perm filename MSG.MSG[1,JMC]1 blob
sn#864659 filedate 1988-12-02 generic text, type C, neo UTF8
COMMENT ⊗ VALID 00399 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00049 00002 ∂01-Oct-88 1355 GOODMAN%uamis@rvax.ccit.arizona.edu reviews of draft
C00054 00003 ∂02-Oct-88 0016 RFC Prancing Pony Bill
C00056 00004 ∂03-Oct-88 0908 pehoushe@Gang-of-Four.Stanford.EDU Q(16) and Q(15) bet
C00062 00005 ∂03-Oct-88 1000 JMC
C00063 00006 ∂03-Oct-88 1001 MPS phone call
C00064 00007 ∂03-Oct-88 1005 lrosenbe@note.nsf.gov Re: coordination theory
C00066 00008 ∂03-Oct-88 1309 JK
C00069 00009 ∂03-Oct-88 1601 @Score.Stanford.EDU:ellis@src.dec.com ACLU
C00071 00010 ∂04-Oct-88 0908 GOODMAN%uamis@rvax.ccit.arizona.edu suggestions for concluding chapter
C00084 00011 ∂04-Oct-88 1045 MEERSMAN%HTIKUB5.BITNET@forsythe.stanford.edu Your text for the China proceedings
C00086 00012 ∂04-Oct-88 1201 gurevich@polya.Stanford.EDU Hello
C00088 00013 ∂04-Oct-88 1310 weening@Gang-of-Four.Stanford.EDU Computer Constellations
C00113 00014 ∂04-Oct-88 1533 MPS Franklin Speller
C00114 00015 ∂04-Oct-88 1811 lincoln@polya.Stanford.EDU Computron usage
C00116 00016 ∂04-Oct-88 2041 ARK Joe Weening
C00117 00017 ∂05-Oct-88 0924 ARK Paying for the Alliant
C00119 00018 ∂05-Oct-88 0947 JK MAD
C00120 00019 ∂05-Oct-88 1356 JK
C00121 00020 ∂05-Oct-88 1530 hoffman@csli.Stanford.EDU re: a follow up appointment?
C00123 00021 ∂05-Oct-88 1544 hoffman@csli.Stanford.EDU re: a follow up appointment?
C00125 00022 ∂05-Oct-88 1551 hoffman@csli.Stanford.EDU re: a follow up appointment?
C00127 00023 ∂05-Oct-88 1639 ginsberg@polya.Stanford.EDU formfeed resumes on 10/6
C00129 00024 ∂06-Oct-88 0007 helen@psych.Stanford.EDU Re: lunch
C00130 00025 ∂06-Oct-88 0751 beeson@ucscd.UCSC.EDU triangles
C00133 00026 ∂06-Oct-88 0959 MPS
C00134 00027 ∂06-Oct-88 1224 MPS Phone call
C00135 00028 ∂06-Oct-88 1450 ginsberg@polya.Stanford.EDU Formfeed to remain on Thursday
C00137 00029 ∂06-Oct-88 1515 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
C00141 00030 ∂06-Oct-88 1612 CLT noise reduction project
C00142 00031 ∂06-Oct-88 1639 jwalton@vax.darpa.mil 1988 Principal Investigators' Conference
C00144 00032 ∂06-Oct-88 1742 jwalton@vax.darpa.mil 1988 Principal Investigators' Conference
C00146 00033 ∂07-Oct-88 0730 @Score.Stanford.EDU:RINDFLEISCH@SUMEX-AIM.Stanford.EDU Action Items from Our Meeting
C00151 00034 ∂07-Oct-88 0911 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSD-CF Rates for 1988-89
C00158 00035 ∂07-Oct-88 0929 golub@na-net.stanford.edu FAX
C00159 00036 ∂07-Oct-88 0951 @Score.Stanford.EDU:ARK@SAIL.Stanford.EDU Re: CSD-CF Rates for 1988-89
C00163 00037 ∂07-Oct-88 1152 VAL exchange with Moscow
C00165 00038 ∂07-Oct-88 1237 VAL next week
C00166 00039 ∂07-Oct-88 1246 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSD-CF Rates for 1988-89
C00168 00040 ∂07-Oct-88 1302 @Score.Stanford.EDU:ball@polya.Stanford.EDU [ball@polya.Stanford.EDU: CSD-CF Rates for 1988-89]
C00176 00041 ∂07-Oct-88 1324 MPS Kyoto
C00177 00042 ∂07-Oct-88 1654 VAL Nonmonotonic seminar - no meeting next week
C00178 00043 ∂07-Oct-88 1724 nilsson@Tenaya.stanford.edu Berlin in April?
C00188 00044 ∂07-Oct-88 2242 @RELAY.CS.NET:masahiko@nuesun.ntt.jp Re: coming to Sendai
C00193 00045 ∂08-Oct-88 1206 ARK re: CSD-CF Rates for 1988-89
C00196 00046 ∂08-Oct-88 1707 ARK re: CSD-CF Rates for 1988-89
C00197 00047 ∂09-Oct-88 1420 daniel@mojave.Stanford.EDU disk use charges
C00200 00048 ∂09-Oct-88 1540 ARK Re: disk use charges
C00201 00049 ∂10-Oct-88 0530 Rich.Thomason@b.gp.cs.cmu.edu JPL Paper
C00204 00050 ∂10-Oct-88 0900 JMC
C00205 00051 ∂10-Oct-88 0913 WIEDERHOLD@SUMEX-AIM.Stanford.EDU Re: disk use charges
C00207 00052 ∂10-Oct-88 1109 hoffman@csli.Stanford.EDU as you're not here and your door is closed
C00209 00053 ∂10-Oct-88 1123 bhayes@polya.Stanford.EDU Psycho Pservey
C00211 00054 ∂10-Oct-88 1146 Rich.Thomason@cad.cs.cmu.edu re: JPL Paper
C00212 00055 ∂11-Oct-88 0133 @Gang-of-Four.Stanford.EDU,@IGNORANT.Stanford.EDU:rivin@GANG-OF-FOUR.Stanford.EDU meeting
C00214 00056 ∂11-Oct-88 0826 tom@polya.Stanford.EDU toner cartridges for Imagen
C00216 00057 ∂11-Oct-88 1340 BALDWIN@Score.Stanford.EDU toner cartridges
C00217 00058 ∂11-Oct-88 1356 Mailer Re: disk use charges
C00219 00059 ∂11-Oct-88 1401 BALDWIN@Score.Stanford.EDU re: toner cartridges
C00220 00060 ∂11-Oct-88 1501 RC.STA@Forsythe.Stanford.EDU
C00222 00061 ∂11-Oct-88 1540 barwise@russell.Stanford.EDU Reply
C00224 00062 ∂11-Oct-88 1546 barwise@russell.Stanford.EDU Re: Reply
C00225 00063 ∂11-Oct-88 1600 minker@jacksun.cs.umd.edu RITA G. MINKER
C00234 00064 ∂11-Oct-88 1609 MPS classes
C00235 00065 ∂11-Oct-88 2029 SHOHAM@Score.Stanford.EDU Re: your computer
C00236 00066 ∂12-Oct-88 1006 KARP@SUMEX-AIM.Stanford.EDU re: JMC's campaign analysis
C00237 00067 ∂12-Oct-88 1011 KARP@SUMEX-AIM.Stanford.EDU bias
C00238 00068 ∂12-Oct-88 1539 @Score.Stanford.EDU:WIEDERHOLD@SUMEX-AIM.Stanford.EDU Re: disk use charges
C00240 00069 ∂12-Oct-88 2236 @cunyvm.cuny.edu:YANG@NORUNIT.BITNET Your travel expenses, IFIP China.
C00243 00070 ∂12-Oct-88 2350 Mailer Re: disk use charges
C00247 00071 ∂13-Oct-88 0841 C.COLE@MACBETH.STANFORD.EDU Bush & Advice: One Anecdote
C00251 00072 ∂13-Oct-88 0900 ullman@polya.Stanford.EDU RT consoles
C00253 00073 ∂13-Oct-88 1255 Mailer Re: disk use charges
C00257 00074 ∂13-Oct-88 1429 pehoushe@Gang-of-Four.Stanford.EDU Is Anybody Out There?
C00259 00075 ∂13-Oct-88 1920 RC.STA@Forsythe.Stanford.EDU
C00262 00076 ∂13-Oct-88 1921 roach@Gang-of-Four.Stanford.EDU Re: Is Anybody Out There?
C00264 00077 ∂13-Oct-88 2014 DAMON@Score.Stanford.EDU New Charge Program
C00266 00078 ∂13-Oct-88 2207 ilan@Gang-of-Four.Stanford.EDU Cockburn
C00269 00079 ∂13-Oct-88 2355 helen@psych.Stanford.EDU Noon Saturday
C00270 00080 ∂14-Oct-88 0025 @CUNYVM.CUNY.EDU:YANG@NORUNIT.BITNET re: Your travel expenses, IFIP China.
C00271 00081 ∂14-Oct-88 1125 Mailer Re: disk use charges
C00275 00082 ∂14-Oct-88 1405 S.SALUT@LEAR.STANFORD.EDU Re: October surprise
C00276 00083 ∂14-Oct-88 1413 S.SALUT@LEAR.STANFORD.EDU re: October surprise
C00277 00084 ∂14-Oct-88 1420 P.EPSTEIN@GSB-WHY.Stanford.EDU re: 3rd party candidates
C00279 00085 ∂14-Oct-88 2133 beeson@ucscd.UCSC.EDU triangles
C00281 00086 ∂14-Oct-88 2136 beeson@ucscd.UCSC.EDU the regular tetrahedron
C00282 00087 ∂15-Oct-88 1038 meyer@theory.lcs.mit.edu hello
C00283 00088 ∂16-Oct-88 1148 beeson@ucscd.UCSC.EDU re: triangles
C00285 00089 ∂16-Oct-88 1149 beeson@ucscd.UCSC.EDU re: the regular tetrahedron
C00286 00090 ∂16-Oct-88 1420 hoffman@csli.Stanford.EDU briefly
C00289 00091 ∂17-Oct-88 0759 meyer@THEORY.lcs.mit.edu hello
C00300 00092 ∂17-Oct-88 0925 meyer@theory.lcs.mit.edu further invites: mac25
C00302 00093 ∂17-Oct-88 0928 meyer@theory.lcs.mit.edu hello
C00304 00094 ∂17-Oct-88 1132 meyer@theory.lcs.mit.edu hello
C00306 00095 ∂17-Oct-88 1150 JSW Thesis
C00307 00096 ∂17-Oct-88 1237 P.EPSTEIN@GSB-WHY.Stanford.EDU re: 3rd party candidates
C00309 00097 ∂17-Oct-88 1356 MPS phone number
C00310 00098 ∂17-Oct-88 1442 VAL Commonsense and Nonmonotonic Reasoning Seminar
C00312 00099 ∂17-Oct-88 1649 VAL Etherington
C00319 00100 ∂17-Oct-88 1842 GLB oral exams
C00320 00101 ∂18-Oct-88 0105 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Kids, don't try this at home!
C00322 00102 ∂18-Oct-88 0327 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM this is closed form?
C00325 00103 ∂18-Oct-88 1033 @RELAY.CS.NET:GOTO@ntt-20.ntt.jp Your visit to Kyoto / My visit to Stanford
C00328 00104 ∂18-Oct-88 1438 VAL msg from Dmitry Lenkov
C00329 00105 ∂18-Oct-88 1511 bhayes@polya.Stanford.EDU Coverup
C00330 00106 ∂18-Oct-88 1554 mkatz@sesame.stanford.edu Re: October surprise
C00333 00107 ∂18-Oct-88 1602 BSCOTT@Score.Stanford.EDU Tenured Faculty Meeting
C00335 00108 ∂18-Oct-88 1736 W.WENTWORTH@MACBETH.STANFORD.EDU Gann initiative
C00340 00109 ∂18-Oct-88 2022 barwise@russell.Stanford.EDU Re: Rota article
C00342 00110 ∂19-Oct-88 0020 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM this is closed form?
C00345 00111 ∂19-Oct-88 0236 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM on the other hand
C00347 00112 ∂19-Oct-88 0751 davism@csd11.nyu.edu letter to Berkeley
C00349 00113 ∂19-Oct-88 0752 davism@csd11.nyu.edu ad
C00352 00114 ∂19-Oct-88 0753 davism@csd11.nyu.edu qualifications & interests
C00363 00115 ∂19-Oct-88 0754 davism@csd11.nyu.edu resume
C00391 00116 ∂19-Oct-88 0920 ARK CSD-CF Rate Idea
C00397 00117 ∂19-Oct-88 1039 VAL kr'89 panel
C00401 00118 ∂19-Oct-88 1113 meyer@theory.lcs.mit.edu hello
C00403 00119 ∂19-Oct-88 1121 ball@polya.Stanford.EDU CSD-CF Rate Idea
C00406 00120 ∂19-Oct-88 1149 ELLIOTT%SLACVM.BITNET@forsythe.stanford.edu Re: bicycling
C00411 00121 ∂19-Oct-88 1246 GLB
C00413 00122 ∂19-Oct-88 1322 CLT qlisp for okuno
C00414 00123 ∂19-Oct-88 1329 kar@polya.Stanford.EDU Applications AI comp.
C00416 00124 ∂19-Oct-88 1333 pimeet@vax.darpa.mil PI Meeting
C00418 00125 ∂19-Oct-88 1406 CLT qlisp for okuno
C00419 00126 ∂19-Oct-88 1413 ginsberg@polya.Stanford.EDU Formfeed to meet tomorrow -- don't forget!
C00420 00127 ∂19-Oct-88 1640 MPS
C00421 00128 ∂19-Oct-88 1642 MPS Dallas
C00422 00129 ∂19-Oct-88 1644 MPS book
C00423 00130 ∂19-Oct-88 1707 meyer@theory.LCS.MIT.EDU no reservation needed
C00426 00131 ∂19-Oct-88 2024 @Gang-of-Four.Stanford.EDU:arg@lucid.com Talk on Parallel Processing and the Butterfly Computer
C00430 00132 ∂20-Oct-88 0858 ELLIOTT%SLACVM.BITNET@forsythe.stanford.edu re: bicycling
C00431 00133 ∂20-Oct-88 1054 pimeet@vax.darpa.mil PI Meeting
C00434 00134 ∂20-Oct-88 1203 pehoushe@Gang-of-Four.Stanford.EDU How to ise NSTACK
C00442 00135 ∂20-Oct-88 1229 @Gang-of-Four.Stanford.EDU,@IGNORANT.Stanford.EDU:rivin@GANG-OF-FOUR.Stanford.EDU How to ise NSTACK
C00444 00136 ∂20-Oct-88 1245 ME Boise
C00445 00137 ∂20-Oct-88 1334 pehoushe@Gang-of-Four.Stanford.EDU How to use NSTACK
C00447 00138 ∂20-Oct-88 1455 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU:kapur@albanycs.albany.edu extension of submission deadline for RTA89
C00450 00139 ∂20-Oct-88 1618 hoffman@csli.Stanford.EDU re: some Symbolic System Forums Announcements
C00452 00140 ∂20-Oct-88 1638 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU:kapur@albanycs.albany.edu extension of submission deadline for RTA89
C00455 00141 ∂20-Oct-88 1650 rick@hanauma.STANFORD.EDU re: Amazon rain forest destruction
C00457 00142 ∂20-Oct-88 1859 JK
C00458 00143 ∂20-Oct-88 2100 JMC
C00459 00144 ∂21-Oct-88 0207 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM hyperbowly
C00465 00145 ∂21-Oct-88 0748 pimeet@vax.darpa.mil PI Meeting
C00467 00146 ∂21-Oct-88 0800 JMC
C00468 00147 ∂21-Oct-88 1019 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
C00470 00148 ∂21-Oct-88 1304 scherlis@vax.darpa.mil PI MEETING PROJECT SUMMARIES
C00475 00149 ∂21-Oct-88 1342 kar@polya.Stanford.EDU Applications AI comp
C00477 00150 ∂21-Oct-88 1434 nfields@vax.darpa.mil SUMMARY ENCLOSED
C00480 00151 ∂21-Oct-88 1441 nfields@vax.darpa.mil summary enclosed
C00485 00152 ∂21-Oct-88 1543 CLT dinner with Susie, etc.
C00486 00153 ∂21-Oct-88 1711 MPS vacation
C00487 00154 ∂22-Oct-88 1002 CLT Hazel
C00488 00155 ∂22-Oct-88 1008 RPG PI Meeting
C00491 00156 ∂22-Oct-88 1011 RPG PI Meeting
C00494 00157 ∂23-Oct-88 0842 barwise@russell.Stanford.EDU common knowledge
C00496 00158 ∂23-Oct-88 1559 beeson@ucscd.UCSC.EDU common sense in word problems
C00499 00159 ∂23-Oct-88 1804 RDZ@Score.Stanford.EDU Dinner?
C00500 00160 ∂24-Oct-88 0923 VAL Commonsense and Nonmonotonic Reasoning Seminar
C00502 00161 ∂24-Oct-88 0948 andy@cayuga.Stanford.EDU JFK's book? (was: Quayle)
C00504 00162 ∂24-Oct-88 1005 CLT painter
C00505 00163 ∂24-Oct-88 1022 CLT notes
C00506 00164 ∂24-Oct-88 1024 CLT umbrella
C00507 00165 ∂24-Oct-88 1031 CLT pullen
C00508 00166 ∂24-Oct-88 1031 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSD-CF Employee List
C00511 00167 ∂24-Oct-88 1452 RWF re: Quayle
C00512 00168 ∂24-Oct-88 1605 chandler@polya.Stanford.EDU I have something for you
C00514 00169 ∂24-Oct-88 1622 RLM@Score.Stanford.EDU Re: Quayle
C00515 00170 ∂25-Oct-88 0454 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM theology question
C00518 00171 ∂25-Oct-88 0955 STAGER@Score.Stanford.EDU Winter text orders
C00520 00172 ∂25-Oct-88 1633 betsy@russell.Stanford.EDU Office Space at CSLI
C00522 00173 ∂25-Oct-88 1741 GLB date of my exam
C00523 00174 ∂25-Oct-88 1829 CLT inamori
C00524 00175 ∂26-Oct-88 1440 chandler@polya.Stanford.EDU Something for you...
C00525 00176 ∂27-Oct-88 0652 harnad%confidence.Princeton.EDU@Princeton.EDU California talks
C00551 00177 ∂27-Oct-88 1047 littell@polya.Stanford.EDU Alex Gorbis
C00553 00178 ∂27-Oct-88 1057 goldberg@polya.Stanford.EDU Industrial Lecturers
C00555 00179 ∂27-Oct-88 1438 scherlis@vax.darpa.mil REVIEWS, PI MEETING, PLANS
C00576 00180 ∂27-Oct-88 1740 JK the 91 function
C00577 00181 ∂27-Oct-88 1852 CLT msg
C00578 00182 ∂27-Oct-88 2341 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM theology question
C00583 00183 ∂28-Oct-88 1128 CLT goto
C00584 00184 ∂28-Oct-88 1213 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
C00586 00185 ∂28-Oct-88 1634 CLT loan
C00587 00186 ∂28-Oct-88 1730 pimeet@vax.darpa.mil PI Meeting - Speech
C00593 00187 ∂29-Oct-88 0918 CLT today
C00594 00188 ∂29-Oct-88 1322 DEK
C00595 00189 ∂30-Oct-88 0359 JSW Electronic mail
C00599 00190 ∂30-Oct-88 1558 A.ERIC@GSB-HOW.Stanford.EDU Macintosh virus information
C00633 00191 ∂31-Oct-88 0807 Mailer re: those whales
C00635 00192 ∂31-Oct-88 0856 VAL Commonsense and Nonmonotonic Reasoning Seminar
C00638 00193 ∂31-Oct-88 1033 RPG
C00639 00194 ∂31-Oct-88 1209 RPG
C00640 00195 ∂01-Nov-88 1055 GC.TLX@Forsythe.Stanford.EDU MSG RCVD FOR J. MCCARTHY
C00645 00196 ∂01-Nov-88 1201 iris@cive.STANFORD.EDU getting ahold of Dr. John Sowa
C00647 00197 ∂01-Nov-88 1207 bill@gatech.edu Need information on Dany Guindi
C00649 00198 ∂01-Nov-88 1217 rivin@polya.Stanford.EDU preprints
C00651 00199 ∂01-Nov-88 1316 VAL Ablex
C00652 00200 ∂01-Nov-88 1327 BYRD@SUMEX-AIM.Stanford.EDU Re: anti-racism
C00654 00201 ∂01-Nov-88 1500 G.GSB@MACBETH.STANFORD.EDU Re: anti-racism
C00656 00202 ∂01-Nov-88 1510 BLUMENTHAL@A.ISI.EDU Final Touches
C00687 00203 ∂01-Nov-88 1809 G.GSB@MACBETH.STANFORD.EDU Re: anti-racism
C00689 00204 ∂01-Nov-88 1824 G.GSB@MACBETH.STANFORD.EDU re: anti-racism
C00693 00205 ∂01-Nov-88 1948 SHOHAM@Score.Stanford.EDU russians
C00695 00206 ∂01-Nov-88 2146 SHOHAM@Score.Stanford.EDU re: russians
C00696 00207 ∂01-Nov-88 2351 qphysics-owner@neat.ai.toronto.edu lethargy?
C00700 00208 ∂02-Nov-88 0633 BLUMENTHAL@A.ISI.EDU re: Final Touches
C00701 00209 ∂02-Nov-88 0850 SHOHAM@Score.Stanford.EDU re: russians
C00702 00210 ∂02-Nov-88 0959 MPS paper on soviet access....
C00703 00211 ∂02-Nov-88 1204 CLT calendar item
C00704 00212 ∂02-Nov-88 1738 Mailer re: anti-racism
C00705 00213 ∂02-Nov-88 1740 kar@polya.Stanford.EDU Reminder: Applications AI questions needed by November 7th
C00708 00214 ∂02-Nov-88 1751 RFC Prancing Pony Bill
C00710 00215 ∂03-Nov-88 0700 pehoushe@Gang-of-Four.Stanford.EDU [elloyd@NOTE.NSF.GOV: NSF support for algorithms and parallel computing systems]
C00717 00216 ∂03-Nov-88 0929 MPS Pigott Account
C00718 00217 ∂03-Nov-88 0957 SHOHAM@Score.Stanford.EDU CSLI evening seminars
C00720 00218 ∂03-Nov-88 1041 SHOHAM@Score.Stanford.EDU re: CSLI evening seminars
C00721 00219 ∂03-Nov-88 1056 scales@polya.Stanford.EDU research mentor
C00724 00220 ∂03-Nov-88 1058 VAL
C00725 00221 ∂03-Nov-88 1159 @b.NSF.GOV:mzemanko@note.nsf.gov Proposal review
C00728 00222 ∂03-Nov-88 1206 weening@Gang-of-Four.Stanford.EDU Information on the "virus"
C00737 00223 ∂03-Nov-88 1253 weening@Gang-of-Four.Stanford.EDU Virus update
C00739 00224 ∂03-Nov-88 1412 clark@sm.unisys.com
C00742 00225 ∂03-Nov-88 1720 VAL Nonmonotonic Reasoning Seminar: Reminder and correction
C00745 00226 ∂03-Nov-88 1832 Mailer re: Article on meat eating
C00753 00227 ∂03-Nov-88 1937 RLM@Score.Stanford.EDU Red meat fun with McDonald's
C00756 00228 ∂04-Nov-88 0718 @b.NSF.GOV:mzemanko@note.nsf.gov Re: Proposal review
C00773 00229 ∂04-Nov-88 1655 betsy@russell.Stanford.EDU CSLI Stanford Faculty Meeting
C00775 00230 ∂05-Nov-88 0910 CLT You haven't sent me the number of the
C00776 00231 ∂05-Nov-88 1012 CLT thanks
C00777 00232 ∂05-Nov-88 1117 singh@sierra.STANFORD.EDU re: Mud-slinging and JMC's heroes [was Re: more Coverup]
C00778 00233 ∂07-Nov-88 0429 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Bessel, yo is my worry now
C00780 00234 ∂07-Nov-88 0630 ghh@confidence.Princeton.EDU AI as CS and the scientific epistemology of the common sense world
C00782 00235 ∂07-Nov-88 0954 betsy@russell.Stanford.EDU re: CSLI Stanford Faculty Meeting
C00784 00236 ∂07-Nov-88 1000 eaf@sumex-aim.stanford.edu [davies@cascade.Stanford.EDU (Byron Davies) : Knowledge quantum ]
C00786 00237 ∂07-Nov-88 1146 TAJNAI@Score.Stanford.EDU Seeking sponsor for Visiting Scholar
C00789 00238 ∂07-Nov-88 1420 scherlis@vax.darpa.mil PI Meeting
C00791 00239 ∂07-Nov-88 1429 honavar@cs.wisc.edu Re: AI as CS and the scientific epistemology of the common sense world
C00793 00240 ∂08-Nov-88 1154 VAL Commonsense and Nonmonotonic Reasoning Seminar
C00795 00241 ∂08-Nov-88 1225 jwalton@vax.darpa.mil [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: Deferred: Connecti]
C00807 00242 ∂08-Nov-88 1507 pullen@vax.darpa.mil ISTO PI Meeting Agenda
C00815 00243 ∂08-Nov-88 1938 qphysics-owner@neat.ai.toronto.edu lethargy?
C00820 00244 ∂08-Nov-88 2106 qphysics-owner@neat.ai.toronto.edu Lethargy?
C00822 00245 ∂09-Nov-88 1204 VAL Special Seminar: Keith Clark's FGCS invited talk
C00824 00246 ∂09-Nov-88 1305 GOODMAN%uamis@rvax.ccit.arizona.edu finished
C00826 00247 ∂09-Nov-88 1333 jwalton@vax.darpa.mil [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: Deferred: Connecti]
C00838 00248 ∂09-Nov-88 1345 postmaster@guvax Returned Network Mail
C00856 00249 ∂09-Nov-88 1743 qphysics-owner@neat.ai.toronto.edu another TR announcement
C00862 00250 ∂10-Nov-88 0225 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Bessel, yo is my worry now
C00868 00251 ∂10-Nov-88 1000 STAGER@Score.Stanford.EDU Text orders???
