bound trivial pursuits they were, in comparison with the
heights skywriting is really capable of reaching.

And don't discount the demographic factor I mentioned either: The Net
is still dominated by computerphiles and students; most of academe is
still computerphobic, and I'm afraid that the best minds are turning
out to be the most conservative in this respect -- reinforced in their
negative view of skywriting by the scouts who (correctly) report to
them the state of affairs you just described. I'm convinced that this
is just a matter of untoward initial conditions though, and I'd like to
do something to shake things up and hasten the skyways toward their
true destiny.

> Electronic discussions fall between communication in journals and
> seminars in the fluidity and intimacy of interchange. But in both
> journals and seminars, there are strong feedback mechanisms that
> discipline the participants. The feedback in journals drives manuscript
> selection and is, obviously, powerful and effective. In a seminar there
> are verbal and non-verbal responses from one's peers. These responses
> have less professional significance than publication decisions, but
> they are immediate and exploit powerful mechanisms in our psychology.
> These mechanisms do not operate effectively on the net, if there are no
> referees or moderators. Communication on the net resembles seminar
> discussion to a small degree, but the personal feedback is delayed and,
> I infer, heavily discounted by the recipient.

All that's needed is that those who normally provide the feedback,
those in whose view we would ordinarily prefer not to make fools of
ourselves -- that college of editors, reviewers, authors, promotion
committees, and other peer evaluators -- should make their presence
prominently felt in the skies. Even in the locker room there are signs
of spontaneous self-restraint when the Dean is in the next row.
Whether by the carrot -- setting high quality examples, and their
extrinsic (laudatory) and intrinsic (intellectual) rewards -- or the
stick (airborne and land-based censure for discourtesy or sloppy
thinking), I'm convinced that skywriting can become as serious and
self-disciplined as the other media of scholarly communication. Right
now the ethos is still too much like a global graffiti board, but
this, as I suggest, is just because of demography and technological
initial conditions.

> Good electronic networks need human editors or moderators. I haven't
> tried being a moderator but it's obviously a lot of work. If we want
> good moderation we will need to reward it, tangibly or intangibly. Your
> idea of a seminar of leading figures, read-only for the rest, is
> excellent. I would happily join the audience. Even your idea, however,
> requires a gatekeeper to assign write privileges and, presumably,
> filter questions from the balcony.

The archival material will need editing, but I don't think skywriting
itself needs much tampering once the altitude is set. The groups at
the highest altitudes could start with a read-write (RW) core population
consisting of the acknowledged leading thinkers in a field, and a
read-only (RO) periphery, consisting of everyone else who is
interested (or qualified and interested, if the discipline wishes to
set further restrictions). The RW's have unrestricted skywriting
privileges (although the group may have a rule that privileges can be
revoked too, if the group, or an appointed subset of it, judges that
they've been abused). ROs could submit candidate postings to RWs, or a
designated subset of RWs, and under their sponsorship they could have
one-time, provisional, or full posting privileges (thereby becoming
RWs).

The system could be implemented at the level of a small
sub-subspecialty, all the way up to an entire academic discipline or
mutlidisciplinary consortium. But the key is that individual RWs should
have full, unmoderated, 24-hour access to the worldwide skyways. That, I
suggest, will provide a medium of communication whose tempo and
territory best match the cognitive and communicative potential of the
human cerebral cortex -- optimally situated between the snail-pace of
letter-writing and journal publication, with their hopelelessly slow
turnaround times for intellectual feedback, and the impulsive and
undisciplined medium of spontaneous blow-away speech (and its
embarassing residue in unedited symposium proceedings).

Stevan Harnad
-----

Finally, Pat Hayes <hayes@parc.xerox.com> (also identified with permission)
added the following:

>  One quick remark. I am always puzzled when people compain about getting
>  too much email. Its VERY easy to delete or just flick past
>  email messages, very little time is wasted, and one can imagine fairly
>  simple software which would help in the task (eg, automatically
>  putting messages from particularly rebarbative people in a backup
>  mailbox for reading at a calmer time, or automatically deleting news
>  bulletins from some organisation or system one no longer has interest
>  in but which refuses to delete one from its mailing list). I suspect
>  that the objection isnt completely rational, and comes from a feeling
>  that one's computer is part of one's personal space, so that all this
>  'junk' is like finding trash in one's living room, rather than lots of
>  letters in one's mailbox.