C00870 00252 ∂10-Nov-88 1206 weening@Gang-of-Four.Stanford.EDU Network connection to USSR
C00871 00253 ∂10-Nov-88 1206 weening@Gang-of-Four.Stanford.EDU
C00875 00254 ∂10-Nov-88 1208 weening@Gang-of-Four.Stanford.EDU
C00882 00255 ∂10-Nov-88 1703 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
C00884 00256 ∂10-Nov-88 2352 harnad@confidence.Princeton.EDU BBS Call For Commentators: The Tag Assignment Problem
C00890 00257 ∂11-Nov-88 0908 weening@Gang-of-Four.Stanford.EDU
C00895 00258 ∂11-Nov-88 0923 pehoushe@Gang-of-Four.Stanford.EDU My Vacation Time
C00897 00259 ∂11-Nov-88 0952 CN.MCS@Forsythe.Stanford.EDU
C00900 00260 ∂11-Nov-88 1412 weening@Gang-of-Four.Stanford.EDU passwords
C00905 00261 ∂13-Nov-88 1109 GKMARH%IRISHMVS.BITNET@forsythe.stanford.edu post-docs in cs
C00909 00262 ∂14-Nov-88 1013 JK
C00910 00263 ∂14-Nov-88 1157 SLOAN@Score.Stanford.EDU [Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>: My Vacation Time]
C00913 00264 ∂14-Nov-88 1219 SLOAN@Score.Stanford.EDU
C00915 00265 ∂14-Nov-88 1316 weening@Gang-of-Four.Stanford.EDU
C00930 00266 ∂14-Nov-88 1446 CN.MCS@Forsythe.Stanford.EDU
C00932 00267 ∂14-Nov-88 1542 CN.MCS@Forsythe.Stanford.EDU
C00934 00268 ∂14-Nov-88 1635 weening@Gang-of-Four.Stanford.EDU
C00942 00269 ∂14-Nov-88 1934 GLB meeting
C00943 00270 ∂14-Nov-88 2152 DEK
C00946 00271 ∂15-Nov-88 0955 VAL Commonsense and Nonmonotonic Reasoning Seminar
C00948 00272 ∂15-Nov-88 1513 debra@russell.Stanford.EDU EVENING SEMINAR
C00950 00273 ∂15-Nov-88 1628 JANLEE%VTVM1.BITNET@forsythe.stanford.edu History of AI and Time Sharing
C00956 00274 ∂15-Nov-88 2226 harnad@confidence.Princeton.EDU BBS Call for Commentators: Are Species Intelligent / J. Schull
C00962 00275 ∂15-Nov-88 2251 harnad@confidence.Princeton.EDU Genetic Similarity Theory: BBS Call for Commentators
C00967 00276 ∂16-Nov-88 1320 weening@Gang-of-Four.Stanford.EDU
C01047 00277 ∂16-Nov-88 1330 ginsberg@polya.Stanford.EDU formfeed to meet tomorrow!
C01048 00278 ∂16-Nov-88 2207 VAL Special seminar on default reasoning
C01049 00279 ∂17-Nov-88 0140 JMC Expired plan
C01050 00280 ∂17-Nov-88 0937 ginsberg@polya.Stanford.EDU Formfeed to meet today!
C01051 00281 ∂17-Nov-88 1313 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSDCF Meeting Discussion Item...Cost Analysis
C01058 00282 ∂17-Nov-88 1532 MPS
C01059 00283 ∂17-Nov-88 1753 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
C01061 00284 ∂17-Nov-88 1810 MGardner.pa@Xerox.COM AIJ Board Members Mtg
C01074 00285 ∂17-Nov-88 2250 ramshaw@src.dec.com the history of the term "bag"
C01076 00286 ∂18-Nov-88 0805 SLOAN@Score.Stanford.EDU Re: reply to message
C01077 00287 ∂18-Nov-88 0813 CN.MCS@Forsythe.Stanford.EDU
C01082 00288 ∂18-Nov-88 1425 GLB oral exam of Gianluigi Bellin.
C01083 00289 ∂18-Nov-88 1436 peters@russell.Stanford.EDU CSLI's Industrial Affiliates Program
C01085 00290 ∂18-Nov-88 2101 @Gang-of-Four.Stanford.EDU:arg@lucid.com new new-qlisp
C01102 00291 ∂19-Nov-88 1525 ME SAIL
C01103 00292 ∂19-Nov-88 1947 weening@Gang-of-Four.Stanford.EDU Inquirer article
C01104 00293 ∂20-Nov-88 1701 hoffman@csli.Stanford.EDU about meeting
C01107 00294 ∂20-Nov-88 2043 ME NIC/DDN registration
C01108 00295 ∂20-Nov-88 2349 BXR Wow!
C01109 00296 ∂21-Nov-88 0124 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Bessel, yo is my worry now
C01121 00297 ∂21-Nov-88 1000 JMC
C01122 00298 ∂21-Nov-88 1035 CLT Knuth dinner party
C01123 00299 ∂21-Nov-88 1051 CN.MCS@Forsythe.Stanford.EDU
C01125 00300 ∂21-Nov-88 1157 PHY
C01126 00301 ∂21-Nov-88 1159 pehoushe@Gang-of-Four.Stanford.EDU Throw Problems
C01128 00302 ∂21-Nov-88 1204 CLT Vacation Time
C01129 00303 ∂21-Nov-88 1209 JANLEE%VTVM1.BITNET@forsythe.stanford.edu History of Time Sharing
C01132 00304 ∂21-Nov-88 1326 weening@Gang-of-Four.Stanford.EDU Meeting with Alliant
C01134 00305 ∂21-Nov-88 1542 hoffman@csli.Stanford.EDU sorry about this morning
C01136 00306 ∂21-Nov-88 1611 pullen@vax.darpa.mil [WASPRAY%UMNACVX.BITNET@CUNYVM.CUNY.EDU: final version of history circular]
C01147 00307 ∂21-Nov-88 1650 scherlis@vax.darpa.mil PI MEETING -- DISCUSSION SUMMARY
C01177 00308 ∂21-Nov-88 2231 weening@Gang-of-Four.Stanford.EDU Soviet access to USENET
C01179 00309 ∂22-Nov-88 0520 JANLEE%VTVM1.BITNET@forsythe.stanford.edu Quick Reply
C01181 00310 ∂22-Nov-88 0854 beeson@ucscd.UCSC.EDU tetrahedron
C01183 00311 ∂22-Nov-88 0943 CLT nsf proposal
C01185 00312 ∂22-Nov-88 1002 ghh@confidence.Princeton.EDU AI as CS and the scientific epistemology of the common sense world
C01187 00313 ∂22-Nov-88 1045 VAL re: nsf proposal
C01188 00314 ∂22-Nov-88 1254 VAL Journal of Automated Reasoning
C01189 00315 ∂22-Nov-88 1321 GLB
C01190 00316 ∂22-Nov-88 1349 jonas@polya.Stanford.EDU cs323
C01193 00317 ∂22-Nov-88 1351 MPS computer chess
C01194 00318 ∂22-Nov-88 1441 GLB
C01195 00319 ∂22-Nov-88 1455 drb@cscfac
C01200 00320 ∂22-Nov-88 1537 weening@Gang-of-Four.Stanford.EDU Qlisp meeting
C01202 00321 ∂22-Nov-88 1549 air@Gang-of-Four.Stanford.EDU abstract from India
C01204 00322 ∂22-Nov-88 1557 MPS PTO
C01205 00323 ∂22-Nov-88 1601 VAL re: NSF report
C01206 00324 ∂22-Nov-88 2008 GLB
C01209 00325 ∂22-Nov-88 2132 hoffman@csli.Stanford.EDU i have printed out
C01211 00326 ∂22-Nov-88 2227 crew@polya.Stanford.EDU new mailing list mtc@polya now exists
C01216 00327 ∂22-Nov-88 2251 harnad@Princeton.EDU BBBS Call for Neuroscience Nominations
C01219 00328 ∂23-Nov-88 0010 rdz@Gang-of-Four.Stanford.EDU What I'm up to
C01221 00329 ∂23-Nov-88 0442 @CUNYVM.CUNY.EDU:GOLUMBIC@ISRAEARN.BITNET
C01227 00330 ∂23-Nov-88 0647 harnad@Princeton.EDU re: BBBS Call for Neuroscience Nominations
C01229 00331 ∂23-Nov-88 0825 tom@polya.Stanford.EDU Dover
C01230 00332 ∂23-Nov-88 0900 JMC
C01231 00333 ∂23-Nov-88 0958 VAL NSF proposal
C01243 00334 ∂23-Nov-88 1055 scales@polya.Stanford.EDU qlisp
C01245 00335 ∂23-Nov-88 1152 drb@cscfac.ncsu.edu Re: reply to message
C01248 00336 ∂23-Nov-88 1157 VAL re: NSF report
C01249 00337 ∂23-Nov-88 1240 CLT unrestricted funds
C01250 00338 ∂23-Nov-88 1245 peters@russell.Stanford.EDU Re: [peters: Re: [peters: CSLI-Japan collaborations] ]
C01252 00339 ∂23-Nov-88 1415 MPS Partytime
C01253 00340 ∂23-Nov-88 1441 alex@jessica.Stanford.EDU TAing your class on NonMonotonic Logic Winter Quarter
C01255 00341 ∂23-Nov-88 1508 bowers@Popserver.Stanford.Edu Winter Schedule Changes
C01260 00342 ∂23-Nov-88 1727 ILAN@Score.Stanford.EDU [Ilan Vardi <ILAN@Score.Stanford.EDU>: Thanksgiving Theorem]
C01265 00343 ∂24-Nov-88 0841 GOODMAN%uamis@rvax.ccit.arizona.edu Happy Thanksgiving
C01267 00344 ∂24-Nov-88 0901 JMC
C01269 00345 ∂25-Nov-88 1602 CLT circus
C01270 00346 ∂25-Nov-88 1700 JMC philooλsophy
C01272 00347 ∂25-Nov-88 2150 beeson@ucscd.UCSC.EDU triangle problem solved!
C01275 00348 ∂26-Nov-88 0739 CLT
C01276 00349 ∂26-Nov-88 1924 pehoushe@Gang-of-Four.Stanford.EDU N-Queue Qlisp Demonstration
C01278 00350 ∂27-Nov-88 0956 harnad@Princeton.EDU Explanatory Coherence: BBS Call for Commentators
C01282 00351 ∂28-Nov-88 0318 cracraft@venera.isi.edu this'll be news by morning...
C01285 00352 ∂28-Nov-88 0744 MPS Filing
C01286 00353 ∂28-Nov-88 0900 JMC
C01288 00354 ∂28-Nov-88 1425 nilsson@Tenaya.stanford.edu DARPA for lunch?
C01290 00355 ∂28-Nov-88 1457 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU,@RITTER.AI.SRI.COM:TYSON@AI.SRI.COM [geoff@wacsvax.OZ: Input refutations]
C01295 00356 ∂29-Nov-88 0706 Rich.Thomason@cad.cs.cmu.edu AI and Philosophical Logic Book
C01297 00357 ∂29-Nov-88 0937 LASHER@Score.Stanford.EDU ["Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>:]
C01299 00358 ∂29-Nov-88 1147 ABRAIDO@Score.Stanford.EDU Surprise for Devika Subramanian
C01302 00359 ∂29-Nov-88 1547 MPS
C01303 00360 ∂29-Nov-88 1551 rpg@lucid.com US/Japan Workshop on Parallel Lisp
C01307 00361 ∂29-Nov-88 1603 GLB
C01308 00362 ∂29-Nov-88 1617 weening@Gang-of-Four.Stanford.EDU Re: US/Japan Workshop on Parallel Lisp
C01310 00363 ∂29-Nov-88 1650 CLT US/Japan Workshop on Parallel Lisp
C01311 00364 ∂29-Nov-88 1714 VAL Commonsense and Nonmonotonic Reasoning Seminar
C01313 00365 ∂29-Nov-88 2011 VAL Re: Commonsense and Nonmonotonic Reasoning Seminar
C01315 00366 ∂29-Nov-88 2121 BEDIT@Score.Stanford.EDU Summary of September computer charges.
C01318 00367 ∂30-Nov-88 0208 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Pi
C01333 00368 ∂30-Nov-88 0743 MPS Conferences
C01334 00369 ∂30-Nov-88 0940 MPS
C01335 00370 ∂30-Nov-88 1326 ginsberg@polya.Stanford.EDU Formfeed to meet on 12/1
C01336 00371 ∂30-Nov-88 1352 pehoushe@Gang-of-Four.Stanford.EDU Patenting a scheduling system?
C01338 00372 ∂30-Nov-88 1525 scales@polya.Stanford.EDU qlisp
C01340 00373 ∂30-Nov-88 1627 MPS Files
C01341 00374 ∂30-Nov-88 1807 weening@Gang-of-Four.Stanford.EDU
C01358 00375 ∂30-Nov-88 2145 ilan@Gang-of-Four.Stanford.EDU Computer chess breakthrough
C01367 00376 ∂30-Nov-88 2244 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM n log n
C01377 00377 ∂01-Dec-88 0355 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU:bundy%aiva.edinburgh.ac.uk@NSS.Cs.Ucl.AC.UK Research Job at Edinburgh
C01381 00378 ∂01-Dec-88 0720 CLT Umbrella
C01385 00379 ∂01-Dec-88 1159 pehoushe@Gang-of-Four.Stanford.EDU Qlisp Special Variables Benchmark
C01391 00380 ∂01-Dec-88 1256 beeson@ucscd.UCSC.EDU Question for circumscription theory
C01393 00381 ∂01-Dec-88 1350 VAL re: Question for circumscription theory
C01395 00382 ∂01-Dec-88 1554 MPS
C01396 00383 ∂01-Dec-88 1558 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
C01398 00384 ∂01-Dec-88 1628 MPS telephone call
C01399 00385 ∂01-Dec-88 1640 STAGER@Score.Stanford.EDU Spring CS309C
C01400 00386 ∂01-Dec-88 1648 BERGMAN@Score.Stanford.EDU NSF salary cap
C01403 00387 ∂01-Dec-88 1853 gurevich@polya.Stanford.EDU
C01404 00388 ∂01-Dec-88 2010 RFC Prancing Pony Bill
C01406 00389 ∂02-Dec-88 0546 Rich.Thomason@cad.cs.cmu.edu AI and Philosophical Logic Book Again
C01409 00390 ∂02-Dec-88 0812 pehoushe@Gang-of-Four.Stanford.EDU CTAK test
C01414 00391 ∂02-Dec-88 0857 gurevich@polya.Stanford.EDU Lunch
C01416 00392 ∂02-Dec-88 1020 chandler@polya.Stanford.EDU Faculty Repoirt
C01418 00393 ∂02-Dec-88 1129 weening@Gang-of-Four.Stanford.EDU
C01427 00394 ∂02-Dec-88 1252 pehoushe@Gang-of-Four.Stanford.EDU CONS Contention
C01429 00395 ∂02-Dec-88 1322 rpg@lucid.com CONS Contention
C01431 00396 ∂02-Dec-88 1358 pehoushe@Gang-of-Four.Stanford.EDU CONS Contention
C01437 00397 ∂02-Dec-88 1356 P.EPSTEIN@GSB-WHY.Stanford.EDU re: Justifying Schultz's decision about Arafat
C01438 00398 ∂02-Dec-88 1557 perrie@sumex-aim.stanford.edu Call from Ed Feigenbaum
C01439 00399 ∂02-Dec-88 1600 JMC
C01440 ENDMK
C⊗;
∂01-Oct-88 1355 GOODMAN%uamis@rvax.ccit.arizona.edu reviews of draft
Received: from rvax.ccit.arizona.edu by SAIL.Stanford.EDU with TCP; 1 Oct 88 13:54:45 PDT
Date: Sat, 1 Oct 88 13:20 MST
From: GOODMAN%uamis@rvax.ccit.arizona.edu
Subject: reviews of draft
To: DUANE.ADAMS@c.cs.cmu.edu, BLUMENTHAL@a.isi.edu, 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, KNEMEYER@a.isi.edu,
MCHENRY@GUVAX.BITNET, OUSTER@ginger.berkeley.edu, Ralston@mcc.com,
CWEISSMAN@dockmaster.arpa
X-VMS-To: @NAS
We now have essentially all of our reviews. These include written or oral
comments from the formal reviewers, the CSTB itself, and several other
reviewers who have read all or large parts of our manuscript. Altogether
over 20 people.
Not bad. I had frankly expected a lot more grief, just on rough statistical
grounds (i.e., from so many people on such a long draft covering so many
topics).
With the exception of one of our shortest middle chapters, each of our
chapters was covered in detail by at least 2 of the written reviews.
As could be expected, we have a healthy number of minor comments,
disagreements, and suggestions on very specific points. Some of these are
contradictory, and some reflect the fetishes of individual reviewers. Most
are helpful, and should enable us to improve the document. Each of the
principal chapter authors will be given copies of the relevant set of minor
comments for his/her chapter.
Of the middle Chs. 3-8, only one came under heavy general criticism (for
"lightness" and organizational problems), as did one section in another
chapter (mainly for definitional problems). Some serious further thought
will have to be given to these by their principal authors.
There was general agreement that we have a weak front end. However, there
was no general agreement as to how to get a better one. We are thinking of
replacing the Executive Summary with a much shorter "report brief" a la
OTA-style reports, and a strengthened introduction.
The reviewers generally thought we did well to go beyond the State SOW,
which really asked for little more than extensive state-of-the-art
tutorials. They also seemed to think we did well with the conclusions we
tried to draw (e.g., avoid getting into policy/politics, listing the most
important technologies to worry about). Many thought that we did not go as
far as we could have (or as far as they would have liked to see) in drawing
conclusions out of our work. None seemed sympathetic to our shortness of
time plight. Most wanted us to move in the direction of a longer, stronger
(rather than "a small number of bullets") final chapter. Marjory and I have
started working on this.
Marjory and I have spent much of last week discussing the reviews as they
have come in, and how we might accommodate or reject these comments.
We will be getting back to you individually and collectively soon. We also
have to write a report to NAS detailing our disposition of almost every
comment before we get approval for publication (standard NAS procedure).
∂02-Oct-88 0016 RFC Prancing Pony Bill
Prancing Pony bill of JMC John McCarthy 2 October 1988
Previous Balance 8.04
Monthly Interest at 1.0% 0.08
Current Charges 4.00 (bicycle lockers)
-------
TOTAL AMOUNT DUE 12.12
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 7/12/88.
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-Oct-88 0908 pehoushe@Gang-of-Four.Stanford.EDU Q(16) and Q(15) bet
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Oct 88 09:08:19 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA05086; Mon, 3 Oct 88 09:04:48 PDT
Date: Mon, 3 Oct 88 09:04:48 PDT
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8810031604.AA05086@Gang-of-Four.Stanford.EDU>
To: rdz@Gang-of-Four.Stanford.EDU
Cc: weening@Gang-of-Four.Stanford.EDU, pehoushek@Gang-of-Four.Stanford.EDU,
dam@WHEATIES.AI.MIT.EDU, jmc@sail
Subject: Q(16) and Q(15) bet
For JMC's and David's info, Ramin and I bet a dollar that Q(15) could
be solved in Qlisp in less than 2 hours, no use of symmetries allowed
(He bet that I couldn't, and I bet that I could, of course). John
should be happy to note that the speed-up is 8 minus epsilon, and that
the parallel code was easy to write, using QDOTIMES, a Qempty-P
implementation of DOTIMES.
To make the bet serve some mildly useful purpose, I calcuated Q(16),
which may not ever have been done before? In 8 processor qlisp,
dynamically spawning tasks in the first 4 columns, the Nqueens solver
found Q(16)=14772512 in 4.8 hours using one symmetry (searched half of
first column, * 2). There was a total of 2.1 minutes idle CPU time.
The predicted serial, no symmetry time is therefore about 76 hours.
Using one more symmetry, Q(17) should be solvable in less than 20
hours.
Status of Q(15) bet: In an unofficial run, 8 processor Qlisp solved
Q(15) (no symmetries) in 1.4 hours, with a minute of idle time, on 8
processors. The number of solutions was 2279184. The code is in
gang-of-four/user/pehoushe/nqueens%.lisp. I'll make an Official Q(15)
run tonight, using the Unix time function. I guess we are even now,
Ramin. I never worked so hard for a dollar.
The main serial Solver is listed below. It searches columns from left
to right, incrementing and decrementing attacked squares in the
unsearched columns. The penultimate column is searched without doing
any incrementing or decrementing. The serial running times are close
to David's, and so, I presume, is the search strategy.
;;; SOLVE counts the number of solutions given N, the current board, b,
;;; the distance from the end of the board, end, and the current unsearched
;;; column i (all columns between i and N-1 are also unsearched).
;;; The recursion is unrolled for quickly searching the penultimate column.
;;; In the actual code, I use +&'s and %svrefs to avoid type checking.
;;; The top-level call to solve is (SOLVE N (NxN array of 0's) (1- N) 0).
(defun solve (n b end i &aux (counter 0)(i+1 (+ i 1))(cur-col (svref b i)))
(if (= 1 end)
(let ((last-col (svref b i+1)))
(dotimes (j n)
(when (= 0 (svref cur-col j))
(dotimes (k (- j 1))
(when (= 0 (svref last-col k)) (setq counter (+ 1 counter))))
(dotimes (k (- n (+ j 2)))
(when (= 0 (svref last-col (+ k j 2)))
(setq counter (+ 1 counter)))))))
(dotimes (j n)
(when (= 0 (svref cur-col j))
(let ((max-iend-j (if (> i j) end (- (- n 1) j)))
(min-end-j (if (< end j) end j))
(j+1 (+ j 1)))
;; mark the rest of the row and diagonals
(dotimes (k end) (attack b (+ i+1 k) j))
(dotimes (k max-iend-j) (attack b (+ i+1 k) (+ j+1 k)))
(dotimes (k min-end-j) (attack b (+ i+1 k) (- j (+ 1 k))))
(setq counter (+ counter (solve n b (- end 1) i+1)))
;; unmark the rest of the row diagonals
(dotimes (k end) (unattack b (+ i+1 k) j))
(dotimes (k max-iend-j) (unattack b (+ i+1 k) (+ j+1 k)))
(dotimes (k min-end-j) (unattack b (+ i+1 k) (- j (+ 1 k))))))))
counter)
∂03-Oct-88 1000 JMC
reopen Stanford civil rights issue
∂03-Oct-88 1001 MPS phone call
Please call Dr. Kornberg, Cell Biology (3-6988).
∂03-Oct-88 1005 lrosenbe@note.nsf.gov Re: coordination theory
Received: from note.nsf.gov by SAIL.Stanford.EDU with TCP; 3 Oct 88 10:05:32 PDT
To: hayes.pa@XEROX.COM
cc: JMC@SAIL.STANFORD.EDU, lrosenberg@note.nsf.gov
Subject: Re: coordination theory
In-reply-to: Your message of 27 Sep 88 09:20:00 -0700.
<880927-092214-5833@Xerox>
Date: Mon, 03 Oct 88 12:42:00 -0400
From: Laurence Rosenberg <lrosenbe@note.nsf.gov>
Message-ID: <8810031242.aa18461@note.nsf.gov>
Pat, for openers, start with Bernardo Huberman's Open Systems workshop
at your very establishment--his book published this year by North-Holland
"Ecology of Computation", contains articles describing at least one thread
of research that we consider to be "coordination
theory". Then, go to Tom Malone's two workshops on Coordination Theory
held this year and last year in Mass. He also has some articles I am
sure he would be glad to share with you. John S. Brown may have
some of them, as he attended the last workshop.
I hope this helps.
Larry
∂03-Oct-88 1309 JK
John --- FYI
------------
∂03-Oct-88 1252 lrosenbe@note.nsf.gov
Received: from note.nsf.gov by SAIL.Stanford.EDU with TCP; 3 Oct 88 12:52:02 PDT
To: Jussi Ketonen <JK@SAIL.STANFORD.EDU>
In-reply-to: Your message of 29 Sep 88 11:57:00 -0700.
<gjpPs@SAIL.Stanford.EDU>
Date: Mon, 03 Oct 88 15:33:26 -0400
From: Laurence Rosenberg <lrosenbe@note.nsf.gov>
Message-ID: <8810031533.aa00517@note.nsf.gov>
dear jussi,
now that i am back in the office let me give you a more thoughtful response
to your proposed research. i continue to think it fits nicely within the
spirit and intent of my coordination theory and technology announcement and
i hope you apply. as to winning--i can only wish you luck. I particularly
think you have assembled a good multi-disciplinary team for the work, one
which is sensitive to the need to understand distributed software systems
in the context of human organizations. as to expected format--it is the same
as for other nsf proposals--the cover page should clearly indicate that the
proposal is targeted for the coordination theory initiative, though. I take
it that you intend the primary submitting institution to be Stanford, so
other institutions, such as Harvard, can sub-contract, or individuals can
be consultants, or they can even come to Stanford, etc. But the whole effort
should funnel thru one institution. i hope this helps. larry.
∂03-Oct-88 1601 @Score.Stanford.EDU:ellis@src.dec.com ACLU
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Oct 88 16:01:20 PDT
Received: from decwrl.dec.com by SCORE.STANFORD.EDU with TCP; Mon 3 Oct 88 15:34:08-PDT
Received: from jumbo.pa.dec.com by decwrl.dec.com (5.54.5/4.7.34)
id AA16749; Mon, 3 Oct 88 15:34:58 PDT
Received: by jumbo.pa.dec.com (5.54.4/4.7.34)
id AA24545; Mon, 3 Oct 88 15:34:52 PDT
Date: Mon, 3 Oct 88 15:34:52 PDT
From: ellis@src.dec.com (John R. Ellis)
Message-Id: <8810032234.AA24545@jumbo.pa.dec.com>
To: jmc@score.stanford.edu
Subject: ACLU
I enjoy reading your comments on su.etc and saw some of your comments
about the ACLU, so I'd thought I'd pass this along: Today's WSJ
editorial page has a piece describing in detail several of the ACLU's
purported main-stream positions, taken directly from the ACLU's own
policy statement.
∂04-Oct-88 0908 GOODMAN%uamis@rvax.ccit.arizona.edu suggestions for concluding chapter
Received: from rvax.ccit.arizona.edu by SAIL.Stanford.EDU with TCP; 4 Oct 88 09:08:01 PDT
Date: Tue, 4 Oct 88 08:47 MST
From: GOODMAN%uamis@rvax.ccit.arizona.edu
Subject: suggestions for concluding chapter
To: DUANE.ADAMS@c.cs.cmu.edu, BLUMENTHAL@a.isi.edu, 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, KNEMEYER@a.isi.edu,
MCHENRY@GUVAX.BITNET, OUSTER@ginger.berkeley.edu, Ralston@mcc.com,
CWEISSMAN@dockmaster.arpa
X-VMS-To: @NAS, THORNTON, GOODMAN
What follows are some thoughts on how we might revise the final
chapter. Please let Marjory or me have your comments and suggestions
ASAP.
---
Most of the people on our committee felt that the original State statement
of work was too limited, and thought it within our purview to draw
conclusions that had more direct relevance to export control (and, to a
lesser extent, the competitiveness issue) than a completely detached
survey/tutorial.
The reviewers felt the same way, and perhaps more so. They are saying we did
not give enough attention and focus to the conclusions that are implicit or
explicit in the analyses of Chs. 3-8. They want more of them to be brought
out in higher profile. Some seemed to sense our tentativeness about going
beyond the State SOW. Others apparently did not read or digest the State
task as explained in the Intro. Some also do not think we said enough about
further work that needs to be done. The view that we should have a longer
and more extensive final chapter clearly dominates the view that we should
have a very short concluding chapter with a small number of bullets.
Let me suggest that we think about the following. The final chapter would
have two sections:
1. A summary discussion that helps to pull together the main and integrating
thoughts from Chs. 3-8. We already have most of this with the material on
globalization, rate of innovation, Soviet tech transfer, etc. We must have
this for 2 important reasons: to pull the main ideas together for the reader
who has read all the separate chapters; and to pull the main ideas together
for those who have not.
2. A section of conclusions/recommendations that make sense to someone who
has read the summary discussion in 1. In addition to what is already there
under recommendations, we should include such material as the list of what
is most important to control and some other things that would help make a
better separation between the summary discussion and concl/recs. This
section might start with a list of fairly general conclusions/propositions,
e.g., focus limited resources for control on a small number of very
important technologies (see below).