-----

∂30-Sep-89  2115	harnad@clarity.Princeton.EDU 	Searle and the Symbol Grunding Problem  
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 30 Sep 89  21:15:34 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/2.22)
	id AA14618; Sat, 30 Sep 89 23:09:06 EDT
Received: by psycho.Princeton.EDU (4.0/1.85)
	id AA09527; Sat, 30 Sep 89 23:05:16 EDT
Date: Sat, 30 Sep 89 23:05:16 EDT
From: harnad@clarity.Princeton.EDU (Stevan Harnad)
Message-Id: <8910010305.AA09527@psycho.Princeton.EDU>
To: harnad@elbereth.rutgers.edu
Subject: Searle and the Symbol Grunding Problem

ON SEARLE'S "INTRINSIC INTENTIONALITY" AND THE SYMBOL GROUNDING PROBLEM

From:  Stevan Harnad

Michael G Dyer <dyer@CS.UCLA.EDU> wrote:
  
>  1. Searle's "grounding"
>  
>  Having read Searle and having recently heard a talk he gave at UCLA, I
>  do not think Searle thinks symbolic systems lack intentionality because
>  they are not grounded in perception. His Chinese box arguments are
>  confusions of levels (of where the intelligence resides) and the
>  inability to accept that intelligence may arise from the component
>  interactions, independent of the types of components used.

You are apparently satisfied with the "Systems Reply." I definitely am
not. Searle has repeatedly said that it's not that he doesn't believe
intelligence arises from "component interactions," because he does
believe that, for example, it arises from the component interactions in
the brain. Nor is Searle committed to any particular kind of component
type. He just holds (correctly, in my view) that his argument has
shown that certain types of system (pure symbol systems, as it turns out,
in my analysis) are the wrong types of system to exhibit intrinsic
intentionality; their intentionality is all parasitic on our
interpretations.

[In my own crtitique of Searle -- Harnad, S. (1989) "Minds, Machines and
Searle" Journal of Experimental and Theoretical Artificial Intelligence
1: 5 - 25 -- I try to sort out systematically what Searle is right and
wrong about and why, and it all boils down to the Symbol Grounding Problem.]

>  In his Chinese box argument he forces the human to be a piece in an
>  intelligent system and then asks that human to introspect, or he asks
>  the human to act as an interpreter of an intelligence system and then
>  again introspect. Instead, he should have been asking the entire "box"
>  to introspect. Only then is he going to get an interesting answer.

This misses the essence of Searle's Argument: The person is not just
being a piece of a system, HE'S DOING EVERYTHING IN THE SYSTEM.
Whatever function you want to point to that a symbol-manipulator is
performing in order to pass the Turing Test, Searle can simply do it
himself. Introspection is the right test for whether the mental state
that is being attributed to the system he is simulating is really there
under conditions where he is performing every last function the system
is performing. That's why Searle says he doesn't see much left to hang
a mind on in the walls and ceiling of the Chinese Room and the chalk on
the blackboard. You're tilting at windmills if you want to insist that
THAT system has a mind. It's not that Searle is claiming NO system can
have a mind, even one of which he is himself a component. He's just
claiming that THAT system -- the pure symbol-cruncher under discussion
in the Chinese Room Argument -- has no mind.

And he's simply right about that, according to me. But, as I show in my
paper, it turns out that a very simple variant is immune to Searle's
objection: One in which the candidate system's functions -- the ones
all of which Searle is committed to performing himself, if his argument
is to go through -- include certain nonsymbolic functions, in
particular, sensory transduction. If, for example, the physical system
that implements the mental state of seeing something -- that state we
all know whether or not we're in just as directly as we all know
whether or not we're understanding Chinese -- includes transducer
function, then Searle cannot simulate it Chinese-Room style. For either
he will have to get only the transducer's output (in which case it's no
wonder he's not seeing, since he is not performing all the functions
the system he is simulating is supposed to perform) or he will have to
BE the transducer, in which case he will be seeing after all.

The reason Searle's argument works with understanding and fails with
seeing is that the candidate system for understanding was a purely
symbolic, whereas the candidate system for sensing was not. It was this
immunity of a seemingly trivial function (transduction) to Searle's
Argument that confirmed my hunch that sensory grounding may be critical
for modeling the mind. (And of course I believe that the place to
ground understanding is in sensory function.)

>  To my knowledge, Searle has never said the following but I will guess
>  that if he were asked whether or not something has to be "alive" to
>  have intentionality, I suspect that Searle would say that it does
>  have to be (I also suspect that Searle would vehemently deny that any
>  future system now being built under the term "artificial life" will
>  ever really be "alive" in any sense <see Langton (ed.) (1989)
>  Artificial Life>)

And I think he would be agnostic about whether or not something had to
be "alive" (by which I assume you mean a natural biological creature
that is not dead) to have a mind. For example, I doubt Searle would
have any problem with artificial neurons, as long as they had the same
causal powers as natural neurons. But Searle's extracurricular beliefs
are irrelevant. Insofar as they are based on the Chinese Room Argument,
he has given strong reasons for doubting that mere symbol-crunchers
("Strong AI") can have minds. The rest is moot.

>  I think the "grounding" Searle wants really has to do with the units of
>  a system being 'grounded' in living tissues (i.e. real neurons, all the
>  way down to quarks and other beasties postulated of reality).
>  In contrast, most AI researchers think that information processing in
>  humans is only loosely coupled to what neural tissues are doing (i.e.
>  the processes keeping the cell alive are separable from those being
>  used to process information of a cognitive nature).

I repeat: although Searle hasn't said so explicitly, I will: The only
kind of artificial system that is vulnerable to his Chinese Room
Argument is a pure symbol manipulating system. When it comes to
nonsymbolic functions, all bets are off. Searle mentioned in passing
that a system with a mind must have all the causal powers of a real
brain, but I'm sure he would agree that many of the causal powers of
the brain may be irrelevant to its having a mind (as Stephen Hawking's
tragic case shows so dramatically), and that all he can really insist
on are those causal powers of the brain that are necessary and
sufficient for having a mind (which begins to sound a little
tautological), and that any system, natural or artificial, that had
those, would have a mind. Apart from that, Searle simply points out,
correctly, that the kind of system he can simulate in the Chinese room
(a pure symbol system) clearly does not have a mind, whereas the brain
clearly does.

So the moral is simply that in the attempt isolate that subset of the
brain's function (or any other possible mind-supporting system's
function) that is necessary and sufficient for having a mind, a pure
symbol crunching module turns out not to the right subset!

>  I've never heard Searle talk about the importance of grounding symbols
>  in perception. If I'm wrong, please point me to his relevant paper(s).

Neither have I. But *I've* talked (and written) about it. And whereas I
first thought I was disagreeing with Searle, he assured me that we were
largely in agreement. So there you have it.

>  If we draw an analogy (dangerous, but permit me) to architecture, my
>  guess is that most AI types would say that a building is e.g. Gothic
>  as long as it's in the Gothic style, even if each brick is made of
>  plastic, while Searle would demand that the actual bricks be made of
>  the right clay and maybe even in the right century.
>  Clearly both the plastic brick and clay brick buildings have different
>  behaviors under stress, but at another level they are both Gothic. It
>  depends on whether you think intelligence has to be 'grounded' in the
>  clay or if it is the way the elements are put together that counts.