From the reviews, I have compiled the list that follows. At least one
reviewer or committee member thinks that each item on the list has not
received proper attention (either none at all or not as much as it might).
All are reasonable enough to deserve consideration for inclusion in the
final chapter (and those not considered important enough, might end up in
the conclusions of the intermediate chapters). We would appreciate input
from you that augments, clarifies, provides good examples for, or
substantively contradicts any of the candidate statements below. These are
listed in no particular order.
---
With the globalization of technology, and increased foreign
capabilities, the US should be more aware of and aquisitive of
what is available abroad to help strengthen us at home. We should
improve the inflow in addition to increasing (trade) or
decreasing (export control) the outflow .
We should clarify the trade-off of (nonpolitical) choices facing the US. The
US has much to lose in certain kinds of export control environments. For
example, delays in export control decisions (esp. West-West) can hurt the
market entry of US companies at certain critical periods.
Small machines are becoming increasingly powerful and capable at a faster
rate than large machines. This is where much of the globalization and
commoditization is taking place. There are some serious potential military
uses for these machines. The US may simply have to learn to live with this.
Control efforts need to be focused. There are several arguments for this,
not the least of which is that the resources for control efforts are
limited, and growing much more slowly than globalization/innovation etc. The
most sensible strategy under the circumstances is to "build high fences"
around a small number of very important technologies. There is general
agreement that we have done a good job identifying these technologies (we
probably should add some key disk storage technologies to the list).
We need a more flexible means for recognizing and accommodating
rapid change in these technologies for export control purposes.
We need much improved national coverage of emerging technologies
and foreign availability. The present system is very large and
slow. A smaller group of especially capable people might be more
effective. Perhaps some sort of "court" system could be used to
decide questions of foreign availability or protectability?
More generally, the US is not well informed about what is going
on abroad. Something must be done about this, and done in such a
way that the improved coverage is made more widely available to
US interests than is now the case with what is done in the
intelligence community.
IC testing equipment must be added to the list of what must be
controlled in "onesies and twosies."
Much more attention needs to be given to non-CMEA foreign efforts
to acquire US technology (esp. East and South Asia).
The discussion of supercomputers and other high speed architectures is much
too focused on hardware. The success or failure of many models and
applications may depend at least as much on other factors as on the power of
the CPU. Heading this other list are: software, storage (both capacity and
accessibility - the latter has not gotten much attention), and I/O. Many
applications depend on a lot more than just raw number crunching of data
that somehow magically gets into the ALU unit.
We have not considered some technologies that could have significant
military applications, but which are not now, nor may ever become, popular
or cost effective to the point of reaching the commodity level. Once
example is bubble memory, which has high cost but certain environment and
portability features that may make it a technology to be controlled.
We pay very little attention to non-physical means of technology transfer.
In many ways, product transfer is a weak form of technology transfer, but
they are the focus of export controls. We might argue that any attempt at
broad controls in the emerging global technological environment would
require controls on non-physical tech transfer mechanisms. We might also
argue that this is probably very difficult, if not impossible, given the
resources at US/CoCom disposal, and the changing nature of that global
environment. It is very important for a future study to follow-up on this
concern.
A study like ours serves various purposes. But to be effective, it needs to
be updated and reconsidered about every 3 years.
We might explicitly repeat more of the chapter-level conclusions in Sect. 2
of the final chapter. There are at least 2 possibilities: simply have a
subsection for each chapter that restates the chapter conclusions; or take
what matters most from each of the chapter-level conclusions and weave that
in with the other material.
∂04-Oct-88 1045 MEERSMAN%HTIKUB5.BITNET@forsythe.stanford.edu Your text for the China proceedings
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 4 Oct 88 10:45:00 PDT
Received: by Forsythe.Stanford.EDU; Tue, 4 Oct 88 10:44:21 PDT
Date: Tue, 4 Oct 88 18:42 N
From: <MEERSMAN%HTIKUB5.BITNET@Forsythe.Stanford.EDU>
Subject: Your text for the China proceedings
To: jmc@sail.stanford.edu
X-Original-To: jmc@sail.stanford.edu
Dear Prof. McCarthy,
Have you been able to process the typed-out text of your presentation in
Guangzhou? We (the editors of the proceedings, that is) are starting to
feel the pressure to deliver the proceedings to the publishers. Can you
please tell us when we may look forward to the edited text? We shall need it
definitely by Oct. 17 lest we upset everybody's schedule.
Note that I am perfectly prepared to make the camera-ready version here if you
deliver just the text.
Sincerely,
Prof. Robert A. Meersman
PS pls acknowledge --thank you
∂04-Oct-88 1201 gurevich@polya.Stanford.EDU Hello
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 4 Oct 88 12:00:58 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA23277; Tue, 4 Oct 88 11:59:25 PDT
Date: Tue, 4 Oct 88 11:59:25 PDT
From: Yuri Gurevich <gurevich@polya.Stanford.EDU>
Message-Id: <8810041859.AA23277@polya.Stanford.EDU>
To: JMC@SAIL
Subject: Hello
This is Yuri Gurevich. I am spending this year hear
(found usually in Pratt's office; Vaughan is on a leave).
I'll teach a course or seminar on something related to logic and CS
in Winter Quarter.
Would you like to have a lunch one day (with me) ?
It can be in the faculty club; they gave me a number too.
-Yuri
∂04-Oct-88 1310 weening@Gang-of-Four.Stanford.EDU Computer Constellations
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 4 Oct 88 13:10:04 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA09072; Tue, 4 Oct 88 13:06:24 PDT
Date: Tue, 4 Oct 88 13:06:24 PDT
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8810042006.AA09072@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Computer Constellations
You might be interested in this paper that Ralph is working on. (He
sent a message to the nethax mailing list asking for comments.) I can
get a printed version, if you don't want to read it online.
@style(fontfamily timesroman, fontscale 12, spread 1.1)
@begin(text)
@modify(heading, below 0.15 inch, above 0.15 inch)
@modify(subheading, below 0.15 inch)
@pageheading(Left "@b(DRAFT) @value[date]", Center "Computer Constellations",
Right "@value(page)")
@heading(Computer Constellations)
@heading(Management of Distributed Computing Resources)
@center[Ralph E. Gorin, Academic Information Resources]
@center[Stanford University]
@center[@b(DRAFT)@ @ @ @value(date)@ @ @ @b(DRAFT)]
@i(The purpose of this draft document is to arouse comment within the
Stanford computing community. The author especially wishes to be
informed of the omissions perceived by readers. Please direct
specific comments to gorin@@jessica.stanford.edu.)
@subheading(Abstract)
Two key challenges face the providers of computer services in the late
1980s: first, to promote more ubiquitous use of computers throughout
the community by providing each person with tools appropriate to his
or her profession without turning that person into a computer systems
administrator; our community of professionals and their support staff
need homogeneous access to heterogeneous computer and data resources.
Second, to develop and deploy appropriate tools by which we, as providers
of computer services, can manage a complex collection of computers;
the managers of computer resources need homogeneous tools to control and
operate heterogeneous computer and data resources.
This paper is a high-level description of the function of a Computer
Constellation, a distributed, object-oriented, capability-based
software environment that addresses both the evolving user needs and
the resulting management needs. The word ``constellation''
describes a brilliant cluster or gathering.
Neither system users nor system administrators have been widely
consulted in the preparation of this draft. Therefore, the
description of the Computer Constellation is probably incomplete.
A meeting of qualified interested parties - mostly administrators
of distributed heterogenous systems - should be held to augment both
the scope and the detail of the present description and to define
a process by which the Constellation can be implemented. Since
an implementation of the Constellation concept necessary spans
the products of many vendors, a vendor-neutral forum is needed for
the continuation of this discussion.
@subheading(Background)
In the beginning, there was batch. There was no interaction.
The goal of timesharing was to provide interactive tools by which
people could accomplish their computing. The implementation of
timesharing brought many people together on the same computer; in
timesharing, we struggled to provide the illusion that each person
operated independent of the others. The systems provided high
function and high complexity. The administration of a timesharing
system could not satisfy every demand of every user: when
administrative requirements (e.g., preventive maintenance, system
backups, software upgrades) would dictate, the system would be
unavailable to all users.
Personal computers took timesharing's goal of interaction to its
logical end: one computer was entirely dedicated to solving one
person's problem. The performance of the computer-human interaction
was brought to a higher standard. The reality of each person
operating independent of others was achieved. Gone were the days of
complex computer systems administered by officious bureaucrats. On
the other hand, each user became personally responsible for the
administrative tasks formerly performed by others.
There are upsets to the serenity of personal computer users: two
problems have arisen. First, most people are not or do not want to be
totally independent of others: from the institutional perspective, we
struggle with personal computers to rebuild the sociology of
timesharing in which users would share results and help each other.
Second, people who use personal-size computers have discovered that
their needs are now more complex; hence, they need more complex
computer environments: in short, there is a growing demand for both
desktop timesharing and desktop access to shared services and
resources.
Two technologies can be applied to address these problems in the world
of personal-sized computers. To provide for communication and
sharing, we add computer networks; to meet the growing complexity of
applications, we provide operating systems that support true
multi-tasking. Of course, the solution to one set of problems
is often the seed for a new generation of problems. Networking
and complex operating environments conspire to make the personal
computing environment harsh and unwelcoming. The computer constellation
is intended to tame the complexities of the user's computer environment
and reduce chaos in the system manager's environment.
@subheading(Constellation Goals)
The computer constellation is software that flexibly supports the
resource-management objectives of one administrative entity which is
charged with the control and operation of a specific, arbitrary
collection of physical computer resources that are linked by a
network. The constellation provides one control point through which
the people who manage these resources can effect whatever operational
policies they determine.
Since organizations are complex, one organization (a corporation, a
university, a government agency) may find it necessary to operate
several independent constellations. Multiple constellations can
co-exist and cooperate on the same network. Each is separately
administered; each implements the policies appropriate to it. The
management decisions made in one constellation (ideally) have no
affect on the operation of other constellations. (The ideal may not
be met when network components are shared between different
constellations.) The administrators of one constellation need not
place limitless ``trust'' in the administrators of another constellation:
each constellation defines what resources and capabilities to share with
others.
Further, the constellation provides one set of techniques by which users of
these resources access and manipulate their files, regardless of where
in the constellation their files happen to be.
Among the computer resources that may be linked in a constellation are
@begin(itemize)
@b(Multi-Servers): these may be general timesharing systems,
compute-servers, super-computers, file servers, print servers, or any
other configuration characterized by being subject to use by several
people simultaneously, or which organizes and handles requests from
different people.
@b(Independent Private Workstations:) a workstation is a
personal-sized computer, connected to a network, that is used by only
one person at a time. A private workstation is owned by (or assigned
to) an individual who is the primary user of the workstation and whose
files are stored on that workstation. A private workstation has
discernable state: it remembers the owner's files from one session to
the next. The owner of an independent workstation is an occasional
user of constellation resources: he or she may make casual use of
facilities such as mail or access to the constellation library. The
owner of an independent workstation surrenders very little of his or
her independence to connect to a constellation on a casual basis. But, the
constellation provides only a limited scope of services to such a user.
@b(Constellation Private Workstations:) a private workstation which,
together with its owner, has extensive formal ties to the constellation.
Such a workstation is expected to be on-line to the constellation
on a (nearly) continuous basis so that mail can be delivered to it,
and so that files can be backed-up remotely. The owner of a constellation
private workstation has relinquished a great deal of control over the
software environment that he or she runs; however, system administration
tasks, e.g., file backup, the installation of new software, are done
on behalf of such users, without their personal intervention.
@b(Public Workstations:) a workstation that is used serially by
different people; the users are not assured further access to a
particular machine after the end of their session. It is necessary
that one idle public workstation be the same as all other idle public
workstations: public workstations are devoid of state particular to
the individual user. The storage of personal files on these systems
is prohibited; file storage for a user of public workstations must be
provided elsewhere.
@end(itemize)
@b(User Perspective: the Look and Feel of the Constellation)
@begin(itemize)
The individual user of the constellation has one user-name and
corresponding password by which all access to the constellation
is controlled.
Use of the constellation (and access to constellation resources) from
a constellation workstation shall be, so far as practicable, identical
to the use (and access) from any constellation workstation. That is,
where practicable, users will use the same commands and techniques to
access constellation resources from all workstations. In particular,
users of constellation private workstations (and on-line independent
workstations) can access their personal files from public
workstations.
Users of public workstations are provided with a standard environment,
free from any effects caused by the previous users of the workstations.
One user profile, changable by the user, defines the customizations
which that person wants applied to the standard environment.
These customizations are applied on behalf of the user whenever
he or she is using the the constellation.
One technique, applied by the user, controls access to that person's
files.
The standard environment (it may initially be the C-Shell) would be
presented as the front-end to a user-selectable windowing environment.
User files (whether stored on constellation private workstations or on
multi-servers) are preserved on backup media to lessen the damage
caused by system accidents or by user mistakes. Backup copies of
files are restored on request. @i(Further topic: archival storage.)
Electronic mail addressed to the user at the constellation will be
delivered to the server specified by the user's profile.
The user can access the constellation's library of licensed software.
The user automatically gets up-to-date, tested, versions of software.
@i(How does the user physically wire an independent personal
machine to the constellation?)
@i(How does an independent private workstation identify itself to the cluster?
How does its owner identify his machine when he accesses it remotely?)
@i(How does the independent system first obtain constellation software?)
@end(itemize)
@b(Management Perspective: Controlling User Access to Resources)
@begin(itemize)
Access to constellation resources is controlled by a database of
capabilities; a set of capabilities is associated with each
user-name and password pair. Inside one constellation, one
person's access to resources is controlled through one user-name
and the associated capabilities. The union of a user's capabilities
defines the extent of the (constellation) resources granted to
that person. Within a constellation, user names are unique.
User names in one constellation are independent of user names
in any other constellation.
Whatever policies govern the use of constellation resources, those
policies are implemented by this database. Thus, alternative polices
for different constellations can readily be accomodated.
Constellation services need be granted only to persons known to the
constellation and only to the extent determined by constellation
management. By management option, certain ``public services'' (e.g.,
anonymous ftp, help, directory services, etc.) can be provided by the
constellation.
Cross-constellation access is supported: a person known both in
constellation ``A'' and in constellation ``B'' (possibly by
different user names) can use a workstation in constellation
``B'' to access the data and resources allowed by ``A''.
At the option of the constellation's management, ``anonymous''
cross-constellation access can be granted from public workstations: at
the option of the manager of constellation ``B'', a person, unknown to
``B'' but known in constellation ``A'', may use a workstation in
constellation ``B'' to access the data and resources allowed by ``A''.
If this use of public workstations is permitted, constellation ``B'' is
used as a conduit through which a user of ``A'' is connected to the ``A''
resources allowed to that person.
Requests for service are authenticated by secure systems in which
passwords are not transmitted between network hosts, and in which
authentication capabilities cannot be conterfeit.
Constellation services can be made available at any location reachable
by network.
A constellation may provide services similar to those provided by
an internet domain. A directory of constellation users and a
directory of constellation host resources is maintained and is
made available electronically both for constellation internal
purposes and for external purposes to the extent determined by
the management of the constellation.
Constellations provide a manageable focal point for the administration
(and cost containment) of shared libraries of software.
By standardizing the user interface, by providing all users with
tested, up-to-date releases of software, the constellation provides an
environment in which support costs can be contained, and in which
users can communicate their knowledge of the environment and learn
from each other.
In a diverse environment, the computer constellation provides first, a
clear partition of responsibilities among organizations and individual
participants, and second, a straightforward means by which those
responsibilities can be discharged.
The constellation will gather accounting and resource usage information for all
constellation systems, in a common format, at one location. An audit trail
for each monitored activity will be collected at one location.
@end(itemize)
@subheading(Conclusion)
The detailed definition of a computer constellation is incomplete.
Although the definition is written without prejudging the nature of an
organization that might use the constellation concept, the definition,
to date, has been influenced only by discussions within the academic
community. To broaden the scope of applicability of constellations,
and to start the effort to define the detatiled requirements of a
constellation, further work should proceed in a vendor-neutral forum
which also can be the focal point for coordinating the implementation
of constellation management software by diverse organizations.
@end(text)
∂04-Oct-88 1533 MPS Franklin Speller
Bookstore called again to say they had this speller.
They would like to know if you want it or shall they put
it on the shelf.
Pat
∂04-Oct-88 1811 lincoln@polya.Stanford.EDU Computron usage
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 4 Oct 88 18:11:03 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA16803; Tue, 4 Oct 88 18:09:38 PDT
Date: Tue, 4 Oct 1988 18:09:36 PDT
Sender: "Patrick D. Lincoln" <lincoln@polya.stanford.edu>
From: "Patrick D. Lincoln" <lincoln@polya.stanford.edu>
To: jmc@sail
Cc: lincoln@polya.Stanford.EDU
Subject: Computron usage
Message-Id: <CMM.0.87.592016976.lincoln@polya.stanford.edu>
Hi. Im a new Phd student in the department, and Nils has pointed me to
you -
I would like to finish a project I started at UT Austin with Bob Boyer,
using his Prover. The fascists here count every computron, and Ive
been told explicitly not to use the only computer accounts ive got on
things like the BMTP. So i would like to ask you if you could sponsor
a bit of computer time for me to finish my project.
What i would like is an account on Gang-of-Four, or Polya that would allow
me to use Bob's prover - Or access to a lisp machine or a sun.
Thank you for any help. pdl.
∂04-Oct-88 2041 ARK Joe Weening
To: bergman@SCORE.STANFORD.EDU, littell@POLYA.Stanford.EDU
CC: ARK@SAIL.Stanford.EDU, JMC@SAIL.Stanford.EDU,
JSW@SAIL.Stanford.EDU, wiederhold@SUMEX-AIM.Stanford.EDU
I'm not sure if I gave you this instruction before or not.
Please charge 1/4 of Joe Weening's RA (that's 12.5%) to
2DMA494 (NSF Paradata) for fall quarter. Thanks.
Arthur
∂05-Oct-88 0924 ARK Paying for the Alliant
To: JMC@SAIL.Stanford.EDU
CC: JSW@SAIL.Stanford.EDU, ARK@SAIL.Stanford.EDU, wiederhold@SUMEX-AIM.Stanford.EDU
As a "good faith" gesture, I've sent a message to the administrators to
have me start paying for a piece of Joe as part of the package that pays
for use of the Alliant. It's important to do that on a timely basis,
because one people charges are allocated, they cannot be transferred.
I propose that we wait until the end of October, and then generate actual
statistics on my project's usage of the Alliant, and our decision on
charging can be based on actual data rather than projections.
Arthur
∂05-Oct-88 0947 JK MAD
FYI: 5 people from the development organization have left MAD during the
last 3 working days. I suspect there is more to come.
∂05-Oct-88 1356 JK
filename: functx.lsp[ekl,jk]
∂05-Oct-88 1530 hoffman@csli.Stanford.EDU re: a follow up appointment?
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Oct 88 15:30:06 PDT
Received: by csli.Stanford.EDU (4.0/4.7); Wed, 5 Oct 88 15:30:03 PDT
Date: Wed 5 Oct 88 15:30:02-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: re: a follow up appointment?
To: JMC@SAIL.STANFORD.EDU
Message-Id: <592093802.0.HOFFMAN@CSLI.Stanford.EDU>
In-Reply-To: <1#mtGk@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Would Friday or next week work? I am temporarily detained from doing much
in the way of Stanford work due to a financial situation of my father's. I
would also like to have a copy of the SSP essay before I see you.
Thanks!
reid
-------
∂05-Oct-88 1544 hoffman@csli.Stanford.EDU re: a follow up appointment?
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Oct 88 15:44:24 PDT
Received: by csli.Stanford.EDU (4.0/4.7); Wed, 5 Oct 88 15:44:22 PDT
Date: Wed 5 Oct 88 15:44:21-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: re: a follow up appointment?
To: JMC@SAIL.STANFORD.EDU
Message-Id: <592094661.0.HOFFMAN@CSLI.Stanford.EDU>
In-Reply-To: <#myw2@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Let's try next week. Today and tomorrow are caught in vice-grips by my
attempting to meet my father's suddenly urgent computer assistance needs.
(and he pays the bill for stanford.) Suggest a time, I am free any time
except Wens afternoon and friday morning.
reid
-------
∂05-Oct-88 1551 hoffman@csli.Stanford.EDU re: a follow up appointment?
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Oct 88 15:51:42 PDT
Received: by csli.Stanford.EDU (4.0/4.7); Wed, 5 Oct 88 15:51:35 PDT
Date: Wed 5 Oct 88 15:51:33-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: re: a follow up appointment?
To: JMC@SAIL.STANFORD.EDU
Message-Id: <592095093.0.HOFFMAN@CSLI.Stanford.EDU>
In-Reply-To: <qmygx@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
I am sometimes reachable at 948-2560 or at 321-8942. both (415). Since, you
have warned me, you need not get in touch with me. I'll just show up Monday
and we'll take it from there (in other words, don't worry if you can't get
in touch with me.
thanks
reid
-------
∂05-Oct-88 1639 ginsberg@polya.Stanford.EDU formfeed resumes on 10/6
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 5 Oct 88 16:39:10 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA13914; Wed, 5 Oct 88 16:37:03 PDT
Date: Wed, 5 Oct 88 16:37:03 PDT
From: Matthew L. Ginsberg <ginsberg@polya.Stanford.EDU>
Message-Id: <8810052337.AA13914@polya.Stanford.EDU>
To: feed@polya.Stanford.EDU
Subject: formfeed resumes on 10/6
Don't forget! Meet at 12.10 in room 252. The extra 10 minutes is
to allow anyone with a class ending at 12 to make it over ... (Like
me, for example.)
See you there!
Matt
∂06-Oct-88 0007 helen@psych.Stanford.EDU Re: lunch
Received: from psych.Stanford.EDU by SAIL.Stanford.EDU with TCP; 6 Oct 88 00:00:39 PDT
Received: by psych.Stanford.EDU (3.2/4.7); Wed, 5 Oct 88 23:54:44 PDT
Date: Wed, 5 Oct 88 23:54:44 PDT
From: helen@psych.Stanford.EDU (Helen Cunningham)
To: JMC@SAIL.Stanford.EDU
Subject: Re: lunch
Hello there and welcome back,
Best times for me are still Saturdays, so let's say NEXT Saturday
October 15. We can discuss the great debates!
Have fun at Great America. I'll be interested in hearing how it was.
-helen
∂06-Oct-88 0751 beeson@ucscd.UCSC.EDU triangles
Received: from ucscd.UCSC.EDU ([128.114.129.2]) by SAIL.Stanford.EDU with TCP; 6 Oct 88 07:51:30 PDT
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA17804; Thu, 6 Oct 88 07:53:03 PDT
Date: Thu, 6 Oct 88 07:53:03 PDT
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8810061453.AA17804@ucscd.UCSC.EDU>
To: jmc@sail.stanford.edu
Subject: triangles
Remember I was searching for a triangle that could be embedded in 4-space
but not in 3-space? The embeddability depends only on the ratio sqrt(k)
of altitude to base, i.e. the tangents are in Q(sqrt(k)). Well, I've
numerically established that any example must have k > 128. That is,
I've examined all k <= 128 and found no counterexample to the following
proposition: The following three are equivalent:
(1) k is a sum of three squares
(2) the equations are solvable in three dimensions
(3) the equations are solvable in four dimensions.
There is no a priori reason for the equivalence of any two of these things,
let alone all three. Note that it's not just that I found no counterexample,
I established that there IS NO counterexample <= 128. Thus one of two
surprising things is true: either there is a counterexample, but the least
such is more than 128, or there is no counterexample, i.e. the same triangles
are embeddable in 3-space and 4-space.
∂06-Oct-88 0959 MPS
Other than the photo album, the only photos I have and
the ones we've been using for passport visa, journals, etc.
There are a couple in there, but with other people.
Stacey called to ask if you would be available on the 15th
between 12-3. She needs a confirmation today as she has to
fax your reply. Her number is 3-2085
Pat
∂06-Oct-88 1224 MPS Phone call
Prof. Hurd called and said he would try you at home.
If not, he would like you to call 854-1901
Pat
∂06-Oct-88 1450 ginsberg@polya.Stanford.EDU Formfeed to remain on Thursday
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 6 Oct 88 14:50:27 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA02961; Thu, 6 Oct 88 14:46:07 PDT
Date: Thu, 6 Oct 88 14:46:07 PDT
From: Matthew L. Ginsberg <ginsberg@polya.Stanford.EDU>
Message-Id: <8810062146.AA02961@polya.Stanford.EDU>
To: feed@polya.Stanford.EDU
Subject: Formfeed to remain on Thursday
Unfortunately (for me), the scheduling of the architectures meeting
seems to be more sporadic than simply every other week. So I fear
that we will have to leave formfeed on Thursdays, instead of moving
it (the architecture folks are meeting on 10/19, for example). This
means that the next meeting is on Thursday, 10/20, at 12.10 in MJH 252.
Matt
∂06-Oct-88 1515 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
THE ALTERNATING FIXPOINT OF LOGIC PROGRAMS WITH NEGATION
Allen Van Gelder
University of California at Santa Cruz
Friday, October 7, 3:15pm
MJH 301
The simplest logic programs use only rules in the form of Horn clauses.
That is, each rule can be described as a goal that can be achieved by
achieving zero or more "positive" subgoals. An important practical
extension is the incorporation of "negative" subgoals, as well. The
idea is that the goal of the rule succeeds when the "positive" subgoals
succeed and the "negative" subgoals fail, in some sense.
Horn clause programs have an intuitive and well-accepted semantics
defined by the least fixpoint of an operator that essentially performs
modus ponens reasoning. Several early attempts to extend this operator
to programs with negative subgoals ran into problems of one sort or
another. For example, a fixpoint proposed by Fitting turned out to
yield a language that is unable to express the complement of transitive
closure.
Two recent proposals to improve matters are named "stable models", due
to Gelfond and Lifschitz, and "well-founded partial models", due to
Van Gelder, Ross, and Schlipf.
Both stable models and well-founded partial models were defined somewhat
nonconstructively, in the sense that certain sets could be recognized
if presented, but no algorithm to construct them from the program was
given.
Today I will describe the Alternating Fixpoint of a logic program,
which gives a construction of the well-founded partial model. The
underlying idea is to monotonically build up a set of negative
conclusions until the least fixpoint is reached. From a fixed set of
negative conclusions, we can derive the positive conclusions that
follow (without deriving any further negative ones), by traditional
Horn clause semantics. The name "alternating" was chosen because the
transformation runs in two passes; the first pass transforms an
underestimate of the set of negative conclusions into an (intermediate)
overestimate; the second pass transforms the overestimate into a new
underestimate; the composition of the two passes is monotonic.
If time permits, I will discuss the expressive power of alternating
fixpoints vis-a-vis "inductive fixpoints".
∂06-Oct-88 1612 CLT noise reduction project
If you think something can be done without infinite time
wasted in useless committee meetings I would be glad
to take part.
∂06-Oct-88 1639 jwalton@vax.darpa.mil 1988 Principal Investigators' Conference
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 6 Oct 88 16:39:49 PDT
Received: from sun16.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA02806; Thu, 6 Oct 88 17:07:37 EDT
Posted-Date: Thu 6 Oct 88 17:08:48-EDT
Received: by sun16.darpa.mil (5.54/5.51)
id AA00234; Thu, 6 Oct 88 17:08:53 EDT
Date: Thu 6 Oct 88 17:08:48-EDT
From: Juanita Walton <JWALTON@vax.darpa.mil>
Subject: 1988 Principal Investigators' Conference
To: ISTO-PI-LIST@vax.darpa.mil
Cc: jwalton@vax.darpa.mil
Message-Id: <592175328.0.JWALTON@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Ladies and Gentlemen:
Please be advised that the conference brochure and other details were only
mailed out this week. If you have not received the information by next
Tuesday (Oct. 11), please feel free to contact Juanita Walton. Also,
please note that the best number on which to reach Ms. Walton is:
703/276-3533.
-------
∂06-Oct-88 1742 jwalton@vax.darpa.mil 1988 Principal Investigators' Conference
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 6 Oct 88 17:42:05 PDT
Received: from sun16.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA02806; Thu, 6 Oct 88 17:07:37 EDT
Posted-Date: Thu 6 Oct 88 17:08:48-EDT
Received: by sun16.darpa.mil (5.54/5.51)
id AA00234; Thu, 6 Oct 88 17:08:53 EDT
Date: Thu 6 Oct 88 17:08:48-EDT
From: Juanita Walton <JWALTON@vax.darpa.mil>
Subject: 1988 Principal Investigators' Conference
To: ISTO-PI-LIST@vax.darpa.mil
Cc: jwalton@vax.darpa.mil
Message-Id: <592175328.0.JWALTON@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Ladies and Gentlemen:
Please be advised that the conference brochure and other details were only
mailed out this week. If you have not received the information by next
Tuesday (Oct. 11), please feel free to contact Juanita Walton. Also,
please note that the best number on which to reach Ms. Walton is:
703/276-3533.
-------
∂07-Oct-88 0730 @Score.Stanford.EDU:RINDFLEISCH@SUMEX-AIM.Stanford.EDU Action Items from Our Meeting
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Oct 88 07:30:16 PDT
Received: from SUMEX-AIM.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Fri 7 Oct 88 07:27:08-PDT
Date: Fri, 7 Oct 88 07:23:44 PDT
From: TC Rindfleisch <Rindfleisch@SUMEX-AIM.Stanford.EDU>
Subject: Action Items from Our Meeting
To: Facil@Score.Stanford.EDU
cc: Rindfleisch@SUMEX-AIM.Stanford.EDU, Nilsson@Score.Stanford.EDU
Message-ID: <12436530742.27.RINDFLEISCH@SUMEX-AIM.Stanford.EDU>
I do not have time to write full minutes from our meeting last Tuesday but do
want to record the items we decided on so Jim Ball can finish his decision
processes and get started with implementation.
1) Backup/archive storage media - we need to alleviate the operational and
medium inefficiencies of dumping large disk stores to industry tape. Based on
preliminary investigations, we authorized CSD/CF to decide on and purchase two
high-density, helical-scan tape storage units for installation on suitable
network servers and to experiment with their use as an alternative to industry
tape. The expected cost of these units is $5-6K each.
2) Network file storage capacity - we need to alleviate access and capacity
problems on existing network file servers (Jeeves in particular). Based on
information presented, we authorized CSD/CF to investigate and purchase an
additional server. A number of vendors have been under consideration,
including (Apollo, IBM, SUN, and DEC) with discount packages that might include
a number of additional workstations for departmental use. The committee
recommended that unless there were an exceptional overriding reason to the
contrary, that the new server be selected from a vendor for which CSD/CF
systems staffing, experience, and support is already in place to minimize
secondary costs of this purchase. The expected hardware cost is expected to be
less than $75K.
The cost of amortizing these purchases (presumably over 5 years) will be borne
through appropriate CSD-CF rate increases and will hopefully be offset somewhat
by reduced tape medium costs and increased staff efficiency.
In other agenda items, the committee reviewed CSD/CF operations for the past
year. These continue to be well managed technically and financially. Some
older systems (SAIL and SCORE in particular) are becoming less and less cost
effective resources. A fuller summary will be sent when I get back from my
trip in 3 weeks.
Please forward any corrections or comments. Tom R.
-------
∂07-Oct-88 0911 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSD-CF Rates for 1988-89
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Oct 88 09:11:40 PDT
Received: from polya.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Fri 7 Oct 88 09:07:46-PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA05715; Fri, 7 Oct 88 09:09:01 PDT
Date: Fri, 7 Oct 88 09:09:01 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
Message-Id: <8810071609.AA05715@polya.Stanford.EDU>
To: Faculty@Score
Cc: Facil@Score
Subject: CSD-CF Rates for 1988-89
The following rates have been sent to the controller's office for
approval. Approval is expected and only minor adjustments may
have to be made. I am sending them to you for planning purposes
only. As soon as the approved rates are available they will be
distributed.
One major change is the job-time or connect time charge. Connect
time was charged at $1.00 per hour for A time, it is now .10 per
hour. Maintaining the connect time charge, even at a low level
helps keep us focused on the fact that much of our attention, and
time goes into the CSD network. Perhaps we can find a better way
of measuring and charging for this resource in the future. CPU and
Disk storage charges were adjusted to compensate for this change
since they represent the real system resources being utilized.
Another change is the fact that printing charges had to be increased
from .09 a page to .10 a page for laser printers and the Dover. Boise
charges were increased from .07 to .09 per page.
-Jim Ball
COMPUTER SCIENCE DEPARTMENT COMPUTER FACILITIES
PROPOSED SERVICE CENTER RATES
9/1/88
TIME WEEKDAY WEEKEND
--------- --------- ---------
"A" RATES = 100% 0000-0800 C C
"B" RATES = 66.67% * "A" RATE 0800-1300 A C
"C" RATES = 33.33% * "A" RATE 1300-1800 A B
1800-2400 B C
--------TIME OF DAY RATES---------
"A" "B" "C" Monthly
---- ---- ---- -----------
Score
Account charge Accts 5.00
Connect time 0.10 0.07 0.03
CPU time Min. 4.25 2.83 1.42
Disk space Mbits/Mo 2.85
Sail
Account charge Accts 5.00
Connect time 0.10 0.07 0.03
CPU time Min. 5.50 3.67 1.83
Disk space Mbits/Mo 2.75
Labrea
Account charge Accts 5.00
Connect time 0.10 0.67 0.33
CPU time Min. 1.25 1.00 0.50
Disk space Mbits/Mo 1.35
Jeeves
Account charge Accts 5.00
Disk space Mbits/Mo 1.08
Polya
Account charge Accts 5.00
Connect time 0.10 0.67 0.33
CPU time Min. 3.00 0.50 0.25
Disk space Mbits/Mo 2.85
Printers
Dovers pages 0.10
Imagen/Apple pages 0.10
Boise pages 0.09
Phototypesetter
Page charges 4.50
Ethernet Maintenance
Monthly charges
Workstations & Minis 33.00
Mainframes & Bridges 330.00
VAX-750 Computer Maint.
Monthly charges
Basic VAX-750 475.00
RA81 Disk Drive 100.00
Kennedy 9300 Tape 200.00
Fujitsu M2351 Disk 50.00
TU78 100.00
CDC 9766 Controller 100.00
Emulex SC758 66.00
8 line term MUX 16.00
∂07-Oct-88 0929 golub@na-net.stanford.edu FAX
Received: from Bravery.stanford.edu by SAIL.Stanford.EDU with TCP; 7 Oct 88 09:28:58 PDT
Received: by Bravery.stanford.edu (4.0/inc-1.5)
id AA00941; Fri, 7 Oct 88 09:34:24 PDT
Date: Fri, 7 Oct 88 09:34:24 PDT
From: golub@na-net.stanford.edu (Gene Golub)
Message-Id: <8810071634.AA00941@Bravery.stanford.edu>
To: nilsson@score.stanford.edu
Subject: FAX
Nils,
Has our FAX machine been installed? I'd like to send some mail abroad.
Gene
∂07-Oct-88 0951 @Score.Stanford.EDU:ARK@SAIL.Stanford.EDU Re: CSD-CF Rates for 1988-89
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Oct 88 09:51:31 PDT
Received: from SAIL.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Fri 7 Oct 88 09:48:34-PDT
Message-ID: <1qnXYk@SAIL.Stanford.EDU>
Date: 07 Oct 88 0950 PDT
From: Arthur Keller <ARK@SAIL.Stanford.EDU>
Subject: Re: CSD-CF Rates for 1988-89
To: ball@POLYA.STANFORD.EDU
CC: Facil@SCORE.Stanford.EDU
Jim,
Overall, rates look reasonable, as the lowering of the connect charges has
not unreasonably raised CPU and disk charges. However, some of the computation
puzzles me. The computation of the "B" and "C" rates for Labrea and for
Polya don't appear to correspond to the "A" rates. Also, while a 10 cent
per hour connect time does make sense for all the other computers, I'm not
convinced it makes sense for Labrea, since it is a file server, and telnet
usage of Labrea slows it down. So I think we should try (on Labrea alone)
to increase the connect and cpu charges so the Disk space charges are as low
as possible. In particular, since Jeeves is cheaper than Labrea for disk
storage, you may find more people wanting Jeeves accounts to store files there!
I think it would be great if Labrea disk storage charges could be cheaper than
all other systems, including Jeeves. Also, if someone wants to have a SUN
connected to Jeeves, are there any other charges besides the disk and account
charges? Do you supply the SUN, and if so, do they pay for it on a monthly
basis? Are there any SUNs available? Thanks.
Arthur
Labrea
Account charge Accts 5.00
Connect time 0.10 0.67 0.33
CPU time Min. 1.25 1.00 0.50
Disk space Mbits/Mo 1.35
Jeeves
Account charge Accts 5.00
Disk space Mbits/Mo 1.08
Polya
Account charge Accts 5.00
Connect time 0.10 0.67 0.33
CPU time Min. 3.00 0.50 0.25
Disk space Mbits/Mo 2.85
∂07-Oct-88 1152 VAL exchange with Moscow
Mints has apparently told my mother that there is no way I can be allowed to
visit with her after the Tallinn conference. (I presume the problem is that
their bureaucracy can't include a personal visit in the itinerary of a trip
like this; I hope they will let me at least to buy a tour to Leningrad.) In
any case, that signigicantly increases my interest in the possibility of
lecturing at the Institute of Philosophy, because my mother will be able to
spend a few days in Moscow.
Also, I haven't heard from Suppes since I sent him the message about the
replies from the potential invitees. If you have a chance, please ask him
whether he's doing anything about that - I'd like to reply to those people
and tell them something positive.
∂07-Oct-88 1237 VAL next week
I'll spend the next week in Torino, Italy, where I'll be giving a talk on
"Negation as failure and introspective reasoning" at the International
Symposium on Methodologies for Intelligent Systems. (The organizers pay all
expenses.) I have the program, if you are interested.
∂07-Oct-88 1246 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSD-CF Rates for 1988-89
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Oct 88 12:46:53 PDT
Received: from polya.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Fri 7 Oct 88 12:35:39-PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA20583; Fri, 7 Oct 88 12:36:52 PDT
Date: Fri, 7 Oct 88 12:36:52 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
Message-Id: <8810071936.AA20583@polya.Stanford.EDU>
To: ARK@SAIL.Stanford.EDU
Cc: Facil@SCORE.Stanford.EDU
In-Reply-To: Arthur Keller's message of 07 Oct 88 0950 PDT <1qnXYk@SAIL.Stanford.EDU>
Subject: CSD-CF Rates for 1988-89
Arthur,
Yep, looks like I have a couple of typos in there I didn't catch. Thanks for
proofreading it for me. I have to see if the stuff I sent over to the
controllers office has the same bug.
-Jim
∂07-Oct-88 1302 @Score.Stanford.EDU:ball@polya.Stanford.EDU [ball@polya.Stanford.EDU: CSD-CF Rates for 1988-89]
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Oct 88 13:01:55 PDT
Received: from polya.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Fri 7 Oct 88 12:58:09-PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA22040; Fri, 7 Oct 88 12:59:24 PDT
Date: Fri, 7 Oct 88 12:59:24 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
Message-Id: <8810071959.AA22040@polya.Stanford.EDU>
To: Faculty@Score
Cc: Facil@Score
Subject: [ball@polya.Stanford.EDU: CSD-CF Rates for 1988-89]
A couple of typos were found in the attached memo. The rates for B
& C time on both Labrea and Polya were incorrect.
Sorry for the inconvenience.
-Jim Ball
Return-Path: <@Score.Stanford.EDU:ball@polya.Stanford.EDU>
Date: Fri, 7 Oct 88 09:09:01 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
To: Faculty@Score
Cc: Facil@Score
Subject: CSD-CF Rates for 1988-89
The following rates have been sent to the controller's office for
approval. Approval is expected and only minor adjustments may
have to be made. I am sending them to you for planning purposes
only. As soon as the approved rates are available they will be
distributed.
One major change is the job-time or connect time charge. Connect
time was charged at $1.00 per hour for A time, it is now .10 per
hour. Maintaining the connect time charge, even at a low level
helps keep us focused on the fact that much of our attention, and
time goes into the CSD network. Perhaps we can find a better way
of measuring and charging for this resource in the future. CPU and
Disk storage charges were adjusted to compensate for this change
since they represent the real system resources being utilized.
Another change is the fact that printing charges had to be increased
from .09 a page to .10 a page for laser printers and the Dover. Boise
charges were increased from .07 to .09 per page.
-Jim Ball
COMPUTER SCIENCE DEPARTMENT COMPUTER FACILITIES
PROPOSED SERVICE CENTER RATES
9/1/88
TIME WEEKDAY WEEKEND
--------- --------- ---------
"A" RATES = 100% 0000-0800 C C
"B" RATES = 66.67% * "A" RATE 0800-1300 A C
"C" RATES = 33.33% * "A" RATE 1300-1800 A B
1800-2400 B C
--------TIME OF DAY RATES---------
"A" "B" "C" Monthly
---- ---- ---- -----------
Score
Account charge Accts 5.00
Connect time 0.10 0.07 0.03
CPU time Min. 4.25 2.83 1.42
Disk space Mbits/Mo 2.85
Sail
Account charge Accts 5.00
Connect time 0.10 0.07 0.03
CPU time Min. 5.50 3.67 1.83
Disk space Mbits/Mo 2.75
Labrea
Account charge Accts 5.00
Connect time 0.10 0.67 0.33
CPU time Min. 1.25 0.83 0.42
Disk space Mbits/Mo 1.35
Jeeves
Account charge Accts 5.00
Disk space Mbits/Mo 1.08
Polya
Account charge Accts 5.00
Connect time 0.10 0.67 0.33
CPU time Min. 3.00 2.00 1.00
Disk space Mbits/Mo 2.85
Printers
Dovers pages 0.10
Imagen/Apple pages 0.10
Boise pages 0.09
Phototypesetter
Page charges 4.50
Ethernet Maintenance
Monthly charges
Workstations & Minis 33.00
Mainframes & Bridges 330.00
VAX-750 Computer Maint.
Monthly charges
Basic VAX-750 475.00
RA81 Disk Drive 100.00
Kennedy 9300 Tape 200.00
Fujitsu M2351 Disk 50.00
TU78 100.00
CDC 9766 Controller 100.00
Emulex SC758 66.00
8 line term MUX 16.00
∂07-Oct-88 1324 MPS Kyoto
I Fed Ex'd the book to Japan and Fax'd the quiz answers
today.
I have to look for a place to live this weekend, so I left
at 1:30.
Pat
∂07-Oct-88 1654 VAL Nonmonotonic seminar - no meeting next week
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
There will be no meeting of the nonmonotonic seminar next Friday, October 14.
-- Vladimir Lifschitz
∂07-Oct-88 1724 nilsson@Tenaya.stanford.edu Berlin in April?
Received: from Tenaya.stanford.edu by SAIL.Stanford.EDU with TCP; 7 Oct 88 17:24:28 PDT
Received: by Tenaya.stanford.edu (4.0/SMI-DDN)
id AA14877; Fri, 7 Oct 88 17:14:52 PDT
Date: Fri, 7 Oct 88 17:14:52 PDT
From: nilsson@Tenaya.stanford.edu (Nils Nilsson)
Message-Id: <8810080014.AA14877@Tenaya.stanford.edu>
To: cheriton@polya, Feigenbaum@sumex, Genesereth@sumex, Goldberg@polya,
Golub@polya, Guibas@polya, Gupta@polya, jlh@sonoma, dek@sail,
latombe@coyote, Linton@polya, zm@sail, jmc@sail, mitchell@polya,
oliger@polya, shoham@polya, weise@polya, wiederhold@polya,
winograd@polya
Cc: et.kck@forsythe, tajnai@score, fullerton@sierra
Subject: Berlin in April?
Does anyone out there want to go to Berlin in April of 1989?
(All expenses paid plus, possibly, an honorarium.)
A German Company named PSI wants to invite 2 or 3 CS faculty members
to come and give talks about the future of CS at a major event that
PSI is having. Details follow. If interested and/or if you have
questions, contact Ken Kaufman (Overseas Studies, x5-4345,
et.kck@forsythe) who is coordinating. [Note to Carolyn Tajnai:
PSI ought to be contacted about joining the Forum.] -Nils
---------
PSI is a software company in Berlin, basically employee owned, which
has been very successful. On April 21, 1989, it will be 20 years old.
The company has very generously supported our German programs, the
base budget at the Berlin Center, The German Consortium which places
Stanford students with German firms in long-term jobs, but most of
all, the Krupp Internship Program. Stanford engineers are well known
at PSI.
On April 21, there will be a major program at the Congress-Halle in
Berlin. Significant Germans and Europeans will be in attendance.
Here is a letter from PSI to Kaufman describing the event:
------
Dear Mr. Kaufman,
This letter summarizes for you and your colleagues our basic
thoughts, ideas, and guidelines for the festivities for the 20th
anniversary of PSI.
1. Background
a) Software is a very recent business activity in Germany. 20
years in the business make PSI an 'old hand' who may
justifiably celebrate this anniversary in style.
b) In Germany PSI leads the market for automation projects for
factories (CIM)
energy supply
PSI not only supplies advanced concepts but all the
relevant software as well. PSI is determined to strengthen
its own position internationally in information technology.
c) The whole event with all its individual components is
geared to promote
Public Relations,
i.e. PSI wants to publicize its name, present itself as a
technically competent and forward-looking partner, and
motivate established clients to further and closer
cooperation.
d) Participants will come from
o medium to higher management of our clients
o government departments
o the Press
o PSI management and selected staff
2. Procedure
The following individual events will be arranged
Thursday (20.04.89)
Afternoon: Press Conference
(A PSI Award or a PSI
Foundation may be announced)
Evening: Reception for our clients and
the Press at the Berlin Senate
Friday (21.04.89)
All day: Symposium
Evening: Large party for clients, the
Press, an PSI staff and their
partners presenting a cross
section of Berlin's cabaret and
music scene.
3. Symposium
a) The venue will be the roof garden of the ICC and we expect
to have 300 guests attending.
10.00h Start
12.00h Lunch
14.00h Start of Part 2
17.00h End
b) We want to use this symposium to publicize our connection
with Stanford University and bask a little in the glamour
of your University and utilize for PSI the reputation and
and competence of your lecturers.
c) We would be very happy if Stanford would deliver lectures
in the time from
10.00 a.m. to 15.30 p.m.
For the last lecture we intend to engage a German or
European philosopher or writer who will approach the
central issue very generally and from a totally different
point of view.
d) So far we have chosen the following title for the central
issue (but would like to hear your ideas on this!):
"Chances and Risks for Information Technology and
Automation over the next 20 Years (or in the year 2000)
with a special outlook on Software."
We wanted the central issue to be fairly far reaching since
we intend not only to cover the technical but also the
economic and socio-political aspects.
e) Because of the large attendance it is impossible to
organize individual working groups. Therefore, the time is
available for lectures only. It would be feasible,
however, that all Stanford lecturers arrange a panel
discussion at the end and also encourage questions from
the audience.
f) Altogether there are 4 hours available to Stanford, i.e. 3
to 4 lecturers will have a chance to speak.
I trust that I have given you and your colleagues some background
information and PSI's ideas on the joint symposium. The selection
of the lecturers and the choice of the exact subjects will, of
course, be up to Stanford. I would be happy to discuss details
either here in Berlin or at Stanford at your convenience.
Please drop me a line to say whether you agree to the basic outline
of the concept.
With best regards,
yours,
Dietrich Jaeschke
∂07-Oct-88 2242 @RELAY.CS.NET:masahiko@nuesun.ntt.jp Re: coming to Sendai
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 7 Oct 88 22:42:31 PDT
Received: from relay2.cs.net by RELAY.CS.NET id aa08011; 8 Oct 88 1:05 EDT
Received: from ntt.jp by RELAY.CS.NET id ab07791; 8 Oct 88 0:48 EDT
Received: by ntt-sh.ntt.jp (3.2/ntt-sh-02) with TCP; Sat, 8 Oct 88 13:33:43 JST
Received: by MECL.NTT.jp (3.2/NTTcs02) with TCP; Sat, 8 Oct 88 13:32:11 JST
Date: Sat, 8 Oct 88 13:32:11 JST
From: Masahiko Sato <masahiko@nuesun.ntt.jp>
Message-Id: <8810080432.AA06058@MECL.NTT.jp>
To: JMC@SAIL.STANFORD.EDU
Cc: clt@SAIL.STANFORD.EDU
In-Reply-To: John McCarthy's message of 23 Sep 88 1458 PDT <cgsp6@SAIL.Stanford.EDU>
Subject: Re: coming to Sendai
I was visiting Chalmers University in Sweden for one month and just
came back. Sorry for the delay of my reply.
As I told you in my previous mail, I will not be in Sendai from
November 7 (Mon) attending an international conference on logic in
Nagoya. Professor Ito will be in Sendai at that time, however, and he says
that he will welcome you and Carolyn. So, if you will be able to come
to Sendai, please let me know the titles of your talk and Carolyn's.
It will be nice if you could provide short abstracts as well. I
might be able to see you in Sendai on Suday, November 6. If you need
any help with regard to the reservations of accomdations in Sendai
and reservations of train tickets etc., please let me know.
I will see you in Kyoto anyway. I look forward to seeing you soon.
masahiko
∂08-Oct-88 1206 ARK re: CSD-CF Rates for 1988-89
John, for your protest to the Controller's office, below are the CSD-CF
rates for SAIL effective 9/1/87 and 4/1/88 and the proposed rates
effective 9/1/88.
Arthur
Effective September 1, 1987
Sail Computer
(DEC10)/WAITS)
Account chg per mo @ 5.00
Connect time hours @ 1.00 0.67 0.33
CPU time Minutes @ 3.98 2.65 1.32
Disk space Mbits/Mo @ 3.75
04/01/88
Sail
Account charge Accts @ 5.00
Connect time hours @ 1.00 0.67 0.33
CPU time Minutes @ 4.00 2.67 1.33
Disk space Mbits/Mo @ 3.00
9/1/88
Sail
Account charge Accts 5.00
Connect time 0.10 0.07 0.03
CPU time Min. 5.50 3.67 1.83
Disk space Mbits/Mo 2.75
∂08-Oct-88 1707 ARK re: CSD-CF Rates for 1988-89
∂08-Oct-88 1539 JMC re: CSD-CF Rates for 1988-89
[In reply to message rcvd 08-Oct-88 12:06-PT.]
Then I don't understand why the message referred to an increase in
disk rates. Or did I misread it?
ARK - Disk rates went up on Polya, for example. Generally CPU charges
went up and disk charges varied. Look on CFRATE.TXT[1,ARK] for a
retrospective of the last few CF rate sheets.
Arthur
∂09-Oct-88 1420 daniel@mojave.Stanford.EDU disk use charges
Received: from mojave.Stanford.EDU by SAIL.Stanford.EDU with TCP; 9 Oct 88 14:20:38 PDT
Received: by mojave.Stanford.EDU (5.59/inc-1.0)
id AA24917; Sun, 9 Oct 88 14:19:57 PDT
Date: Sun, 9 Oct 88 14:19:57 PDT
From: daniel@mojave.Stanford.EDU (Daniel Weise)
Message-Id: <8810092119.AA24917@mojave.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 09 Oct 88 1357 PDT <Ooxvd@SAIL.Stanford.EDU>
Subject: disk use charges
1. The charges are way out of line with costs for disk file -
by a factor larger than 10 and perhaps larger than 100. SAIL's
present (somewhat reduced) charges of $2.75 per megabit month
would pay for RAM chips to store the files in two months and for
buying disk units in one month.
I think your math is off here:
$2.75 MBit/Month = $22.00 MByte/Month.
A MByte of memory runs around $400 for micros, double that
for Suns and mainframes. So it would take 18 months to buy RAM.
Also, to store 300MByte on disk at CSDCF rates would be
$6600 per month. Since I can buy a 330MByte disk from HP
for $7000, it seems that one could buy a disk in 1 month, not
two.
I have the breakdown of CSDCF costs. It shows 2/3's of its
costs come from salaries. For example, they list networking
costs at 46K year, of which 30K is salary. There's no breakdown
of how they came to this figure.
Daniel
∂09-Oct-88 1540 ARK Re: disk use charges
David Cheriton is ultimately right about "dynosaur" shared computing.
The solution is to get your own workstations and fileserver and to
pay for the service from CSD-CF that you really need. For me, that
will be software support and file backup and restoration. I plan to
get a collection of SUN workstations before SAIL's demise.
Arthur
∂10-Oct-88 0530 Rich.Thomason@b.gp.cs.cmu.edu JPL Paper
Received: from B.GP.CS.CMU.EDU by SAIL.Stanford.EDU with TCP; 10 Oct 88 05:30:05 PDT
Date: Sun, 9 Oct 1988 20:21:27 EDT
From: Rich Thomason <thomason@B.GP.CS.CMU.EDU>
To: jmc@sail.stanford.edu
Cc: thomason
Subject: JPL Paper
Message-ID: <CMM.0.88.592446087.thomason@B.GP.CS.CMU.EDU>
John,
I have been in touch with the publisher. They have to go ahead
with the JPL issue. However, they are publishing a book in parallel with
the issue, and can add extra papers to the book.
I would very much like to include your paper with the others
in the book. If you could send me electronic copy by 10 December I
could see that it is published in a later issue of the Journal of
Philosophical Logic (August 89 at latest) and included in the book,
which will appear more or less instantly.
It would help if I could have generically Tex formatted copy.
(That just means, so I can tell what the formulas look like.) I would
then format it here and print it so that it looks more or less like a
JPL article. Without camera ready copy of this sort, we could not make
the deadline for the book. If you get it to me by 10 December, I can
send you hard copy to proofread.
Do you think this is feasable? Or do you expect to win another
fancy time consuming prize?
--Rich
∂10-Oct-88 0900 JMC
Karl Cohen
∂10-Oct-88 0913 WIEDERHOLD@SUMEX-AIM.Stanford.EDU Re: disk use charges
Received: from SUMEX-AIM.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Oct 88 09:13:07 PDT
Date: Mon, 10 Oct 88 09:06:45 PDT
From: Gio Wiederhold <WIEDERHOLD@SUMEX-AIM.Stanford.EDU>
Subject: Re: disk use charges
To: JMC@sail.Stanford.EDU
cc: ullman@score.Stanford.EDU
In-Reply-To: <Ooxvd@SAIL.Stanford.EDU>
Message-ID: <12437335928.23.WIEDERHOLD@SUMEX-AIM.Stanford.EDU>
I agree of course, but then I am a database freak.