The problem is that "Gothic Style," if it's anything at all, is an
objective, observable property. (So is behavior under stress.) In
principle we can all agree that any structure that has certain
geometric properties, irrespective of any other properties, is Gothic.
Unfortunately, this is not true of understanding (or seeing, or any of
the other subjective properties of having a mind). For even if a system
has all the objective, observable properties of a system that has a
mind, it may not have a mind. And that can be true as a simple matter
of fact, not just a metaphysical possibility, as Searle's Chinese Room
Argument shows.

But here I actually part ways with Searle. Searle rejects all versions
of the Turing Test, very much for the above reasons, whereas I reject
only the traditional symbolic version of the Turing Test (symbols in,
symbols out, and presumably just symbol-crunching in between), because
of the Symbol Grounding Problem. The robotic version -- what I've
called the Total Turing Test (TTT) -- is good enough for me, especially
since that's all I can ever have to go on even with my fellow human
beings. Appearances could be deceiving there too (as the "Other Minds
Problem" reminds us), but as far as I'm concerned, Total Turing power
is power enough to convince me.

Besides, no one knows what the relevant causal powers of the brain are
anyway, so that certainly can't be the way to settle the matter. The
brain's equivalent of your clay's "behavior under stress" may or may
not be one of the properties any system must have in order to have a
mind. To put it another way, it's not clear how much of the TTT will
have to include not only the body's behavior, but the brain's. Both,
after all, are objective, observable properties. The more you capture,
the better your grounds for confidence. I just think (total) bodily
performance capacity is probably a sufficient constraint.

>  The issue is not whether or not future AI systems will have HUMAN
>  intentionality, but whether they will have Searle's (elusive)
>  intentionality that's good enough, say, to quarantee them human
>  rights.

Searle's "elusive" intentionality is the only kind there is. It's the
stuff the systems that really see or feel or understand [English or
Chinese] have (and know at first hand that they have, because they know
exactly what it feels like to see or feel or understand [English or
Chinese]) and that systems that only act exactly as if they see or feel
or understand [English or Chinese] don't have (but don't know they
don't have, because there's nobody home). It's just that I don't happen
to believe you can get a system to ACT (TTT) exactly as if it sees...
etc., without its really having a mind.

Pure symbol-crunchers, on the other hand, cannot by their very nature
have ALL the requisite powers, because some of the powers are
nonsymbolic  -- and, according to my theory, the symbolic ones must be
(nonmodularly) grounded in the nonsymbolic ones; so an autonomous
symbol crunching mudule will have no mental powers at all.

>  2. Harnad "Grounding"
>  
>  But Searle's 'grounding' is different (I assume) from the way you use
>  the term "grounding". You have pointed out (correctly, I believe) that
>  the relationship of symbols to perceptions is a more important and
>  fundamental problem than people may have realized.

And I tried to show above exactly what the connection between my
grounding problem and Searle's intrinsic intentionality problem is.

>  AI researchers have argued that one can go a long way toward
>  intelligence without worrying about grounding (e.g. we imagine that we
>  are working with machines that are like Helen Keller). The problem is
>  that even Helen Keller had the (incredibly complex) sense of touch and
>  movement through the environment.

Exactly. I know of no one -- certainly not Helen Keller or Stephen
Hawking -- who is even remotely like an autonomous symbolic module.
There is no such nonhuman species either. The symbol grounding problem
may be the reason why. And of course the reason AI could go as far as
it did with its mind-like feats of symbol crunching was that the
symbol-crunching is always parasitic on our human interpretations: From
a TTT standpoint, AI's feats are all mere toys.

>  As we try to figure out where symbols come from, we will discover that
>  the grounding problem is major and will effect the nature of the
>  symbols we build and the way we manipulate them.

Especially if we have TTT ambitions.