I have often moved usage away fron CSD-CF because of these policies,
and expect to continue to do so. I am afraid that is less effort than
fight battles with administrators. It does raise total cost for others.
In general, having interesting data on-line (not just textual, but also raw
stuff) motivates use of computers, because now even simple programs can
produce interesting results.
Gio
-------
∂10-Oct-88 1109 hoffman@csli.Stanford.EDU as you're not here and your door is closed
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Oct 88 11:09:11 PDT
Received: by csli.Stanford.EDU (4.0/4.7); Mon, 10 Oct 88 11:09:10 PDT
Date: Mon 10 Oct 88 11:09:09-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: as you're not here and your door is closed
To: jmc@SAIL.STANFORD.EDU
Message-Id: <592510149.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
I presume that you needed to cancel. When you have time, send me another
time to meet.
reid
-------
∂10-Oct-88 1123 bhayes@polya.Stanford.EDU Psycho Pservey
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Oct 88 11:23:43 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA28761; Mon, 10 Oct 88 11:22:15 PDT
Date: Mon, 10 Oct 88 11:22:15 PDT
From: Barry Hayes <bhayes@polya.Stanford.EDU>
Message-Id: <8810101822.AA28761@polya.Stanford.EDU>
To: jmc@sail
Subject: Psycho Pservey
John-
Once or twice I've failed to notice that someone was black, although
it's a rarity for me. I recall someone claiming on some news group
[sorry] that he or she sometimes failed to notice someone's sex. I
don't think this is quite the "blocking" you mentioned, but it does
have to do with how the mind files things. Me? I always notice
someone's sex, and find it fairly unbelievable that someone doesn't.
-Barry
∂10-Oct-88 1146 Rich.Thomason@cad.cs.cmu.edu re: JPL Paper
Received: from CAD.CS.CMU.EDU by SAIL.Stanford.EDU with TCP; 10 Oct 88 11:46:28 PDT
Date: Mon, 10 Oct 1988 14:43:56 EDT
From: Rich Thomason <thomason@CAD.CS.CMU.EDU>
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: JPL Paper
In-Reply-To: Your message of 10 Oct 88 0836 PDT
Message-ID: <CMM.0.88.592512236.thomason@CAD.CS.CMU.EDU>
John,
Great! It could even be Dec. 10.
--Rich
∂11-Oct-88 0133 @Gang-of-Four.Stanford.EDU,@IGNORANT.Stanford.EDU:rivin@GANG-OF-FOUR.Stanford.EDU meeting
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 01:33:23 PDT
Received: from Ignorant.Stanford.EDU by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01093; Tue, 11 Oct 88 01:29:07 PDT
Date: Tue, 11 Oct 88 01:29 PDT
From: Igor Rivin <rivin@GANG-OF-FOUR.Stanford.EDU>
Subject: meeting
To: qlisp@GANG-OF-FOUR.Stanford.EDU
Message-Id: <19881011082931.1.RIVIN@IGNORANT.Stanford.EDU>
The next meeting will be this wednesday at noon in MJH301.
I will give a survey of parallel garbage collection algorithms.
CU there.
∂11-Oct-88 0826 tom@polya.Stanford.EDU toner cartridges for Imagen
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 08:26:17 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA19284; Tue, 11 Oct 88 08:24:56 PDT
Date: Tue, 11 Oct 88 08:24:56 PDT
From: Tom Dienstbier <tom@polya.Stanford.EDU>
Message-Id: <8810111524.AA19284@polya.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 10 Oct 88 1544 PDT <40seG@SAIL.Stanford.EDU>
Subject: toner cartridges for Imagen
John, I'll check this out. The Imagens that we support DONOT use toner
cartridges. We actually install toner from a box. I'll let you know
what I come up with.
tom
∂11-Oct-88 1340 BALDWIN@Score.Stanford.EDU toner cartridges
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 13:40:47 PDT
Date: Tue 11 Oct 88 13:37:57-PDT
From: Julie Baldwin <BALDWIN@Score.Stanford.EDU>
Subject: toner cartridges
To: jmc@Sail.Stanford.EDU
cc: tom@Score.Stanford.EDU
Message-ID: <12437647443.21.BALDWIN@Score.Stanford.EDU>
In response to your message to Tom, we get our toner from stores.
The stores number is 380309 and they sell for $71.46 per cartridge.
Julie
-------
∂11-Oct-88 1356 Mailer Re: disk use charges
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 11 Oct 88 13:55:58 PDT
Received: from Cabernet.ms by ArpaGateway.ms ; 11 OCT 88 13:51:46 PDT
Date: 11 Oct 88 13:51 PDT
From: hayes.pa@Xerox.COM
Subject: Re: disk use charges
In-reply-to: John McCarthy <JMC@SAIL.Stanford.EDU>'s message of 09 Oct 88
13:57 PDT
To: John McCarthy <JMC@SAIL.Stanford.EDU>
cc: faculty@SCORE.STANFORD.EDU, su-computers@SAIL.Stanford.EDU
Message-ID: <881011-135146-6304@Xerox>
Its easy for me, since I dont have any over there, but Im behind you on
this files issue. I recently logged in to SAIL to send a mail message, and
promptly got a bill for around $50 for the filespace occupied by two years
back mail, which I was obliged to flush.
There is a general tendency for administrative systems to start doing
things to suit themselves rather than the people for whose convenience they
exist. If theres no good reason to overcharge, dont do it. And if there
is, lets all see it stated in clear nonlegalistic prose.
Pat Hayes
∂11-Oct-88 1401 BALDWIN@Score.Stanford.EDU re: toner cartridges
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 14:00:55 PDT
Date: Tue 11 Oct 88 13:58:02-PDT
From: Julie Baldwin <BALDWIN@Score.Stanford.EDU>
Subject: re: toner cartridges
To: JMC@Sail.Stanford.EDU
cc: tom@Score.Stanford.EDU, BALDWIN@Score.Stanford.EDU
In-Reply-To: <1x0xBv@SAIL.Stanford.EDU>
Message-ID: <12437651099.21.BALDWIN@Score.Stanford.EDU>
We have never had any problems getting them.
-------
∂11-Oct-88 1501 RC.STA@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 11 Oct 88 15:01:27 PDT
Date: Tue, 11 Oct 88 15:00:44 PDT
To: jmc@sail
From: "Stacey Green" <RC.STA@Forsythe.Stanford.EDU>
To: Professor McCarthy
From: Stacey Green (rc.sta)
Re: November presentation to Kankeiren
Thank you for your e-mail of 10/4 stating that you have arranged to
stay in Japan an extra day. At this point, our staff member in Kyoto
has been given permission for you to speak on November 15th. Though
I don't have the exact times yet, I am pretty sure it will be
between noon and 3:00, as I imagine you would have to leave for the
airport at around 3:00. I'm not sure who has arranged your return
ticket (perhaps you have?) but I was asked to make sure that you
remind the Inamori Foundation that you will be returning from Osaka
(not Tokyo) on a United flight (Fl #810?).
I will let you know of the time, format, and location as soon as I
know. Thank you very much for agreeing to speak. Professor Aoki
and others involved in the Kyoto project send their regards.
Stacey Green
∂11-Oct-88 1540 barwise@russell.Stanford.EDU Reply
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 15:40:33 PDT
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Tue, 11 Oct 88 15:41:22 PDT
To: jmc@sail
Subject: Reply
Date: Tue, 11 Oct 88 15:41:20 PDT
From: Jon Barwise <barwise@russell.Stanford.EDU>
John, I got an contribution for my "Computers and Math" column (NOtice
of the AMS) from Gian-Carlo Rota and Paul Stein called
"Mathematics and AI" I know you will disagree with it. I wonder
if you would like to write a reply to it? I think it would be needed
in late November. The piece is 16 pages and your reply could be
4 or 5 pages. Jon
∂11-Oct-88 1546 barwise@russell.Stanford.EDU Re: Reply
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 15:46:14 PDT
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Tue, 11 Oct 88 15:46:59 PDT
To: JMC@SAIL.Stanford.EDU
Subject: Re: Reply
In-Reply-To: Your message of 11 Oct 88 15:42:00 PDT.
<1T0yci@SAIL.Stanford.EDU>
Address: CSLI, Stanford University, Stanford, CA 94305 (415) 723-0110
Date: Tue, 11 Oct 88 15:46:54 PDT
From: Jon Barwise <barwise@russell.Stanford.EDU>
OK, it is on its way.
∂11-Oct-88 1600 minker@jacksun.cs.umd.edu RITA G. MINKER
Received: from gyre.umd.edu by SAIL.Stanford.EDU with TCP; 11 Oct 88 16:00:09 PDT
Received: from jacksun.cs.umd.edu by gyre.umd.edu (5.58/4.7)
id AA10065; Tue, 11 Oct 88 18:57:23 EDT
Received: by jacksun.cs.umd.edu (5.54/3.14)
id AA04814; Tue, 11 Oct 88 19:00:13 EDT
Date: Tue, 11 Oct 88 19:00:13 EDT
From: minker@jacksun.cs.umd.edu (Jack Minker)
Return-Path: <minker@jacksun.cs.umd.edu>
Message-Id: <8810112300.AA04814@jacksun.cs.umd.edu>
To: JMC@sail.stanford.edu
Subject: RITA G. MINKER
Rita G. Minker
April 28, 1927 - October 11, 1988
Rita G. Minker, early worker in the field of computer
programming, died on October 11, 1988 of cancer at the age
of 61. Mrs. Minker received a B.S. degree with High Honors
in Mathematics from Douglass College in 1948 and a M.A.
degree in Mathematics from the University of Wisconsin in
1950.
In the summer of 1950 Mrs. Minker started to work at
the prestigious Bell Telephone Laboratories in Murray Hill,
New Jersey. She programmed network problems for one of the
early digital computers, the Bell Relay Machine. She was
among the first computer programmers in the United States.
On June 24, 1951 she married Jack Minker, who she had met at
the University of Wisconsin. The couple moved to Buffalo,
New York, where Mrs. Minker was employed as a mathematician
at the Cornell Aeronautical Laboratories. She worked on
electronic analog computers on which she simulated the per-
formance of missile systems. In 1952 she was hired by RCA
in Camden, New Jersey and became the second computer pro-
grammer, and the first woman programmer to work at that com-
pany. She programmed the Bizmac, RCA's first computer.
In 1953 Mrs. Minker took time off from the computing
profession to raise a family. In April 1964, when her two
children were enrolled in school, Mrs. Minker returned to
work as a mathematician and computer programmer in the newly
formed Division of Computer Research and Technology (DCRT)
at the National Institutes of Health (NIH)in Bethesda, Mary-
land. She was one of the charter members of this division,
formed to service the computer needs of medical researchers
at the NIH. Although the computing profession had made sig-
nificant progress during the time she was raising her fam-
ily, Mrs. Minker was able to rapidly learn the new technol-
ogy and recapture her skills as a programmer and mathemati-
cian. She served as Head, Training Unit in DCRT from 1968 -
1975, and instituted training courses to permit medical
researchers to learn how to program and work with computers
and become familiar with statistical methods. In 1975,
after having built-up the Training Division, she joined the
Statistical Software Section, Laboratory of Statistical and
Mathematical Methodology of the DCRT. She was able to par-
ticipate and assist medical researchers with their program-
ming and statistics problems. She was also in charge of
consulting on and maintaining SPSS, a major statistical
package.
Mrs. Minker was co-author of a number of medical jour-
nal articles on the schistosomyacin disease and was
acknowledged for her assistance in numerous medical journal
articles. Together with her husband, she published an arti-
cle in the Annals of the History of Computing, which traced
the historical developments in the optimization of boolean
expressions and related problems.
Mrs. Minker had a long bout with cancer. She first
contracted breast cancer in 1975. The disease reoccurred in
1985. Because of her illness she was forced to retire from
the government in April 1988, exactly 24 years after she was
hired at the NIH. Mrs. Minker is survived by her son,
Michael Saul Minker who resides with his wife Katharine in
Chevy Chase, Maryland; by her daughter, Sally Anne Minker,
who lives in Bethesda, Maryland; by her husband, Jack
Minker, of Bethesda, Maryland; her father, Louis H. Goldberg
and step-mother, Anna Goldberg of North Miami Beach,
Florida; and brother, Sanford H. Goldberg of West Orange,
New Jersey.
Funeral services will be held at:
10:00AM Thursday, October 13, 1988
Congregation Beth El
8215 Old Georgetown Road
Bethesda, Maryland
Contributions may be made to the American Cancer Society in
her memory.
The family will receive condolence calls during the
period October 13, 1988 through the evening of October 18,
1988 at the home of:
Jack Minker
6913 Millwood Road
Bethesda, Maryland 20817
∂11-Oct-88 1609 MPS classes
CS 323 - Advanced Artificial Intelligence
2:45 - 4:00 - Skilling Aud
VTSS - 1:15 - 2:30 Tues-Thu
Bldg 60 Room 62P
∂11-Oct-88 2029 SHOHAM@Score.Stanford.EDU Re: your computer
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Oct 88 20:29:46 PDT
Date: Tue 11 Oct 88 16:45:42-PDT
From: Yoav Shoham <SHOHAM@Score.Stanford.EDU>
Subject: Re: your computer
To: JMC@Sail.Stanford.EDU
In-Reply-To: <1D0yrY@SAIL.Stanford.EDU>
Message-ID: <12437681622.25.SHOHAM@Score.Stanford.EDU>
Either way - I guess if I come get it it will be an excuse to chat.
Yoav
-------
∂12-Oct-88 1006 KARP@SUMEX-AIM.Stanford.EDU re: JMC's campaign analysis
Received: from SUMEX-AIM.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Oct 88 10:06:53 PDT
Date: Wed, 12 Oct 88 10:00:29 PDT
From: Peter Karp <KARP@SUMEX-AIM.Stanford.EDU>
Subject: re: JMC's campaign analysis
To: JMC@SAIL.Stanford.EDU
In-Reply-To: <D18ZK@SAIL.Stanford.EDU>
Message-ID: <12437869996.36.KARP@SUMEX-AIM.Stanford.EDU>
Interesting comments; I would agree. I hope we're not in for any big
surprises if he's elected.
-------
∂12-Oct-88 1011 KARP@SUMEX-AIM.Stanford.EDU bias
Received: from SUMEX-AIM.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Oct 88 10:11:40 PDT
Date: Wed, 12 Oct 88 10:05:15 PDT
From: Peter Karp <KARP@SUMEX-AIM.Stanford.EDU>
Subject: bias
To: jmc@sail.Stanford.EDU
Message-ID: <12437870865.36.KARP@SUMEX-AIM.Stanford.EDU>
Tell me, is my liberal bias showing through with my comments about
Bush's mud-slinging, or would you concur that he's been running a
sleazier campaign than Dukakis?
Peter
-------
∂12-Oct-88 1539 @Score.Stanford.EDU:WIEDERHOLD@SUMEX-AIM.Stanford.EDU Re: disk use charges
Received: from SUMEX-AIM.Stanford.EDU by SAIL.Stanford.EDU with TCP; 12 Oct 88 15:38:28 PDT
Date: Wed, 12 Oct 88 15:31:26 PDT
From: Gio Wiederhold <WIEDERHOLD@SUMEX-AIM.Stanford.EDU>
Subject: Re: disk use charges
To: hayes.pa@Xerox.COM
cc: JMC@SAIL.Stanford.EDU, faculty@score.Stanford.EDU,
su-computers@SAIL.Stanford.EDU
In-Reply-To: <881011-135146-6304@Xerox>
Message-ID: <12437930245.85.WIEDERHOLD@SUMEX-AIM.Stanford.EDU>
The lower disk charges, as set intially on Polya, caused me to invest time
to move files there. Now they are similar to all systems. Rather than fight
administrators I'll equip my workstations with adequate storage. That does
put the administrative burden on me and my students, but I can plan at least
ahead. Gio.
-------
∂12-Oct-88 2236 @cunyvm.cuny.edu:YANG@NORUNIT.BITNET Your travel expenses, IFIP China.
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 12 Oct 88 22:36:20 PDT
Received: from cunyvm.cuny.edu by RELAY.CS.NET id aa12671; 13 Oct 88 1:17 EDT
Received: from NORUNIT.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.1) with BSMTP id 0232; Wed, 12 Oct 88 10:15:54 EDT
Date: Wed, 12 Oct 88 13:04:43 ECT
To: jmc%sail.stanford.edu@RELAY.CS.NET
From: YANG%NORUNIT.BITNET@cunyvm.cuny.edu
Comment: CROSSNET mail via SMTP@INTERBIT
Subject: Your travel expenses, IFIP China.
Dear Prof. John McCarthy,
We have received your invoice for the China trip, sent to Prof.
Arne Solvberg. Solvberg is on his Sabbatical year at the Naval
Postgraduate School, Monterey (Aug. '88 - Aug. '89).
However, we would like to have copies of the tickets you mentioned
in the invoice. I am very sorry for any inconvenience it may cause
you, but the economic section we use for book-keep for
this conference has very strict rules for refunding
travel expenses, and they insist to have at least copies of the tickets.
If you like, you may send the copies through the fax,
to the number:
+47-4-594466. Attn.: Jianhua Yang.
Sincerely yours,
Jianhua Yang
(for the organization committee)
∂12-Oct-88 2350 Mailer Re: disk use charges
Received: from Pescadero (Pescadero.Stanford.EDU) by SAIL.Stanford.EDU with TCP; 12 Oct 88 23:50:33 PDT
Received: by Pescadero (5.54/Ultrix2.0-B)
id AA21303; Wed, 12 Oct 88 23:51:04 PDT
Date: Wed, 12 Oct 88 23:51:04 PDT
From: "David Cheriton" <cheriton@pescadero.stanford.edu>
Message-Id: <8810130651.AA21303@Pescadero>
To: WIEDERHOLD@sumex-aim.Stanford.EDU, hayes.pa@xerox.com
Subject: Re: disk use charges
Cc: JMC@sail.Stanford.EDU, faculty@score.Stanford.EDU,
su-computers@sail.Stanford.EDU
In-Reply-To: <12437930245.85.WIEDERHOLD@SUMEX-AIM.Stanford.EDU> from Gio
Wiederhold <WIEDERHOLD@SUMEX-AIM> on Wed, 12 Oct 88 15:31:26 PDT
I find some of this disk use charge discussion strange.
For one, it is true that storage is cheaper now than it has ever been.
I have an optical disk with 2 Gbytes that would not only allow JMC to
keep things on line but, as a WORM, precludes ever deleting. $300 per
2 gigabyte platter plus a jukebox to keep them on-line. However,..
Most of CSD-CF budget is people, and that's the big cost. So, when Gio
talks about taking the "adminstrative burden" on himself and students,
either he can manage systems in less time thatn CSD-CF or else his time
is cheaper to him, or ??
I see two major problems.
For one, we (the computing community) have done a poor job to date of
automating backup/archive, etc. procedures and its not even clear that the
derivative is positive. I understand that Sail and Tops-20 are better than
Unix in this regard (not saying much).
We are really in the dark ages when it comes to self-maintaining storage
systems - so CSD-CF incurs significant person time to provide a reliable
storage system - the alteratnive right now is unreliable service.
Secondly, we are running a enormous range of systems, from Sail to
score to Vax/Unix systems to SUN file servers, all of which are different
and brain-damaged in ways that will make our grandchildren shake their
heads in wonder. We pay to survive in this menegarie.
David C.
∂13-Oct-88 0841 C.COLE@MACBETH.STANFORD.EDU Bush & Advice: One Anecdote
Received: from MACBETH.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 08:40:56 PDT
Date: Thu 13 Oct 88 08:34:55-PDT
From: George Cole <C.COLE@MACBETH.STANFORD.EDU>
Subject: Bush & Advice: One Anecdote
To: jmc@SAIL.STANFORD.EDU
Message-ID: <12438116563.81.C.COLE@MACBETH.STANFORD.EDU>
My father related this to me. He had heard Bush taking a position on farm
policy which my father thought mistaken. Since my father had just retired from
a senior post in Cargill, he had a great deal of specific information on which
to base his opinion. He also is a Republican of long standing, and so knew a
close advisor to George Bush. He forwarded the material and his own evaluation
to that advisor to get to George Bush. Four days later, Bush changed his
position on that issue to one much more attuned to my father's informed choice.
I agree with his assessment: Bush can use new information to change his mind
and position on issues which are not at the focus of the political process
where the accusation of "hypocrisy" out-shouts a necessarily complex
explication of the change. That is at least a cue of how Bush might work out as
President: on the peripheral issues, he may do very well indeed.
I'm also satisfied that Bush is emotionally mature so that accusations
of "wimp" do not provoke an emotional need to prove himself. In international
as well as national politics this level of personal remoteness is very, very
desireable. Dukakis seems to be handling this fairly well, too, though
The reports that I have heard that Dukakis' advisors and thus Iron Mike
are ignoring non-Boston Democratic advisors are chilling; they suggest a lack
of institutionalized methods for countering the natural tendency to "close in"
the circle of advisors.
I'm not pleased with either candidate, and definitely displeased with
both campaigns. I'd like to have some solid "intelligence" to make a decision
on, rather than by anecdotes, innuendo, and attempts to unravel carefully-built
media promotions to uncover substance in the slather of style.
George
-------
∂13-Oct-88 0900 ullman@polya.Stanford.EDU RT consoles
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 08:57:59 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA17903; Thu, 13 Oct 88 08:56:36 PDT
Date: Thu, 13 Oct 88 08:56:36 PDT
From: Jeffrey D. Ullman <ullman@polya.Stanford.EDU>
Message-Id: <8810131556.AA17903@polya.Stanford.EDU>
To: jmc@sail, plotkin@polya.Stanford.EDU
Subject: RT consoles
Cc: mumick@polya.Stanford.EDU, phipps@polya.Stanford.EDU
If I recollect the RT that John gave to Serge had three displays,
including two megapel displays. If those are still around,
they could solve two problems.
1) Inderpal Mumick could use one of the displays for his RT.
2) Geoff Phipps' keyboard was destroyed by someone dripping
an unknown liquid onto it. We could use one keyboard as a replacement.
---jeff
∂13-Oct-88 1255 Mailer Re: disk use charges
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 13 Oct 88 12:55:20 PDT
Received: from Semillon.ms by ArpaGateway.ms ; 13 OCT 88 12:49:26 PDT
Date: 13 Oct 88 12:48 PDT
From: hayes.pa@Xerox.COM
Subject: Re: disk use charges
In-reply-to: Gio Wiederhold <WIEDERHOLD@SUMEX-AIM.Stanford.EDU>'s message
of Wed, 12 Oct 88 15:31:26 PDT
To: Gio Wiederhold <WIEDERHOLD@SUMEX-AIM.Stanford.EDU>
cc: hayes.pa@Xerox.COM, JMC@SAIL.Stanford.EDU, faculty@score.Stanford.EDU,
su-computers@SAIL.Stanford.EDU
Message-ID: <881013-124926-4526@Xerox>
Perhaps saying this is politically unwise, but the problem and your
solution seem to me to have a uniquely American flavor. Its a sort of 80s
bedtime story. The central organisation whose function is to serve the
needs of the community becomes unusable because it wants to charge the
market rate for its services, presumably feeling that to give the stuff
away to those who cant pay for it is somehow faintly immoral. Your
response is to accept its redefinition of its role as another agent in the
free market, and just find a cheaper way to look after yourself. The net
result is that those who, like yourself, are willing and able to somehow
get hard cash to buy what you want, wind up getting along OK, but those who
are unable or unwilling to get sufficiently rich, get screwed. And even
among those who win, more and more of their time gets used up in looking
for money, and cooperation and communication become more and more difficult
as the market forces encourage idiosyncracy. How many more kinds of
mutually incompatible "workstation with adequate storage" are there likely
to be on campus soon? Why is France, not the USA, the first country to
have a national computer-communication network in place?
Still, I really love it here and will remain loyal no matter who wins the
election. Really.
Pat
∂13-Oct-88 1429 pehoushe@Gang-of-Four.Stanford.EDU Is Anybody Out There?
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 14:29:36 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02656; Thu, 13 Oct 88 14:27:52 PDT
Date: Thu, 13 Oct 88 14:27:52 PDT
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8810132127.AA02656@Gang-of-Four.Stanford.EDU>
To: Qlisp@Gang-of-Four.Stanford.EDU
Subject: Is Anybody Out There?
In case anyone is interested, the NSTACK system can spawn tasks at a
sustained rate of 1.35 million every 10 seconds.
New-Qlisp can spawn 1200 or so tasks before breaking. I cannot
get any sustained spawning rates because of this.
If anyone would like to try my system, just drop me a note. Aside
from myself, the only one who has tried using it is Igor. He's used
it twice, that I know of.
-Dan
∂13-Oct-88 1920 RC.STA@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 13 Oct 88 19:19:56 PDT
Date: Thu, 13 Oct 88 17:44:19 PDT
To: jmc@sail
From: "Stacey Green" <RC.STA@Forsythe.Stanford.EDU>
To: Professor McCarthy
From: Stacey Green (rc.sta or 3-2085)
Date: 10/13/88
Re: Speech at Kankeiren
I've been informed by our staff in Kyoto that the Kankeiren would
like you to address a group of businessmen from 2:00-3:30 on
November 15. The location will be at the Kankeiren building,
located at:
Nakanoshima Center Building
2-27 Nakanoshima 6-chome
Kita-ku, Osaka 530 Japan
Should you have any questions about these arrangements after you
arrive in Japan, please contact our staff there:
Ted Eimon
Rune Shimogamo 114
1-3 Tadehara-cho, Takano
Sakyo-ku Kyoto
Tel: 075-791-4460
Fax: 075-791-4225
Ted has requested that I send him a short personal profile of you as
soon as possible. I imagine they would like this information so
that they can provide this to the audience they will be inviting.
Could you or perhaps your assistant send me this as soon as possible
by electronic mail? I am at rc.sta@forsythe.
Thank you very much for your prompt attention.
Rgds.
Stacey Green
∂13-Oct-88 1921 roach@Gang-of-Four.Stanford.EDU Re: Is Anybody Out There?
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 19:20:57 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02901; Thu, 13 Oct 88 15:13:12 PDT
Date: Thu, 13 Oct 88 15:13:12 PDT
From: Kelly Roach <roach@Gang-of-Four.Stanford.EDU>
Message-Id: <8810132213.AA02901@Gang-of-Four.Stanford.EDU>
To: Qlisp@Gang-of-Four.Stanford.EDU, pehoushe@Gang-of-Four.Stanford.EDU
Subject: Re: Is Anybody Out There?
What you need to do is to find out THE person in Lucid who is
responsible for the part of QLISP you want to alter. You then have
to nag nag nag about how your changes are not only GOOD but a priority
for Lucid. Finally, you have to hold THE person's hands for a month
or so and make repeat inquiries about progress to see that anything
gets done. I've been "out there" a few years and thats the way things
seem to get done.
Kelly
∂13-Oct-88 2014 DAMON@Score.Stanford.EDU New Charge Program
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 20:14:02 PDT
Date: Thu 13 Oct 88 20:10:34-PDT
From: Daemon Koronakos <DAMON@Score.Stanford.EDU>
Subject: New Charge Program
To: facilities@Score.Stanford.EDU
Message-ID: <12438243204.12.DAMON@Score.Stanford.EDU>
Hi folks,
I've written a new version of chg on POLYA. This version uses the
datafiles produced by my new accounting system.
The program is fully described in the man page (type "man cf_charge"
on POLYA).
Please run the program and give me comments and suggestions.
Jobtime is not included as yet (disk and cpu only, for now). I will
include this in the future.
Damon
-------
∂13-Oct-88 2207 ilan@Gang-of-Four.Stanford.EDU Cockburn
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 22:07:02 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA03918; Thu, 13 Oct 88 22:07:29 PDT
Date: Thu, 13 Oct 88 22:07:29 PDT
From: Ilan Vardi <ilan@Gang-of-Four.Stanford.EDU>
Message-Id: <8810140507.AA03918@Gang-of-Four.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
Cc: ilan@score
In-Reply-To: John McCarthy's message of 13 Oct 88 2146 PDT <r1c8Y@SAIL.Stanford.EDU>
Subject: Cockburn
Yes, the theme is that chess players are a psychopathic repressed
bunch of guys (he quotes heavily from Reuben Fine's terrible book
``The Psychology of the Chessplayer'') and he goes on with the
canonical examples: Steinitz challenging God to Pawn and move etc.,
i.e., stuff that you can read in any Time Magazine article in chess.
By the way, the book came out about the same time as Jim Fixx'
``Puzzles for the superintelligent'' which he wrote when he
still weighed 250 lbs. I was thinking of writing a letter to
Marylin Mach vos Savant, who is listed in the Guiness book as
having the highest IQ in the world. I have her book ``The omni IQ
quiz book.'' It contains a number of sequences you are supposed to
complete, e.g.
v1,2,2,5,2,2,6,?
that you need a large brain to figure out. I was going to send her
R.K. Guy's article in the last Monthly about ``The strong law of
small numbers.'' Also, I was going to point out to her that
what she is really looking for in a sequence is Kolmogorov Complexity,
i.e., the shortest program that generates the numbers etc...
If you want some light reading re: chess, I suggest ``The Queen's Gambit''
by Walter Tevis (he wrote ``The Hustler'' and ``The man who fell to earth'')
∂13-Oct-88 2355 helen@psych.Stanford.EDU Noon Saturday
Received: from psych.Stanford.EDU by SAIL.Stanford.EDU with TCP; 13 Oct 88 23:55:23 PDT
Received: by psych.Stanford.EDU (3.2/4.7); Thu, 13 Oct 88 23:50:23 PDT
Date: Thu, 13 Oct 88 23:50:23 PDT
From: helen@psych.Stanford.EDU (Helen Cunningham)
To: JMC@SAIL.Stanford.EDU
Subject: Noon Saturday
Let's meet at the psych dept. as usual. You want to meet behind,
as we agreed before?
-h
∂14-Oct-88 0025 @CUNYVM.CUNY.EDU:YANG@NORUNIT.BITNET re: Your travel expenses, IFIP China.
Received: from CUNYVM.CUNY.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 00:25:11 PDT
Received: from NORUNIT.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.1) with BSMTP id 6956; Fri, 14 Oct 88 03:24:35 EDT
Date: Fri, 14 Oct 88 08:24:34 ECT
To: JMC@SAIL.Stanford.EDU
From: YANG%NORUNIT.BITNET@CUNYVM.CUNY.EDU
Comment: CROSSNET mail via SMTP@INTERBIT
In-Reply-To: <41Djf@SAIL.Stanford.EDU>
Subject: re: Your travel expenses, IFIP China.
thanks!
Jianhua Yang
∂14-Oct-88 1125 Mailer Re: disk use charges
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 11:25:43 PDT
Date: Fri 14 Oct 88 11:20:55-PDT
From: William "Chops" Westfield <BILLW@Score.Stanford.EDU>
Subject: Re: disk use charges
To: JMC@Sail.Stanford.EDU
cc: faculty@Score.Stanford.EDU, su-computers@Sail.Stanford.EDU
In-Reply-To: <Ooxvd@SAIL.Stanford.EDU>
Message-ID: <12438408927.9.BILLW@Score.Stanford.EDU>
1. The charges are way out of line with costs for disk file -
by a factor larger than 10 and perhaps larger than 100. SAIL's
present (somewhat reduced) charges of $2.75 per megabit month
would pay for RAM chips to store the files in two months and for
buying disk units in one month.
This is true of modern technology RAM and disks. Unfortunately, SAIL
is NOT a modern computer, and the most modern disk technology it can use
dates back to 1981 or so. Although the storage capacity of one of these
drives charged at current rates will pay for a drive in a month, this is
only because you can purchase such obsolete drives on the used market for
less than their original price... Worse, these drives have a half life of
about one year, and are physically very large. You can not add more drives
to SAIL because there is no room in the machine room (or, more accurately,
any room there is is better spent on more modern machines, which can fit
4 to 10 times the disk in the same floor space).
Therefore, SAIL (and Score, which uses the same disk technology) have
a fixed amount of disk space available. It is not enough for EVERYONE
to permanantly save all of their files, and I suspect that the current
charging policies are as much to discorage "wasteful" use of the disk as
anything else (and I suppose that this can be considered "evil" - at least
I consider taxes imposed solely to reduce consumption "evil" in other cases).
(Score's file system of over 2E9 Bytes (which was very large for a filesystem
when it was created), has less than 3% empty space...)
By the way, what is your estimate of they amount of space necessary to
save all your files? I would suspect that the current microcomputer
"standard" size of 40 Mbytes is too small. The 250 MB on a NEXT computer
might be enough. Would cheap disk space make you happy if it was not
available to SAIL?
Bill Westfield
-------
∂14-Oct-88 1405 S.SALUT@LEAR.STANFORD.EDU Re: October surprise
Received: from LEAR.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 14:05:03 PDT
Date: Fri 14 Oct 88 14:00:50-PDT
From: Alex Bronstein <S.SALUT@LEAR.STANFORD.EDU>
Subject: Re: October surprise
To: JMC@SAIL.STANFORD.EDU
In-Reply-To: <101xBf@SAIL.Stanford.EDU>
Message-ID: <12438438040.150.S.SALUT@LEAR.STANFORD.EDU>
Are you talking of the trade deficit figures?
Alex
-------
∂14-Oct-88 1413 S.SALUT@LEAR.STANFORD.EDU re: October surprise
Received: from LEAR.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 14:12:56 PDT
Date: Fri 14 Oct 88 14:08:45-PDT
From: Alex Bronstein <S.SALUT@LEAR.STANFORD.EDU>
Subject: re: October surprise
To: JMC@SAIL.STANFORD.EDU
In-Reply-To: <l2rNu@SAIL.Stanford.EDU>
Message-ID: <12438439482.150.S.SALUT@LEAR.STANFORD.EDU>
Ah.. I see. I don't have enough American experience (7 years) to have noticed
a pattern, but that makes sense.
Alex
-------
∂14-Oct-88 1420 P.EPSTEIN@GSB-WHY.Stanford.EDU re: 3rd party candidates
Received: from GSB-WHY.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 14:20:53 PDT
Date: Fri 14 Oct 88 14:19:26-PDT
From: David L. Epstein <P.EPSTEIN@GSB-WHY.Stanford.EDU>
Subject: re: 3rd party candidates
To: JMC@Sail.Stanford.EDU
In-Reply-To: <l2rR9@SAIL.Stanford.EDU>
Message-ID: <12438441426.12.P.EPSTEIN@GSB-WHY.Stanford.EDU>
You are indeed correct; I confused Madison and Hamilton in my posting. Sorry
about that, but I don't think it affects my general point about third parties.
From the primordial soup of early American politics emerged, finally, the
Whig party and the Democrat-Republican party. Ever since then, there have
only been two major parties. At one point the Whigs may have been termed
a third party, but that would have been only a temporary lable during the
initial sorting out of factions.
David Epstein
-------
∂14-Oct-88 2133 beeson@ucscd.UCSC.EDU triangles
Received: from ucscd.UCSC.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 21:33:08 PDT
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA12377; Fri, 14 Oct 88 21:36:15 PDT
Date: Fri, 14 Oct 88 21:36:15 PDT
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8810150436.AA12377@ucscd.UCSC.EDU>
To: jmc@sail.stanford.edu
Subject: triangles
I now believe the same triangles are embeddable in 3-space and 4-space.
I almost have it proved: I only need to prove the following
intuitively quite believable thing: given a vector of integers (u,v,w),
then the unit circle in the plane of vectors orthogonal to (u,v,w) does
contain a rational point.
I probably won't have time to work on this any more until at least
November, as I'm trying to get my AI-based educational software ready
for a show-and-tell in Columbus, Ohio at the end of this month. That's
why I bothered to tell you about it before finishing.
∂14-Oct-88 2136 beeson@ucscd.UCSC.EDU the regular tetrahedron
Received: from ucscd.UCSC.EDU by SAIL.Stanford.EDU with TCP; 14 Oct 88 21:36:25 PDT
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA12477; Fri, 14 Oct 88 21:39:31 PDT
Date: Fri, 14 Oct 88 21:39:31 PDT
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8810150439.AA12477@ucscd.UCSC.EDU>
To: jmc@sail.stanford.edu
Subject: the regular tetrahedron
The natural generalization of the triangle question is not to
polygons but to simplices. For example, is the regular tetrahedron
embeddable in three-space with its vertices at lattice points?
∂15-Oct-88 1038 meyer@theory.lcs.mit.edu hello
Received: from stork.LCS.MIT.EDU by SAIL.Stanford.EDU with TCP; 15 Oct 88 10:37:54 PDT
Received: by stork.LCS.MIT.EDU
id AA01253; Sat, 15 Oct 88 13:37:43 EDT
Date: Sat, 15 Oct 88 13:37:43 EDT
From: meyer@theory.lcs.mit.edu
Message-Id: <8810151737.AA01253@stork.LCS.MIT.EDU>
To: jmc@sail.stanford.edu
Subject: hello
Let me know if you get this.
Regards,
Prof. Albert R. Meyer
MIT Lab. for Computer Science
Chairman, Project MAC 25th Anniversary Celebration
∂16-Oct-88 1148 beeson@ucscd.UCSC.EDU re: triangles
Received: from ucscd.UCSC.EDU ([128.114.129.2]) by SAIL.Stanford.EDU with TCP; 16 Oct 88 11:48:35 PDT
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA03310; Sun, 16 Oct 88 11:51:32 PDT
Date: Sun, 16 Oct 88 11:51:32 PDT
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8810161851.AA03310@ucscd.UCSC.EDU>
To: JMC@SAIL.Stanford.EDU, beeson@ucscd.UCSC.EDU
Subject: re: triangles
Not so fast! It turns out the "intuitively quite believable thing"
is false! There's no rational point on the circle perpendicular to
(1,1,1). So the triangle problem will require a bit more number
theory. Too bad, I would love to go on working on it now while I'm
into it, but I really can't afford to.
∂16-Oct-88 1149 beeson@ucscd.UCSC.EDU re: the regular tetrahedron
Received: from ucscd.UCSC.EDU ([128.114.129.2]) by SAIL.Stanford.EDU with TCP; 16 Oct 88 11:49:26 PDT
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA03321; Sun, 16 Oct 88 11:52:18 PDT
Date: Sun, 16 Oct 88 11:52:18 PDT
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8810161852.AA03321@ucscd.UCSC.EDU>
To: JMC@SAIL.Stanford.EDU, beeson@ucscd.UCSC.EDU
Subject: re: the regular tetrahedron
No, I haven't even tried to tackle the regular tetrahedron problem, as I
would like to finish up triangles first.
∂16-Oct-88 1420 hoffman@csli.Stanford.EDU briefly
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 16 Oct 88 14:20:42 PDT
Received: by csli.Stanford.EDU (4.0/4.7); Sun, 16 Oct 88 14:22:11 PDT
Date: Sun 16 Oct 88 14:22:09-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: briefly
To: jmc@SAIL.STANFORD.EDU
Message-Id: <593040129.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
I will call you monday or tuesday of next week, but this note is principally
a reminder of your upcoming talk, at 3:15 on Oct. 21 on Formalizing Common
Sense Knowledge and Reasoning in Mathematical Logic. As I expect that you
will draw a larger crowd than most speakers, I am trying to change the
forum to a larger room. (I will notify you.) Could you send some brief
abstract or introduction so that I may post it to students and the CSLI
calender? (4-5 lines is sufficient.)
Also, after much work, Searle has tentatively agreed to a debate, which means
I think he will do it, but it will take more work on my part. I have not
yet organized a debate, and was wondering what sort of format would you like?
Would it be acceptable to have it televised/video taped for future reference?
(Nilsson said that he would be willing to organize this and thinks that it
would be a good idea for educational purposes for students.) Are there any
relevant factors which I am forgetting?
thanks and see you soon
reid
-------
∂17-Oct-88 0759 meyer@THEORY.lcs.mit.edu hello
Received: from stork.LCS.MIT.EDU ([18.26.0.191]) by SAIL.Stanford.EDU with TCP; 17 Oct 88 07:59:05 PDT
Received: by stork.LCS.MIT.EDU
id AA00327; Mon, 17 Oct 88 10:58:20 EDT
Date: Mon, 17 Oct 88 10:58:20 EDT
From: meyer@THEORY.lcs.mit.edu
Message-Id: <8810171458.AA00327@stork.LCS.MIT.EDU>
To: JMC@sail.stanford.edu
In-Reply-To: John McCarthy's message of 15 Oct 88 1121 PDT <102Yng@SAIL.Stanford.EDU>
Subject: hello
Glad to be in touch with you.
We haven't yet gotten your RSVP, so I wanted to inquire about--and
encourage--your attendance at our Project Mac 25th Anniversary Celebration
Oct. 26-27, especially the Testimonial Banquet on Oct. 26. Many of your old
friends and admirers are looking forward to the opportunity to acknowledge
the intellectual contributions you made here at MIT which helped set the
stage for the founding of Project MAC.
Below is the email announcement of our Celebration plans. I look forward to
hearing from you.
Regards, A.
------------------------------------------------------
Date: Tue, 30 Aug 88 11:54 EDT
To: All@theory.LCS.MIT.EDU, BBoards@theory.LCS.MIT.EDU
Subject: MIT Project MAC 25th Anniversary Research Symposium
From: MAC25-REGISTRATION@xx.lcs.mit.edu
*****************************************************************
MIT COMPUTER SCIENCE RESEARCH SYMPOSIUM
IN CELEBRATION OF THE
25th ANNIVERSARY OF THE FOUNDING OF PROJECT MAC
OCTOBER 26-27, 1988
MIT, CAMBRIDGE, MA
*****************************************************************
Sponsored by the MIT
Laboratory for Computer Science
and
Industrial Liaison Program
CONFERENCE DESCRIPTION: The symposium will cover a full range of Computer
Science research ongoing at MIT LCS and AI Lab--the two labs which
grew from the original ``Project MAC'' founded in 1963. Leading researchers
from the faculty and staff of the laboratories will highlight current
research and future activities in multiprocessors; distributed systems;
intelligent systems (AI), linguistics and robotics; cryptology, complexity
and random computation theory; parallel algorithms and programming languages;
and computers and economic productivity. The symposium will be of interest
to those seeking an overview of research as well as to specialists.
LECTURES OPEN TO THE PUBLIC: without charge, after seating by invited
and ILP affiliated guests.
PLACE: Kresge Auditorium, MIT.
*****************************************************************
SCHEDULE AND PROGRAM
TUESDAY, October 25
REGISTRATION (5PM-8PM) at Kresge Auditorium
RECEPTION (6PM-9PM) at the MIT Museum (Invited and ILP affiliated guests only)
WEDNESDAY, October 26
REGISTRATION (7:45AM-continuing) at Kresge Auditorium
WELCOMING REMARKS (8:45AM-9AM)
The MIT Administration
Michael L. Dertouzos, LCS Director
Albert R. Meyer, Symposium Chair
SESSION 1 (9AM-Noon) Chair: Fernando J. Corbato
John V. Guttag, Why Programming is Too Hard and What to Do About It
Nicholas P. Negroponte, Beyond the Desktop Metaphor
Barbara H. Liskov, Issues in Distributed Computing
Robert W. Scheifler, Windows in Time: The X Window System
David D. Clark, The Changing Nature of Computer Networks
LUNCH (Noon-1:30PM)
SESSION 2 (1:30PM-2:20PM) Chair: Robert M. Fano
Michael L. Dertouzos, Computers for Productivity
SESSION 3 (2:25PM-5:00PM) Chair: Randall Davis
Peter Szolovits, Knowledge-Based Systems
Ramesh S. Patil, An Expert System for Arrhythmia Detection in Noise
Berthold K.P. Horn, Parallel Networks for Vision
Rodney A. Brooks, Artificial Creatures
Marc H. Raibert, Robots that Run
TESTIMONIAL BANQUET (6:30PM-11:00PM) (By Invitation)
THURSDAY, Oct. 27
REGISTRATION (8:45AM-continuing) at Kresge Auditorium
SESSION 4 (9AM-Noon) Chair: Frederick C. Hennie, III
Harold Abelson, Computation as a Framework for Engineering Education
Albert R. Meyer, Observing Concurrent Processes
Michael F. Sipser, We Still Don't Know if P=NP
Shafi Goldwasser, The Quest for Provably Unbreakable Codes
Silvio Micali, Nothing but the Truth: Zero-Knowledge Protocols
Ronald L. Rivest, Learning Theory: What's Hard and What's Easy
LUNCH (Noon-1:30PM)
SESSION 5 (1:30PM-2:20PM) Chair: Marvin L. Minsky
Joel Moses, Cultural Biases in CS and AI
SESSION 6 (2:25PM-5:00PM) Chair: Jack B. Dennis
Arvind, A Dataflow Approach to General Purpose Parallel Computing
William J. Dally, Fine-Grain Concurrent Computing
Charles E. Leiserson, New Machine Models for Synchronous Parallel Algorithms
Gerald J. Sussman, Dynamicist's Workshop: Automatic Preparation, Execution,
and Analysis of Numerical Experiments
*****************************************************************
ABSTRACTS: Detailed abstracts of the above talks is availabe upon request.
INVITATIONS: The symposium lectures are open to the public without charge.
Lunch will be provided for invited and ILP affiliated guests, while
the banquet is for invited guests and their companions. Invitations
are being sent to alumni and scientific collaborators of Project MAC/LCS/AI,
contract monitors and similar liaison officers from other organizations,
and other laboratory affiliates.
Completing the registration form below will also serve as a request for an
invitation if you have not received one.
*****************************************************************
REGISTRATION FORM
TITLE (Mr. Ms. Dr. ...):
FIRSTNAME:
MIDDLE INITIAL:
LASTNAME:
POSITION (Vice President,...):
COMPANY:
DEPARTMENT/DIVISION:
ADDRESS:
CITY:
STATE:
COUNTRY:
ZIP:
TEL:
EMAIL-ADDRESS:
I WOULD LIKE TO ATTEND (mark with `x'):
October 25, RECEPTION:
October 26, SYMPOSIUM:
LUNCH:
BANQUET:
October 27, SYMPOSIUM:
LUNCH:
BANQUET COMPANION'S NAME:
REGISTRANT'S AFFILIATION
Former MAC/LCS/AI Lab member or student. Group:
Year:
Other MAC/LCS/AI affiliation
(funding officer, research collaborator,...):
Year:
Lab-member reference:
ILP affiliated (mark with `x'):
No affiliation, just want to register for
the symposium (mark with `x'):
*****************************************************************
SEND Registration and further inquiries by EMAIL to
Internet: MAC25-registration@XX.LCS.MIT.EDU
or by REGULAR MAIL to
Professor Albert R. Meyer, Chairman
Project MAC 25th Anniversary Symposium
MIT Laboratory for Computer Science
545 Technology Square
Cambridge, MA 02139
tel: (617) 258-8215
-------
∂17-Oct-88 0925 meyer@theory.lcs.mit.edu further invites: mac25
Received: from stork.LCS.MIT.EDU ([18.26.0.191]) by SAIL.Stanford.EDU with TCP; 17 Oct 88 09:22:08 PDT
Received: by stork.LCS.MIT.EDU
id AA00362; Mon, 17 Oct 88 12:21:47 EDT
Date: Mon, 17 Oct 88 12:21:47 EDT
From: meyer@theory.lcs.mit.edu
Message-Id: <8810171621.AA00362@stork.LCS.MIT.EDU>
To: JMC@sail.stanford.edu
Subject: further invites: mac25
Barbara Liskov suggested that some of the members of your original LISP
development team, eg, Steve Russell, should also have been invited to the
Project MAC celebration. I'd be grateful if you could give me names and
addresses of any such folks you think appropriate (including Russell, whose
address we don't have).
Regards, A.
∂17-Oct-88 0928 meyer@theory.lcs.mit.edu hello
Received: from stork.LCS.MIT.EDU ([18.26.0.191]) by SAIL.Stanford.EDU with TCP; 17 Oct 88 09:27:24 PDT
Received: by stork.LCS.MIT.EDU
id AA00371; Mon, 17 Oct 88 12:27:02 EDT
Date: Mon, 17 Oct 88 12:27:02 EDT
From: meyer@theory.lcs.mit.edu
Message-Id: <8810171627.AA00371@stork.LCS.MIT.EDU>
To: JMC@sail.stanford.edu
In-Reply-To: John McCarthy's message of 17 Oct 88 0916 PDT <l3X9T@SAIL.Stanford.EDU>
Subject: hello
Date: 17 Oct 88 0916 PDT
From: John McCarthy <JMC@sail.stanford.edu>
[In reply to message sent Mon, 17 Oct 88 10:58:20 EDT.]
If you can pay travel expenses, yes, otherwise no.
-----------
I'll see what I can do; be back to you asap.
Regards, A.
∂17-Oct-88 1132 meyer@theory.lcs.mit.edu hello
Received: from stork.LCS.MIT.EDU by SAIL.Stanford.EDU with TCP; 17 Oct 88 11:32:30 PDT
Received: by stork.LCS.MIT.EDU
id AA00421; Mon, 17 Oct 88 14:32:13 EDT
Date: Mon, 17 Oct 88 14:32:13 EDT
From: meyer@theory.lcs.mit.edu
Message-Id: <8810171832.AA00421@stork.LCS.MIT.EDU>
To: JMC@sail.stanford.edu
In-Reply-To: John McCarthy's message of 17 Oct 88 0940 PDT <h3XPL@SAIL.Stanford.EDU>
Subject: hello
Date: 17 Oct 88 0940 PDT
From: John McCarthy <JMC@sail.stanford.edu>
[In reply to message sent Mon, 17 Oct 88 12:27:02 EDT.]
I have a call in to someone in New York who may have Russell's address -
said to be in California.
------------
Happy to pay your travel expenses. Look forward to seeing you. Let me know
if you need help with arrangements.
Regards, A.
∂17-Oct-88 1150 JSW Thesis
Do you have some time this afternoon to go over my thesis work?
I want to make sure that I'm preparing a reasonable set of things
for the draft and the orals talk.
∂17-Oct-88 1237 P.EPSTEIN@GSB-WHY.Stanford.EDU re: 3rd party candidates
Received: from GSB-WHY.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Oct 88 12:37:50 PDT
Date: Mon 17 Oct 88 12:36:16-PDT
From: David L. Epstein <P.EPSTEIN@GSB-WHY.Stanford.EDU>
Subject: re: 3rd party candidates
To: JMC@Sail.Stanford.EDU
In-Reply-To: <v2r#w@SAIL.Stanford.EDU>
Message-ID: <12439209076.8.P.EPSTEIN@GSB-WHY.Stanford.EDU>
Yes, I think that the Republicans were indeed the new kid on the block after
the civil war. I am not very familiar with the dynamics of the change, but it
seems that I remember that there was a mass movement in realigning the party
system in wake of the civil war. The original party contours were shaken up
in the Reconstruction period, aligning the North against the South. In any
case, the point is that there were still never more than two major parties
at any one time, and for the party labels to even change took a civil war.
Thus the inevitability of a national two party system is still confirmed even
in the wake of the upheaval following the Civil War.
David Epstein
-------
∂17-Oct-88 1356 MPS phone number
Al Ewert's office called. You wanted a number. Steve Russell of
Isix?? (the secretary thinks) 415-578-1900.
Pat
∂17-Oct-88 1442 VAL Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
FIRST ORDER THEORIES OF QUANTIFICATION
Arkady Rabinov (AIR@SAIL)
Stanford University
Friday, October 21, 3:15pm
MJH 301
Many applications of logic to AI require that propositions be reified, i.e.,
made elements of the domain of reasoning. This is needed, in particular, in
formalizing reasoning about knowledge and about preconditions of actions.
Difficulties arise when the quantificational structure of propositions is
essential. We show how ideas from combinatory logic can be used to deal with
this problem. This approach allows us to construct first order theories in
which lambda-abstraction and quantification can be easily expressed as terms
of the language.
∂17-Oct-88 1649 VAL Etherington
Here is a message from Etherington and my draft reply. I'd like to
know your opinion before I send it out.
∂11-Oct-88 1500 ether%allegra@research.att.com kr'89 panel
Received: from research.att.com (ATT.ARPA) by SAIL.Stanford.EDU with TCP; 11 Oct 88 15:00:10 PDT
From: ether@allegra.att.com
Date: Tue, 11 Oct 88 17:45:16 edt
>From: allegra!ether (David Etherington)
To: arpa!val@sail.stanford.edu
Subject: kr'89 panel
Hi Vladimir.
I'm in charge of organizing an afternoon on Nonmonotonic Reasoning at
the Toronto KR'89 conference (organized by Reiter, Brachman, & Levesque,
from May 15-18, '89). I hope that you've already heard about the
conference, and are planning to come. The special session will be
Tuesday PM, from about 13:30 - 17:00, with a 1/2 hr break, in parallel
with a single session of submitted papers.
The current outline is:
1hr: invited talk
1/2 hr: summary of Munich challenges
1/2 hr: break
1.5 hr: Panel
I'd like to ask you to participate in the panel session. I haven't
finalized the size or composition of the panel, although it will
probably be 4-5 participants. The members will be chosen in the hopes
of making it both lively and thought-provoking.
The panel will be something like "NMR: An Idea Whose Time Has Gone?".
The idea will be to discuss whether there is any future in the
'classical' approaches to nonmonotonicity (e.g., do they have any hope
of surviving existing challenges and/or scaling up to real problems)
or whether other approaches (e.g., qualitative reasoning, probability,
etc) aren't more appropriate paradigms.
Context will be provided by the invited talk (tentatively, Judea Pearl
on the relationship between nonmonotonic and probabilistic reasoning)
and the summary slot, which is intended to present a cohesive overview
of what went on at the Munich workshop, especially re: the challenges
that Poole and others brought up, and why they are important.
Whether or not you are interested, I would appreciate it if you could
please let me know what you think of the idea, and anyone you think
should be involved.