>  3. Connectionism and Systematicity
>  
>  Distributed representations aren't all the way to systematicity yet,
>  but there are a number of researchers that have been building methods
>  for dynamically forming distributed representations that act a lot like
>  symbols, yet still have a "microsemantics". For instance, see
>  Miikkulainen and Dyer in Intern Joint Conf on Neural Networks (IJCNN),
>  Wash D. C. 1989. and in Touretzky, Hinton, Sejnowski (eds.) Proc of
>  Connectionist Summerschool See also Pollack in Cognitive Science
>  conference 1988 & 89, See Dolan' s UCLA dissertaion on Tensor
>  manipulation networks. See also Lee, Flowers, Dyer in 1989 Cogn Sci
>  conf., See Sumida and Dyer in IJCAI-89.
>  
>  Of course, once we get started moving these distributed representations
>  around like symbols, then Fodor et al. can use the argument that it's
>  "just" an implementation. I think we do need symbols and, unlike
>  Chuchland and Sejnowski, I don't expect high-level cognition without
>  them. But automatically forming symbol representations through the
>  symbolic and perceptual tasks being demanded of them will give us more
>  robust intelligent systems and explain currently unexplained
>  phenomena.

I too think that symbols enter into cognition at some level. I just
think they have to be grounded bottom-up in nonsymbolic function. And
the difference between a pure symbol crunching system and a nonsymbolic
system that can climb to a symbolic level by learning through sensorimotor
interactions in the world is just the difference between theft and
honest toil. I also think that the requisite nonsymbolic function will
include a lot of transduction and pure analog processing (not just
connectionist networks).

And even when a symbolic level is "trained up" in such a hybrid system
we will still not have a pure symbol system, because pure symbol
manipulation is based only on the arbitrary shape of the symbol tokens,
whereas in this "dedicated" symbol system the symbol tokens would also
be grounded (via feature-detecting neural nets) in the sensory
projections of the objects to which they refer. I predict that
there will be no isolable, autonomous symbolic module in a grounded
hybrid system with TTT power.

>  One problem with symbolic implementations is that they use something
>  like ASCII -- i.e.  arbitrary, static and predetermined -- forcing us 
>  into the use of pointers where maybe they're not needed. What we want is 
>  something that's formed from perception (plus lots of innate wiring 
>  that extracts features from those perceptions -- that wiring we get 
>  "for free" from evolution) and modified by interacting with other 
>  "distributed symbols representations" (DSRs). Instead of ASCII for CAT 
>  we get a DSR that reconstructus images of cats moving, plus 
>  reconstructing abstract information about cats when we need that type
>  of information. Take a look at Miikkkulainen & Dyer in IJCNN-89 for 
>  how this can be done. (Notice that although the perceptual part in 
>  this system is incredibly primitive it's actually being used to help 
>  keep bindings straight when doing script role binding!) Look also at 
>  Nenov and Dyer in IEEE ICNN 1987 San Diego conference.

It's no coincidence that symbol manipulation uses ASCII, because in a
formal symbol system the shapes of the symbol tokens must be arbitrary:
they cannot resemble or be in any other causal way related to the
objects or states of affairs they stand for. If the symbol tokens are
constrained in any way, e.g., by either resembling or being causally
connected to the objects they stand for, it's another ball game. (I
would add only that I still believe a significant amount of
feature-learning is done during an organism's lifetime, rather than its
evolutionary history -- especially in the case of human beings.)


>  4. Fodor's innateness
>  
>  I doubt if Fodor remembers, but he visited Yale University years ago
>  (when I was a grad student there) and he gave a talk on innateness. to
>  Schank's group. Well, everyone started asking about what concepts were
>  innate. As I recall, we got him to admit that "telephone" was innate,
>  but "Chicago" didn't seem to be innate. It was really quite amazing. It
>  seemed to be the case that he postulated innateness in any case in
>  which he couldn't figure out an alternative.

Since Fodor believes that the intersection of the detectable features of
the members of (many? most? all?) categories is empty, he probably
believes in something worse then mere innateness (for that allows for
feature-detectors shaped by evolutionary "learning"). Such Cartesian or
Platonic "Spandrels" would have to be the residue of what I've dubbed the
"Big Bang Theory of the Origin of Knowledge." Too big a bullet for me
to bite, if my thinking were to lead me inexorably in that
direction...

Stevan Harnad

[Michael Dyer has responded to this; his response will be posted
in the next mailing.]