Thanks very much,
David Etherington.
allegra!ether
ether@research.att.com
-----
Dear David,
Thank you for your message. I know about the Toronto conference, I have agreed
to be on its program committee, and I'm going to submit a paper. I like your
plan to have an afternoon on Nonmonotonic Reasoning, with Judea invited to
speak about his work, and with a summary of "Munich challenges" presented. (I
presume you're asking Poole to do the summary part.)
But I am puzzled by the idea of discussing "whether there is any future in
the 'classical' approaches to nonmonotonicity (e.g., do they have any hope
of surviving existing challenges and/or scaling up to real problems)". I am
also embarassed by the tentative title of the panel, even if it is meant
entirely as a joke. Is there any serious crisis that I am not aware of and
that prompts you to talk about "any hope of surviving"? Surely it's not the
existence of "challenges," i.e., open problems. I will ask whether NMR is an
idea whose time has gone when I see that a year or two passes without new
challenging problems discovered, so that we don't have new topics for our
research.
I would rather suggest a panel on "Open Problems in Nonmonotonic Reasoning,"
or something like this. For instance, you can ask people like David Poole,
Matt Ginsberg and Kurt Konolige to describe the open problems as they see
them (including the "Munich challenges"), and ask the other members of the
panel to respond. I would be happy to take part in a panel of this kind.
The discussion will be more constructive if a draft description of the
"challenges" can be made available to us in advance.
What do you think?
--Vladimir
∂17-Oct-88 1842 GLB oral exams
To: JMC@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU, jcm@POLYA.Stanford.EDU,
CLT@SAIL.Stanford.EDU
I think I should take my orals. Would it be good for everyone a month from today,
Thursday 17 November (or something like that)?
Gianluigi
∂18-Oct-88 0105 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Kids, don't try this at home!
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 18 Oct 88 01:05:37 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 327755; Tue 18-Oct-88 03:59:17 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73114; Tue 18-Oct-88 00:55:47 PDT
Date: Tue, 18 Oct 88 00:53 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Kids, don't try this at home!
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: rwg@RUSSIAN.SPA.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19881018075319.5.RWG@TSUNAMI.SPA.Symbolics.COM>
k↑2 a[k-1] - 1
a[1] := 1, a[k] := --------------,
a[k-1] + k↑2
tan(π/sqrt(2)) + tanh(π/sqrt(2))
then a[∞] = -------------------------------- .
tan(π/sqrt(2)) - tanh(π/sqrt(2))
∂18-Oct-88 0327 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM this is closed form?
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 18 Oct 88 03:27:23 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 327766; Tue 18-Oct-88 06:25:51 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73125; Tue 18-Oct-88 03:22:21 PDT
Date: Tue, 18 Oct 88 03:19 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: this is closed form?
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: rwg@RUSSIAN.SPA.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19881018101952.7.RWG@TSUNAMI.SPA.Symbolics.COM>
i - 1 i
(-----)! (- -)!
2 2
i LOG(---------------)
∞ i i + 1
==== K (-)! (- -----)!
\ (- 1) 2 2
> ATAN(------) = - ---------------------- = - 5.066709032166229819851B-1
/ K 2
====
K = 1
∂18-Oct-88 1033 @RELAY.CS.NET:GOTO@ntt-20.ntt.jp Your visit to Kyoto / My visit to Stanford
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 18 Oct 88 10:33:30 PDT
Received: from relay2.cs.net by RELAY.CS.NET id ae00277; 18 Oct 88 12:21 EDT
Received: from ntt.jp by RELAY.CS.NET id aa12400; 18 Oct 88 11:54 EDT
Received: by ntt-sh.ntt.jp (3.2/ntt-sh-02) with TCP; Wed, 19 Oct 88 00:15:38 JST
Received: by MECL.NTT.jp (3.2/NTTcs02) with TCP; Wed, 19 Oct 88 00:14:19 JST
Date: Wed 19 Oct 88 00:16:18
From: Shigeki Goto <Goto@ntt-20.ntt.jp>
Subject: Your visit to Kyoto / My visit to Stanford
To: jmc%sail.stanford.edu@nuesun.ntt.jp, clt%sail.stanford.edu@nuesun.ntt.jp
Cc: goto@ntt-20.ntt.jp
Message-Id: <12439412974.12.GOTO@NTT-20.NTT.JP>
Dear John and Carolyn,
Congratulations on John's Kyoto Prize.
I will attend the awarding ceremony and the lecture on November 10 and
11 in Kyoto.
Incidentally, I will visit Stanford University with one of NTT
directors. He will have a meeting, NTT-Stanford steering committee,
on October 27 and I should help him. I guess Prof. Nils Nilsson will
attend it. Taking advantage of this opportunity, I will give my talk
at CSLI on October 31 (Mon).
See you soon at Stanford or Kyoto.
-- Shigeki --
sg@sail or goto%ntt-20.ntt.jp@relay.cs.net
-------
∂18-Oct-88 1438 VAL msg from Dmitry Lenkov
He says you can write a check payable to Samizdat Bulletin. If you wish, I
can pass it to him tonight.
∂18-Oct-88 1511 bhayes@polya.Stanford.EDU Coverup
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Oct 88 15:11:32 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA09451; Tue, 18 Oct 88 15:11:42 PDT
Date: Tue, 18 Oct 88 15:11:42 PDT
From: Barry Hayes <bhayes@polya.Stanford.EDU>
Message-Id: <8810182211.AA09451@polya.Stanford.EDU>
To: JMC@sail
Subject: Coverup
I admit, timing is all, but Coverup was made last year...
∂18-Oct-88 1554 mkatz@sesame.stanford.edu Re: October surprise
Received: from sesame.stanford.edu by SAIL.Stanford.EDU with TCP; 18 Oct 88 15:54:39 PDT
Received: by sesame.stanford.edu (5.57/Ultrix2.4-C)
id AA09364; Tue, 18 Oct 88 15:53:46 PDT
Date: Tue, 18 Oct 88 15:53:46 PDT
From: mkatz@sesame.stanford.edu (Morris Katz)
Message-Id: <8810182253.AA09364@sesame.stanford.edu>
To: JMC@sail.stanford.edu (John McCarthy)
Subject: Re: October surprise
Newsgroups: su.etc
In-Reply-To: <l4rYc@SAIL.Stanford.EDU>
Organization: Stanford University
In article <l4rYc@SAIL.Stanford.EDU> you write:
>There is a tradition in American politics that one side or the
>other will present startling charges just beore the election.
>The traditional name is the October surpris. It needs just
>enough credibility to last through the elction. Here it is
>October 18, and we have posters for Coverup right on schedule.
>Line up, suckers. Dukakis has to decide whether to endorse it.
>I gather that so far he has said nothing.
In case you didn't notice, this movie was publicized by such "liberal" journals
as the New York Times well over a month ago. Just because stanford students
seem to be slow on picking up on the issue does not mean it was pulled as an
Oct. surprise. Furthermore, the original accusations aired over 6 months ago,
but were not well publicized. I am sure if that had been we would have heard
the traditional cries about the liberal press. You can't have it both ways.
--
--
Morry Katz
katz@polya.stanford.edu
∂18-Oct-88 1602 BSCOTT@Score.Stanford.EDU Tenured Faculty Meeting
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Oct 88 16:02:14 PDT
Date: Tue 18 Oct 88 15:58:17-PDT
From: Betty Scott <BSCOTT@Score.Stanford.EDU>
Subject: Tenured Faculty Meeting
To: Binford@Coyote.Stanford.EDU, Feigenbaum@SUMEX-AIM.Stanford.EDU,
Golub@Score.Stanford.EDU, ZM@Sail.Stanford.EDU, JMC@Sail.Stanford.EDU
cc: BScott@Score.Stanford.EDU
Message-ID: <12439507996.45.BSCOTT@Score.Stanford.EDU>
I need your votes on the appointment of Fan Chung as Professor of Computer
Science. The papers are available in Phyllis Winkler's Office, MJH 326,
if you wish to see them.
Will appreciate your votes as soon as possible.
Thanks,
Betty
-------
∂18-Oct-88 1736 W.WENTWORTH@MACBETH.STANFORD.EDU Gann initiative
Received: from MACBETH.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 18 Oct 88 17:36:32 PDT
Date: Tue 18 Oct 88 17:31:46-PDT
From: Robert Wentworth <W.WENTWORTH@Macbeth.Stanford.EDU>
Subject: Gann initiative
Sender: W.WENTWORTH@MACBETH.STANFORD.EDU
To: su-etc@MACBETH.STANFORD.EDU
cc: jmc@SAIL.STANFORD.EDU
Reply-To: W.WENTWORTH@MACBETH.STANFORD.EDU
Message-ID: <12439525014.96.W.WENTWORTH@MACBETH.STANFORD.EDU>
JMC writes:
>I favor the Gann initiative. As an AIDS victim, he has had a lot of
>time to think about it,
Yes, Gann is an AIDS victim. But, having contracted the disease
through a blood transfusion (which I suspect is not the major
transmission route for AIDS, certainly not now that donor blood is
tested), he may well have a skewed or even vindictive perspective on
the issue. He does not suffer from the stigma of having contracted
the disease sexually, nor, I suspect, does he identify with those for
whom this is an issue. Also, the solution he proposes is one that
might have helped prevent infection via transfusion (now a relatively
moot issue), but which health officials believe to be inefficient at
best at best at combatting current forms of transmission.
(Also, while Gann is the official sponsor, doesn't this look an awful
lot like the old LaRouche AIDS initiatives?)
> and it only proposes to treat AIDS as other
>communicable diseases have been treated in the past.
The analogy to other communicable diseases for which this has been
done is weak. The epidemiological conditions are not the same.
A large majority of AIDS carriers are members of well-identified
high-risk groups, who have been exposed with rather high probability.
As a result, most money spent on contact tracing (which is expected
be quite expensive) will be spent to warn people of dangers that they
already know about.
>The opposition
>is based on the idea that homosexuals have been oppressed, and since
>AIDS mostly afflicts them, we should change past public health practice
>in order to accomodate their preferences.
The stigma associated with homosexuality and sexually transmitted
disease is an important part of the argument against the initiative,
but it is compelling only in conjunction with the other unique aspects
of the AIDS epidemic.
>While the homosexual
>organizations oppose nonvoluntary AIDS testing, and aggressive contact
>tracing, there is every reason to believe that these measures will
>reduce the number of homosexuals who die of AIDS. Compulsory contact
>tracing has greatly reduced other infectious diseases.
"COMPULSORY CONTACT TRACING HAS GREATLY REDUCED OTHER INFECTIOUS
DISEASES"??!!
Don't you think this has something to do with the fact that other
infectious diseases are treatable??
Public health officials concur that the Gann initiative, Prop 102,
will AT BEST be counterproductive and wasteful -- and many believe
it would be a public health DISASTER, making a bad situation much
worse.
Vote no on 102!
-------
∂18-Oct-88 2022 barwise@russell.Stanford.EDU Re: Rota article
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Oct 88 20:22:35 PDT
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Tue, 18 Oct 88 20:24:58 PDT
To: JMC@SAIL.Stanford.EDU
Subject: Re: Rota article
In-Reply-To: Your message of 18 Oct 88 15:45:00 PDT.
<l4sff@SAIL.Stanford.EDU>
Address: CSLI, Stanford University, Stanford, CA 94305 (415) 723-0110
Date: Tue, 18 Oct 88 20:24:56 PDT
From: Jon Barwise <barwise@russell.Stanford.EDU>
Oh, dear, sounds bad. I assumed he knew what he was talking about. I
guess I had better write him and either reject the paper (which is a
bit hard since I invited him to submit something) or at least suggest
that he improve it, update it, and check up on his facts.
Thanks. Why don't you forget this version.
Jon
∂19-Oct-88 0020 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM this is closed form?
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 19 Oct 88 00:20:09 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 328020; Wed 19-Oct-88 03:18:19 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73210; Wed 19-Oct-88 00:14:51 PDT
Date: Wed, 19 Oct 88 00:12 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: this is closed form?
To: rwg@RUSSIAN.SPA.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19881018101952.7.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19881019071219.8.RWG@TSUNAMI.SPA.Symbolics.COM>
Date: Tue, 18 Oct 88 03:19 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
i - 1 i
(-----)! (- -)!
2 2
i LOG(---------------)
∞ i i + 1
==== K (-)! (- -----)!
\ (- 1) 2 2
> ATAN(------) = - ---------------------- = - 5.066709032166229819851B-1
/ K 2
====
K = 1
Oh, bleep. Sobvious. The middle expression simplifies to
i-1 i
arg(--- !) - arg(- !). Express each x! in terms of (x+n)!, and let n→∞.
2 2
∂19-Oct-88 0236 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM on the other hand
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 19 Oct 88 02:36:21 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 328026; Wed 19-Oct-88 05:34:47 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73213; Wed 19-Oct-88 02:31:21 PDT
Date: Wed, 19 Oct 88 02:28 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: on the other hand
To: rwg@RUSSIAN.SPA.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19881018101952.7.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19881019092848.9.RWG@TSUNAMI.SPA.Symbolics.COM>
∞
====
\ 6
> ATAN(----------) = ARG(BINOMIAL(3 i, i)) = 1.225344775226358086828B0
/ 2
==== K (K + 7)
K = 1
∂19-Oct-88 0751 davism@csd11.nyu.edu letter to Berkeley
Received: from csd11.nyu.edu ([128.122.128.37]) by SAIL.Stanford.EDU with TCP; 19 Oct 88 07:51:04 PDT
Received: by csd11.nyu.edu (3.2/25-eef)
id AA05832; Wed, 19 Oct 88 10:51:37 EDT
Date: Wed, 19 Oct 88 10:51:37 EDT
From: Martin Davis <davism@csd11.nyu.edu>
Message-Id: <8810191451.AA05832@csd11.nyu.edu>
To: jmc@sail.stanford.edu
Subject: letter to Berkeley
Hi John,
This is the promised prompt, it being about a month since my previous message.
The letter is due by November 1. To save you retrieving, I will send in 3
successive messages the relevant documents:
1. ad describing the position & giving the address to which the letter
needs to be sent.
2. statement of my "qualifications and interests" which I included with my
letter of application.
3. my CV.
Letters have already been sent by Moschavakis and Hilary Putnam.
And ... thanks very much,
Martin
∂19-Oct-88 0752 davism@csd11.nyu.edu ad
Received: from csd11.nyu.edu ([128.122.128.37]) by SAIL.Stanford.EDU with TCP; 19 Oct 88 07:52:08 PDT
Received: by csd11.nyu.edu (3.2/25-eef)
id AA05836; Wed, 19 Oct 88 10:52:35 EDT
Date: Wed, 19 Oct 88 10:52:35 EDT
From: Martin Davis <davism@csd11.nyu.edu>
Message-Id: <8810191452.AA05836@csd11.nyu.edu>
To: jmc@sail.stanford.edu
Subject: ad
EDUCATION IN MATHEMATICS, SCIENCE, AND TECHNOLOGY
UNIVERSITY OF CALIFORNIA, BERKELEY
Pending Budgetary approval, three professorships are to be filled
by the Graduate School of Education, starting July 1989. Applicants at all
levels (tenure and tenure-track) are encouraged to apply. Successful
applicants will join Berkeley's outstanding program of research, doctoral
education, and professional training concerned with education in
mathematics, science, and technology. The program is strongly
interdisciplinary, with an emphasis on the educational applications of
cognitive science and computers. Candidates should preferably possess a
strong background in a mathematical, computational, or cognitive science.
They should have demonstrable excellence in research, teaching, and
applied interests in instructional improvement.
Applicants should send a resume, a statement describing
qualifications and interests, and the names of at least three references
to:
EMST Search Committee
Graduate School of Education
University of California, Berkeley
Berkeley, CA 94720
so as to be received by November 1, 1988. They should ask their references
to send letters directly to the Committee by this deadline.
∂19-Oct-88 0753 davism@csd11.nyu.edu qualifications & interests
Received: from csd11.nyu.edu ([128.122.128.37]) by SAIL.Stanford.EDU with TCP; 19 Oct 88 07:52:44 PDT
Received: by csd11.nyu.edu (3.2/25-eef)
id AA05840; Wed, 19 Oct 88 10:53:19 EDT
Date: Wed, 19 Oct 88 10:53:19 EDT
From: Martin Davis <davism@csd11.nyu.edu>
Message-Id: <8810191453.AA05840@csd11.nyu.edu>
To: jmc@sail.stanford.edu
Subject: qualifications & interests
QUALIFICATIONS AND INTERESTS
Martin Davis
As a researcher I am best described as a mathematical
logician with a longstanding interest in computer science. In the
field of automated deduction (which is of course an important
part of artificial intelligence), I am regarded as a "pioneer."
The anthology "Automation of Reasoning: Classical Papers on
Computational Logic 1957-1966" published in 1983 begins with a
historical article that I wrote and includes four research papers
of which I was author or coauthor. In fact the very first
research paper in the collection is my report dated 1957 on a
computer program I had written (in 1954) to carry out
Pressburger's procedure for additive arithmetic. The program was
written (in absolute binary) for a computer now housed in
Washington in the Smithsonian. Although my technical interests
have been quite varied, I've continued from time to time, to make
contributions to this field, and I am an editor of the Journal of
Automated Reasoning. I was asked to write an article for the
Encyclopedia of Artificial Intelligence and to be a commentator
for an article in The Behavioral and Brain Sciences.
My book "Computability and Unsolvability" which appeared
in 1958 was the first attempt in print to present an exposition
of the mathematical work of Godel, Turing, Church, and Post in a
manner that made contact with the already rapidly developing
computer technology. This book has been translated into Chinese,
Italian, and Japanese. It continues to live as a Dover reprint,
and it has been described both by logicians and by computer
scientists as a "classic." My book "Applied Nonstandard Analysis"
has also been translated: into Japanese and Russian.
Mathematicians know me best for my work on the tenth
problem on the famous list that David Hilbert had presented in
1900. Hilbert had asked for an algorithm to test polynomial
equations (in many unknowns) for the possession of integer
solutions. In a paper that I coauthored with Hilary Putnam and
Julia Robinson in 1961 (based in part on results from my
dissertation), we proved that if the class of equations is
expanded to permit exponents as unknowns, no such algorithm is
possible. We also showed that if even one polynomial equation
could be found whose solutions (in a suitable sense) grow
exponentially, then the same could be said of Hilbert's problem
as originally posed. In 1970, Yuri Matijasevich completed the
work by exhibiting such an equation.
After Matijasevich's breakthrough, I made some additional
contributions to the subject, and wrote one and coauthored
another expository articles on Hilbert's tenth problem. One,
coauthored with Reuben Hersh, was written for the Scientific
American and received the coveted Chauvenet Prize of the
Mathematical Association of America. A second, written for the
American Mathematical Monthly, was addressed to mathematicians.
It received two prizes, the Lester R. Ford Prize of the
Mathematical Association of America and the Leroy P. Steele Prize
of the American Mathematical Society. I was especially gratified
to receive these awards because I have always been very
interested in mathematical exposition. My abilities as an
expositor were also recognized by the Mathematical Association of
America by appointing me Earle Raymond Hedrick Lecturer. I was
asked to contribute an essay to the collection "Mathematics
Today" which was intended to help non-mathematicians to
appreciate contemporary mathematics. I also wrote the section on
"Theoretical Computer Science" for a report prepared in 1982 by
the National Research Council for the Congress of the United
States.
I have always liked to teach and have been a popular teacher.
In the Fall of 1972, I was presented the "Award for Superior
Teaching" by the Courant Institute Student Council. I have just been
an invited participant in a panel discussion at Logic Colloquium '88
in Padua, Italy on the teaching of logic. I have taught "in-service"
courses for high school teachers and Saturday courses for
mathematically gifted high school students. At New York University,
I've had many conversations with Anneli Lax about her innovative
basic course in "mathematical thinking"; a number of my ideas
apparently proved to be useful. In recent years my teaching has
mostly been at the graduate level, and I have had 21 successful
doctoral students, some of whom have had Ph.D. students of their own.
In the last few years my research has concentrated in two
quite disparate directions: a project to develop a theory of
software testing and a study of the historical interplay of logic
and computer science. In the first of these, we have sought appropriate
mathematical models of aspects of the empirical process that is
called software testing. Although we have obtained some
interesting insights, the field remains quite daunting. The other
direction, I find a fascinating exercise in intellectual history,
and would want to continue with it should I become a member of
the Graduate School of Education at Berkeley.
In addition to the letters of recommendation that are
being sent, you might also wish to consult some of the following
people at Berkeley: John Addison and Leon Henkin in Mathematics,
William Craig in Philosophy, Manual Blum and Richard Karp in
Computer Science, all of whom know me reasonably well.
I have just begun a term as chair of the Computer Science
Department at NYU, and were I to come to Berkeley, it would
represent a radical career shift. But, at least on my side, it
feels like a very appropriate one. Berkeley itself is a great
attraction for many reasons. But more important, the shift in
career focus excites me. While my computer science research has
been largely theoretical, I really do thoroughly enjoy using
computers in a hands-on mode. Although, I have never been
involved in serious research concerning the use of computers for
educational purposes, the challenging possibilities have long fascinated me.
Finally, the thought that my work might actually have an effect
in alleviating the critical social problem of widespread ignorance
of science and mathematics is very appealing.
∂19-Oct-88 0754 davism@csd11.nyu.edu resume
Received: from csd11.nyu.edu ([128.122.128.37]) by SAIL.Stanford.EDU with TCP; 19 Oct 88 07:53:27 PDT
Received: by csd11.nyu.edu (3.2/25-eef)
id AA05844; Wed, 19 Oct 88 10:54:00 EDT
Date: Wed, 19 Oct 88 10:54:00 EDT
From: Martin Davis <davism@csd11.nyu.edu>
Message-Id: <8810191454.AA05844@csd11.nyu.edu>
To: jmc@sail.stanford.edu
Subject: resume
Martin Davis
Courant Institute of Mathematical Sciences
New York University
251 Mercer Street
New York, NY 10012
(212)-998-3102
Born on March 8, 1928 in New York City
EDUCATION:
City College of New York, B.S. 1948
Princeton University, M.A. 1949, Ph.D. 1950
ACADEMIC POSITIONS:
Professor of Mathematics, Courant Institute of Mathematical Sciences,
New York University, 1965- . Joint appointment in Mathematics and
Computer Science, 1969 - . Chair of Computer Science 1988- .
Associate Professor and Professor of Mathematics, Belfer Graduate
School of Science, Yeshiva University, 1960-65.
Research Scientist and Adjunct Associate Professor of Mathematics, New
York University, 1959-60.
Assistant Professor and Associate Professor of Mathematics, Hartford
Graduate Division, Rensselaer Polytechnic Institute, 1956-59.
Assistant Professor of Mathematics, Ohio State University, 1955-56.
Assistant Professor of Mathematics, University of Calif. Davis, 1954-55.
Visiting Member, School of Mathematics, Inst. for Adv. Study, 1952-54.
Research Instructor in Mathematics, University of Illinois, 1950-52.
VISITING APPOINTMENTS:
Westfield College, University of London, England, 1968-69.
Belfer Graduate School of Science, Yeshiva University, 1970-71.
University of California, Berkeley (Mathematics and Computer Science)
and Stanford University (Artificial Intelligence Laboratory), 1976-77.
University of California, Santa Barbara, 1978-79.
University of California, Berkeley 1983-84.
HONORS AND AWARDS:
Leroy P. Steele Prize, American Mathematical Society, January 1975.
Chauvenet Prize, Mathematical Association of America, January 1975.
Lester R. Ford Prize, Mathematical Assoc. of America, January 1975.
Earle Raymond Hedrick Lecturer 1976, Mathematical Association of America.
Fellow of the A.A.A.S., January 1982.
Guggenheim Foundation Fellowship, 1983-84.
CONSULTED AT VARIOUS TIMES FOR:
Bell Telephone Laboratories, Murray Hill, N.J.
Applied Logic Inc., Princeton, N.J.
IBM Research Laboratories, Yorktown Heights, N.Y.
Rand Corporation, Santa Monica, California.
PROFESSIONAL SERVICE:
Editorial Boards: Journal of Symbolic Logic, Journal of the Association
for Computing Machinery. Journal of Automated Reasoning.
Served on MAA award committees for Chauvenet Prize and Hedrick
Lecturer.
Served on American Mathematical Society Nominations Committee.
Served on Nominations Committee, Section A, A.A.A.S.
Chairman, Nominations Committee, Association for Symbolic Logic
Chairman, Committee on Academic Freedom, Tenure, and Employment
Security of the American Mathematical Society.
Wrote section on "Theoretical Computer Science" for "Outlook for
Science and Technology - The Next Five Years," 1982, prepared by the
National Research Council for the Congress of the United States.
Served on Program Committee of Fifth Conference on Automated
Deduction; Local Arrangements Chairman for Sixth Conference on
Automated Deduction.
Chairman of committee to select first winner of prize for a
"landmark" contribution to automatic theorem proving.
Member of Program Committee for 1989 meeting on "Logic in
Computer Science."
INVITED PRESENTATIONS AT PROFESSIONAL CONFERENCES:
"Applications of Recursive Function Theory to Number Theory,"
Symposium on Recursive Function Theory, American Mathematical Society,
New York, April 1961.
"Eliminating the Irrelevant from Mechanical Proofs," Symposium on
Experimental Arithmetic, American Mathematical Society, April 1962.
"Unsolvable Problems," Symposium on Mathematical Theory of Automata,
New York, April 1962.
"First Order, Second Order, and Higher Order Logic," Association for
Symbolic Logic, Washington, D.C., December 1963.
"Diophantine Equations and Recursive Sets," and "Recursive Functions -
An Introduction," NATO Advanced Study Institute on Automata, Ravello,
Italy, June 1964.
"Computability," Symposium on System Theory, New York 1965.
"One Equation to Rule Them All," New York Academy of Sciences, March
1968.
"Hilbert's Tenth Problem," London Mathematical Society, London, England,
March 1969.
"Speed-up Theorems and Diophantine Equations," Courant Computer
Science Symposium on Computational Complexity, New York, October 1971.
"The Unsolvabilty of Hilbert's Tenth Problem," Joint meeting of American
Mathematical Society and Association for Symbolic Logic, St. Louis,
March 1972.
"Whither Mechanical Theorem-Proving?" Tagung uber automatisches
Beweisen, Mathematisches Forschungsinstitut, Oberwolfach, Germany,
January 1976.
"Three Lectures on Some Mathematical Applications of Logic: I.
Unsolvable Problems; II. Diophantine Sets; III. Nonstandard
Analysis," Earle Raymond Hedrick Lectures, Mathematical Association of
America, Toronto, August 1976.
"Takeuti Models and the Foundations of Quantum Mechanics," Symposium
on Infinitesimals, Iowa City, May 1977.
"Boolean-Valued Models in Set Theory, Analysis, and Quantum Mechanics,"
Chauvenet Symposium, Mathematical Association of America, Atlanta,
January 1978.
"What is a Computation," Symposium on Mathematics Today, American
Association for the Advancement of Science, Houston, January 1979.
"The Prehistory and Early History of Automated Deduction," Fourth
Workshop on Automated Deduction, Austin, February 1979.
"Why Didn't Godel Have Church's Thesis?" Recursion Theoretic Aspects of
Computer Science, Purdue University, May 1981.
"Diophantine Representation of Arithmetic Propositions," New York
Academy of Sciences, March 1982.
"Formal Proof and Mathematical Practice," Symposium on New Kinds of
Mathematical Proof (with Hilary Putnam, K.I. Appel, Stephen Cook, Marvin
Minsky, and Marshall Stone), American Philosophical Association,
Philadelphia, April 1982.
"Logic and Computation," Symposium on Philosophy of Computation, Brown
University, November 1982.
"Relations between Mathematical Logic and Computer Science," Symposium
on Mathematical Logic, Research Institute of Mathematical Sciences,
University of Kyoto, Japan, October 1983.
"Teaching the Incompleteness Theorem," Panel Discussion on
Teaching of Logic, Logic Colloquium '88, Padua, Italy, August
1988.
"Trends in Logic: Relations with Computer Science," Panel
Discussion on Trends in Logic, Logic Colloquium '88, Padua,
Italy, August 1988.
Doctoral Dissertations Supervised
1. Eric Wagner, "Uniformly Reflexive Structures: Towards an Abstract
Theory of Computability," Columbia University, 1963.
2. Robert DiPaola, "On Pseudo-Complements of Recusively Enumerable
Sets," Yeshiva University, 1964.
3. Donald Loveland, "Recusively Random Sequences," New York University,
1964.
4. Robert Case, "Partial Predicates," Yeshiva University, 1966.
5. Martin Zuckerman, "Finite Versions of the Axiom of Choice," Yeshiva
University, 1967.
6. Saul Levy, "Computational Equivalence," Yeshiva University, 1970.
7. John Denes (now called John Grant), "Definable Automorphisms in Model
Theory," New York University, 1970.
8. Richard Gostanian, "The Next Admissable Ordinal," New York
University, 1971.
9. Donald Perlis, "Ackermann's Set Theory and Related Topics," New York
University, 1972.
10. Daniel Gogol, "Models of Formulas in Various Languages," Yeshiva
University, 1973.
11. Keith Harrow, "Sub-Elementary Classes of Functions and Relations,"
New York University, 1973.
12. William Gewirtz, "Investigations in the Theory of Descriptive
Complexity," New York University, 1974.
13. Edward Schwartz, "Existential Definability in Terms of Some
Quadratic Functions," Yeshiva University, 1974.
14. Barry Jacobs, "Alpha Computational Complexity," New York
University, 1975.
15. Richard Rosenberg, "Recusively Enumerable Images of Arithmetic
Sets," New York University, 1976.
16. Jean-Pierre Keller, "Abstract Forcing and Applications," New York
University, 1977.
17. Allen Goldberg, "On the Complexity of the Satisfiability Relation,"
New York University, 1979.
18. Moshe Koppel, "Bases of Recursively Enumerable Relations," New York
University, 1980.
19. Ron Sigal, "Undecidable Complexity Statements in a Hierarchy of
Extensions of Primitive Recursive Arithmetic," New York University,
1983.
20. Elia Weixelbaum, "Formal Languages with Oracles," New York
University, 1983.
21. Eugenio Omodeo, "Decidability and Proof Procedures for Set Theory
with a Choice Operator," New York University, 1984.
PUBLICATIONS AND REPORTS:
1."On the Theory of Recursive Unsolvability," Doctoral Dissertation,
Princeton University, May 1950.
2."On the Existence of Certain Orderings in the Plane," Control Systems
Laboratory, University of Illinois, Report I-28, May 1951.
3."Arithmetical Problems and Recursively Enumerable Predicates," Journal
of Symbolic Logic, vol. 18(1953), pp. 33-41.
4."Mathematical Procedures for Decision Problems: A Program for
Presburger's Algorithm for Additive Number Theory on the Institute for
Advanced Study Digital Computer," Office of Ordnance Research Project
1333, Institute for Advanced Study, Princeton, October 1954.
5."A Note on Universal Turing Machines," Automata Studies, C.E. Shannon
and J. McCarthy, editors, Annals of Mathematics Studies, Princeton
University Press, 1956.
6."The Definition of Universal Turing Machine," Proceedings of the
American Mathematical Society, vol.8(1957), pp. 1125-1126.
7.Computability and Unsolvability, McGraw-Hill, New York 1958; reprinted
with an additional appendix, Dover 1983.
8.(with Hilary Putnam) "Reductions of Hilbert's Tenth Problem," Journal
of Symbolic Logic, vol.23(1958), pp. 183-187.
9.(with Hilary Putnam) "Feasible Computational Methods in the
Propositional Calculus," Rensselaer Polytechnic Institute, October 1958.
10.(with Hilary Putnam) "A Computational Proof Procedure; A Finitely
Axiomatizable System for Elementary Number Theory; On Hilbert's Tenth
Problem," AFOSR Report TR59-124, Rensselaer Polytechnic Institute,
October 1959.
11."A Program for Presburger's Algorithm," Summaries of Talks Presented
at the Summer Institute for Symbolic Logic, Cornell University, 1957,
Institute for Defense Analyses, 1960, pp. 215-223; reprinted in, Siekmann,
Jorg and Graham Wrightson (eds), Automation of Reasoning, vol. 1,
Springer Verlag, 1983, pp. 41-48.
12."Computable Functionals of Arbitrary Finite Type," Summaries of Talks
Presented at the Summer Institute for Symbolic Logic, Cornell
University, 1957, Institute for Defense Analyses, 1960, pp. 242-246.
13.(with Hilary Putnam) "A Computing Procedure for Quantification
Theory," Journal of the Association for Computing Machinery, vol.7(1960),
pp. 201-215; reprinted in, Siekmann, Jorg and Graham Wrightson (eds),
Automation of Reasoning, vol. 1, Springer Verlag, 1983, pp. 125-139.
14.(with Hilary Putnam and Julia Robinson) "The Decision Problem for
Exponential Diophantine Equations," Annals of Mathematics, vol.74(1961),
pp. 425-436.
15."Aspects of Mechanical Theorem-Proving," Proceedings of Third
International Congress on Cybernetics, Namur, Belgium, 1961, pp. 415-418.
16.(with George Logemann and Donald Loveland) "A Machine Program for
Theorem Proving," Communications of the Association for Computing
Machinery, vol.5(1962), pp. 394-397; reprinted in, Siekmann, Jorg and Graham
Wrightson (eds), Automation of Reasoning, vol. 1, Springer Verlag, 1983,
pp. 267-270.
17."Unsolvable Problems: A Review," Proceedings of the Symposium on
Mathematical Theory of Automata, 1962, pp. 15-22.
18."Applications of Recursive Function Theory to Number Theory,"
Proceedings of Symposia in Pure Mathematics, vol. 5(1962), pp. 135-138.
19.(with Hilary Putnam) "Diophantine Sets over Polynomial Rings," Illinois
Journal of Mathematics, vol.7(1963), pp. 251-255.
20."Extensions and Corollaries of Recent Work on Hilbert's Tenth
Problem," Illinois Journal of Mathematics, vol.7(1963), pp. 246-250.
21."Eliminating the Irrelevant from Mechanical Proofs," Proceedings of
Symposia in Applied Mathematics, vol.15(1963), pp. 15-30. reprinted in,
Siekmann, Jorg and Graham Wrightson (eds), Automation of Reasoning, vol.
1, Springer Verlag, 1983, pp. 315-330.
22.(with T.J. Chinlund, P.G. Hinman, and M.D. McIlroy) "Theorem-Proving by
Matching," Bell Telephone Laboratories, June 1964.
23.(editor) The Undecidable, Raven Press 1965.
24."Recursive Functions - An Introduction," Automata Theory, E.R.
Caianello, editor, Academic Press,1966, pp. 153-163.
25."Diophantine Equations and Recursively Enumerable Sets," Automata
Theory, E.R. Caianello, editor, Academic Press,1966, pp. 146-152.
26."Computability," Proceedings of the Symposium on System Theory,
Brooklyn, N.Y. 1966, pp. 127-131.
27.Lectures on Modern Mathematics, Gordon and Breach, 1967.
28.First Course in Functional Analysis, Gordon and Breach, 1967.
29."Recursive Function Theory," Encyclopedia of Philosophy, Paul
Edwards, editor, Macmillan and Free Press, 1967, vol.7, pp. 89-95.
30."One Equation to Rule Them All," Rand Memorandum, RM-5495-PR,
February 1968.
31."One Equation to Rule Them All," Transactions of the New York
Academy of Sciences, Sec. II, vol.30(1968), pp. 766-773.
32."An Explicit Diophantine Definition of the Exponential Function,"
Communications on Pure and Applied Mathematics, vol.24(1971), pp. 137-145.
33.(with Reuben Hersh) "Nonstandard Analysis," Scientific American,
vol.226(1972), pp. 78-86.
34."On the Number of Solutions of Diophantine Equations," Proceedings
of the American Mathematical Society, vol.35(1972), pp. 552-554.
35."Hilbert's Tenth Problem is Unsolvable," American Mathematical
Monthly, vol.80(1973), pp. 233-269; reprinted in Davis, Martin,
Computability and Unsolvability, Dover 1983.
36.(with Reuben Hersh) "Hilbert's Tenth Problem," Scientific American,
vol.229(1973), pp. 84-91; reprinted in Abbott, J.C. (ed.) The Chauvenet
Papers, vol. 2, pp. 555-571, Math. Assoc. America, 1978.
37.(with N. Z. Shapiro) "Uncrackable Data Banks," Rand Report R-1382-NSF,
November 1973.
38."Speed-up Theorems and Diophantine Equations," Computational
Complexity, Randall Rustin, editor, Algorithmics Press, 1973, pp. 87-95.
39.(with Yuri Matijasevic and Julia Robinson) "Hilbert's Tenth Problem:
Diophantine Equations: Positive Aspects of a Negative Solution,"
Proceedings of Symposia in Pure Mathematics, vol.28(1976), pp. 323-378.
40.Applied Nonstandard Analysis, Interscience-Wiley, 1977.
41.(with Rona Gurkewitz and David L. Yarmush) "LOGIK, A Special Purpose
Language for Writing Theorem-Provers," Courant Institute Report
IMM-413, October 1976.
42."Minimal Entailment," Stanford University Artificial Intelligence
Laboratory, December 1976.
43.(with J.T. Schwartz) "Extensibility of Verifiers" in Courant Institute
Computer Science Report NSO-12, October 1977.
44."Unsolvable Problems," Handbook of Mathematical Logic, Jon Barwise,
editor, North-Holland, 1977, pp. 567-594.
45."A Relativity Principle in Quantum Mechanics," International Journal
of Theoretical Physics, vol.16(1977), pp. 867-874.
46."What is a Computation?" Mathematics Today: Twelve Informal Essays,
L. A. Steen, editor, Springer-Verlag, 1978, pp. 241-267.
47.(with J.T. Schwartz) "Metamathematical Extensibility for Theorem
Verifiers and Proof-Checkers," Computers and Mathematics with
Applications, vol.5(1979), pp. 217-230.
48."Notes on the Mathematics of Non-Monotonic Reasoning," Artificial
Intelligence, vol.13(1980), pp. 73-80.
49."Obvious Logical Inferences," Proceedings of the Seventh Joint
International Congress on Artificial Intelligence, 1981, pp. 530-531.
50.(with Elaine J. Weyuker) "Pseudo-Oracles for Non-Testable Programs,"
ACM '81 Conference Proceedings, pp. 254-257.
51."Why Godel Didn't Have Church's Thesis," Information and Control,
vol.54 (1982), pp. 3-24.
52."Lectures at 'Atlanta State'," Annals of the History of
Mathematics, vol.4(1982), pp. 370-371.
53."The Prehistory and Early History of Automated Deduction," Siekmann,
Jorg and Graham Wrightson (eds), Automation of Reasoning, vol. 1,
Springer Verlag, 1983, pp. 1-28.
54.(with Elaine J. Weyuker) Computability, Complexity, and Languages,
Academic Press, 1983.
55.(with Elaine J. Weyuker) "A Formal Notion of Program-Based Test Data
Adequacy," Information and Control, vol.56(1983), pp. 52-71.
56."Church's Thesis," Encyclopedia of Artificial Intelligence, John Wiley,
1987.
57."Mathematical Logic and the Origin of Modern Computers," Studies
in the History of Mathematics, pp. 137-165. Mathematical
Association of America, 1987. Reprinted in The Universal Turing
Machine - A Half-Century Survey, Rolf Herken, editor, pp. 149-174.
Verlag Kemmerer & Unverzagt, Hamburg, Berlin 1988; Oxford University
Press, 1988.
58.(with Elaine J. Weyuker) "Metric Space Based Test Data Adequacy
Criteria," The Computer Journal, vol. 31(1988), pp. 17-24.
59."Influences of Mathematical Logic on Computer Science," in The
Universal Turing Machine - A Half-Century Survey, Rolf Herken,
editor, pp. 315-326. Verlag Kemmerer & Unverzagt, Hamburg, Berlin
1988; Oxford University Press, 1988.
∂19-Oct-88 0920 ARK CSD-CF Rate Idea
To: ball@POLYA.STANFORD.EDU
CC: ARK@SAIL.Stanford.EDU, facil@SCORE.STANFORD.EDU,
wiederhold@SUMEX-AIM.Stanford.EDU, JMC@SAIL.Stanford.EDU
This note is about scalable costs. It is relatively easy to expand the
amount of disk space available, while relatively difficult to expand the
amount of CPU time available. We can easily wheel in another disk drive
when we run out of disk space, but getting another CPU is an expensive
solution. Since people can buy their own disk drives, it appears that the
revenue base for CSD-CF is well protected by encouraging users to use
CSD-CF disk space by pricing it low enough to compete with the incremental
cost of providing additional disk space. The costs of running each
computer that is not recovered through storage charges and connect charges
can be recovered through CPU charges alone, and this will tend to
discourage overuse of the CPU and reduce the likelihood that a new CPU is
needed.
Here is an example of how the rates could be computed. I'm using numbers
that may not approximate the actual numbers, but you can try them with the
"real" numbers and see how that works out.
1. Take the capacity of a disk drive 600 megabytes
2. Take one-half of #1 300 megabytes
(On the average, the new disk drive will be half full.)
3. Convert to megabits, since that's the charging scheme
2400 megabits
4. Take the cost of one disk drive $18,000
5. Divide #4 by #3 $7.50/megabit
6. Divide this by 12 $0.625/megabit
The value in #6 is the suggested cost per megabit per month. It is
set to recover the cost of a disk drive in one year. Since capital
equipment is depreciated over 5 years, the remainder is available to
pay for backup and other costs of running that computer.
7. Take the cost of one disk drive (#4 above) $18,000
8. Divide by 60 to get monthly depreciation $300
9. Multiple #6 by #3 above to get monthly income $1,500
10. Amount used for backup and other costs $1,200
for the newest (half full) disk drive
11. Double #9 $3,000
(This is income for next to last disk drive)
12. Subtract #8 from #11 $2,700
Amount used for backup and other costs
I've made the assumption that the new disk drive is, on the average, half
full with chargeable data, and that there is a fixed amount of system
files which are on the first disk drives that comes with the CPU that are
considered part of the cost of the CPU. (That is, they are a fixed cost,
while disk space is a variable cost.)
------
Let me make a separate proposal for consideration. Is CSD-CF willing to
trade charges for capital equipment? Suppose I agree to buy for CSD-CF an
$18,000 disk drive. Can I get a credit for $18,000 worth of CSD-CF usage
charges? This is how SAIL used to pay for expansion before there were
usage charges and a cost center. The reason PI's might want to do this is
that $18,000 worth of usage charges costs an ordinary grant over $31,000
with overhead, while an $18,000 disk drive costs the same grant only
$18,000 (no overhead). The advantage to CSD-CF is that it does not have
to invest unrestricted funds that only get recovered over 5 years.
Effectively, this is a way to recover costs of capital equipment in one
year. The remaining 4 years are then "gravy".
Arthur
∂19-Oct-88 1039 VAL kr'89 panel
∂19-Oct-88 0753 ether%allegra@research.att.com kr'89 panel
Received: from research.att.com (ATT.ARPA) by SAIL.Stanford.EDU with TCP; 19 Oct 88 07:53:00 PDT
From: ether@allegra.att.com
Date: Wed, 19 Oct 88 10:25:30 edt
>From: allegra!ether (David Etherington)
To: arpa!VAL@SAIL.Stanford.EDU
In-Reply-To: Vladimir Lifschitz's message of 18 Oct 88 1316 PDT <v4qZv@SAIL.Stanford.EDU>
Subject: kr'89 panel
Vladimir:
You are right about the title. The intention is not
to suggest that the field is dead, but rather to try
to address some of the criticisms of the paradigm.
Specifically, these include McDermott's "Critique of
Pure Reason", the "Why not use Probablility?" position,
and Ken Forbus' comment (as Commonsense reasoning area
chair on the AAAI-88 panel), that "Last time I looked,
the nonmonotonic reasoning people were still talking
about tweety, ...", suggesting that there is no reason
to believe that the field will ever get beyond toy
problems. There are also problems of dealing with
inconsistent theories, and tractability questions.
I remain optomistic (my CSCSI invited talk, for example,
was a catalogue of some of these "woes", and my perception
of ways that they might be overcome within the paradigm
of nonmonotonic reasoning), but I think it could be
stimulating and informative to have some people raise
these questions, and have some well-thought-out replies.
I think that the idea I have in mind is much like what you
suggested, except taking the "challenges" from outside as
well as inside the core of people working in the field.
I had hoped to be able to circulate a prospectus among
the panel members, consisting of a summary of each person's
main theme(s), allowing everyone to make sure that they
are thinking about the right questions and to prepare
refutations/solutions in advance.
I think it is important for us to have you on the panel,
so I'm willing to accommodate your input to a large degree.
How does "Challenges for Nonmonotonic Reasoning" (or something
like that) sound as a title, with something like the outline
I gave above, sound?
Thanks for your ideas!
...David.
Replying-To: ether@allegra.att.com
Reply-Subject: re: kr'89 panel
Reply-Text:
[In reply to message sent Wed, 19 Oct 88 10:25:30 edt.]
David,
Thanks for your reply. I think I understand your idea better now, and I'll
be glad to take part in the panel.
--Vladimir
∂19-Oct-88 1113 meyer@theory.lcs.mit.edu hello
Received: from stork.LCS.MIT.EDU by SAIL.Stanford.EDU with TCP; 19 Oct 88 11:13:25 PDT
Received: by stork.LCS.MIT.EDU
id AA01242; Wed, 19 Oct 88 14:13:12 EDT
Date: Wed, 19 Oct 88 14:13:12 EDT
From: meyer@theory.lcs.mit.edu
Message-Id: <8810191813.AA01242@stork.LCS.MIT.EDU>
To: JMC@sail.stanford.edu
In-Reply-To: John McCarthy's message of 19 Oct 88 1015 PDT <D4Y0X@SAIL.Stanford.EDU>
Subject: hello
Date: 19 Oct 88 1015 PDT
From: John McCarthy <JMC@sail.stanford.edu>
[In reply to message sent Mon, 17 Oct 88 14:32:13 EDT.]
Could someone make me a hotel reservation convenient to the celebration,
Tuesday and Wednesday nights?
-----------------
I've asked the Celebration secretary, Barbara Wollan, to make your
reservation. She'll get back to you. Regards, A.
∂19-Oct-88 1121 ball@polya.Stanford.EDU CSD-CF Rate Idea
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Oct 88 11:21:49 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA06534; Wed, 19 Oct 88 11:22:00 PDT
Date: Wed, 19 Oct 88 11:22:00 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
Message-Id: <8810191822.AA06534@polya.Stanford.EDU>
To: ARK@SAIL.Stanford.EDU
Cc: ARK@SAIL.Stanford.EDU, facil@SCORE.STANFORD.EDU,
wiederhold@SUMEX-AIM.Stanford.EDU, JMC@SAIL.Stanford.EDU
In-Reply-To: Arthur Keller's message of 19 Oct 88 0920 PDT <P4Xsx@SAIL.Stanford.EDU>
Subject: CSD-CF Rate Idea
Arthur,
An interesting idea. The current problem is that the smallest element of
cost has been assigned to the whole system cost base, not looked at
incrementally. This tends to cause the newest additions to be "loaded"
with the old base, this is the case with SAIL and all other systems.
I'm currently working on an analysis of the present model with an eye
toward looking at the incremental costs. It may result in a much higher
charge for the non-expandible resources, like CPU's. I have been concerned
with making major changes to the charging structure. The recent reduction
of connect time charges was the first major change. I knew that I could
defend a change like that since the communications hardware no longer
has the constraint of a port per user with Ethernet. It is not quite
as easy to explain a change of this type, even though it appears that
we may have to.
Thanks for the input
-Jim
∂19-Oct-88 1149 ELLIOTT%SLACVM.BITNET@forsythe.stanford.edu Re: bicycling
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 19 Oct 88 11:48:58 PDT
Received: by Forsythe.Stanford.EDU; Wed, 19 Oct 88 11:49:47 PDT
Date: 19 Oct 88 11:47 PST
From: ELLIOTT%SLACVM.BITNET@Forsythe.Stanford.EDU
To: JMC @ SAIL.STANFORD.EDU
Subject: Re: bicycling
Date: 19 October 1988, 11:46:42 PST
From: Bloom, Elliott (415)-926-2469 ELLIOTT at SLACVM
To: JMC at SAIL.STANFORD.EDU
Subject: Re: bicycling
In-Reply-To: JMC AT SAIL.STANFORD.EDU -- 10/19/88 02:43
Dear John,
Yes, we will go on Sat. at about 9:30am. Happy to have you
come along if you wish.
Greetings,
Elliott
∂19-Oct-88 1246 GLB
To: JMC@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU, jcm@POLYA.Stanford.EDU,
CLT@SAIL.Stanford.EDU
∂17-Oct-88 1842 GLB oral exams
To: JMC@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU, jcm@POLYA.Stanford.EDU,
CLT@SAIL.Stanford.EDU
I think I should take my orals.
Would it be good for everyone a month from today,
Thursday 17 November (or something like that)?
Gianluigi
-----
∂18-Oct-88 1314 JK
Fine with me. I teach Tuesday and Thursday mornings.
-----
∂17-Oct-88 2205 CLT oral exam
John and I will be away from Nov 7 through 19
Also on Thursdays Timothy would have to come too!
What about sometime the week of Nov 21?
-----
∂18-Oct-88 1637 jcm@ra.stanford.edu Re: oral exams
Offhand, Nov 17 seems OK to me.
-----
∂18-Oct-88 2345 sf@csli.Stanford.EDU
Wednesday afternoons are good for me, and Tu-Th after 2:30. The 18th
would have been OK, but since that's a problem for JMC, also the week of
the 21st would be OK, say the 22nd or 23d. Or the first week in Dec.
********
From the above messages it looks like Nov. 22 or 23 should be good.
As external member, I thought of Ivan Sag (Linguistics) as someone
who might have some interest in the topic.
Gianluigi
∂19-Oct-88 1322 CLT qlisp for okuno
Date: 19 Oct 88 1303 PDT
From: Dick Gabriel <RPG@SAIL.Stanford.EDU>
Subject: Okuno
To: pehoushe@GANG-OF-FOUR.STANFORD.EDU
DARPA has vetoed sending Qlisp outside the US. Please do not send a tape.
You can send my regrets if you wish.
-rpg-
Given that DARPA seems to not balk at getting mileage out of the
prize the Japanese have awarded you, do you think we have a
basis for complaint at there not allowing us to send a tape to Okuno?
∂19-Oct-88 1329 kar@polya.Stanford.EDU Applications AI comp.
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Oct 88 13:29:03 PDT
Received: from LOCALHOST by polya.Stanford.EDU (5.59/25-eef) id AA16693; Wed, 19 Oct 88 13:29:11 PDT
Message-Id: <8810192029.AA16693@polya.Stanford.EDU>
To: jmc@sail
Subject: Applications AI comp.
Date: Wed, 19 Oct 88 13:29:01 -0700
From: kar@polya.Stanford.EDU
On behalf of the Applications Comp. Committee I am writing to you
to solicit questions for the AI section of the Applications Comp.
which is to be held next January.
The text for this section is the whole of Elaine Rich's book
"Artificial Intelligence". (Note that there is a separate AI
section on the Theory comp. based on Genesereth and Nilsson's book.)
You are invited to submit *one question* (or more if you feel
enthusiastic) by November 7th. Please give hard copies to either
myself or Prof. Wiederhold; electronic copies can be sent to
wiederhold@sumex. An accompanying sketch solution would also
be appreciated, as the solutions are being prepared alongside the
exam for this comp. Please ensure that the questions remain
confidential.
Thank you for your cooperation.
Ken Ross.
∂19-Oct-88 1333 pimeet@vax.darpa.mil PI Meeting
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 19 Oct 88 13:33:13 PDT
Posted-Date: Wed 19 Oct 88 13:46:03-EDT
Received: by vax.darpa.mil (5.54/5.51)
id AA08113; Wed, 19 Oct 88 13:46:05 EDT
Date: Wed 19 Oct 88 13:46:03-EDT
From: Mark Pullen <PIMEET@vax.darpa.mil>
Subject: PI Meeting
To: ISTO-PI-LIST@vax.darpa.mil
Message-Id: <593286363.0.PIMEET@VAX.DARPA.MIL>
Mail-System-Version: <VAX-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Ladies and Gentlemen:
Just a reminder that the cut-off for hotel reservations at the Hyatt is
this Friday, October 21. Please be sure to take care of this soon.
Also, for those of you who were told that there were no rooms available
at the Hyatt and subsequently booked the Hilton, please feel free to
change your reservations to the Hyatt. There really is plenty of room!
Hyatt Reservations: 214/453-8400.
Juanita Walton
-------
∂19-Oct-88 1406 CLT qlisp for okuno
Even better, to whom should we make the argument?
Would it be worth discussing this with Goto and his NTT
colleagues when they are here?
∂19-Oct-88 1413 ginsberg@polya.Stanford.EDU Formfeed to meet tomorrow -- don't forget!
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Oct 88 14:13:27 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA20753; Wed, 19 Oct 88 14:12:18 PDT
Date: Wed, 19 Oct 88 14:12:18 PDT
From: Matthew L. Ginsberg <ginsberg@polya.Stanford.EDU>
Message-Id: <8810192112.AA20753@polya.Stanford.EDU>
To: feed@polya.Stanford.EDU
Subject: Formfeed to meet tomorrow -- don't forget!
See you there; 12.15 as usual.
Matt
∂19-Oct-88 1640 MPS
Mr. Suzuki phoned from Tokyo. He would like you to call
regarding your upcoming trip.
He gave me his office number 265-4232. The area code is 3
Pat
∂19-Oct-88 1642 MPS Dallas
There is a Delta flight to SFO
7-day advance w/25% penalty 376, otherwise full fare 476
American to San Jose is same deal, but prices are different
418 with penalty and 494 full fare. Both are coach.
Do you want either one of them?
Pat
∂19-Oct-88 1644 MPS book
Please look for the book Abstract Objects. Thanks
Pat
P.S. I hope I have not forgotten anything
Pat
∂19-Oct-88 1707 meyer@theory.LCS.MIT.EDU no reservation needed
Received: from theory.LCS.MIT.EDU by SAIL.Stanford.EDU with TCP; 19 Oct 88 17:07:10 PDT
Received: by theory.LCS.MIT.EDU
id AA08961; Wed, 19 Oct 88 20:06:53 EDT
Date: Wed, 19 Oct 88 20:06:53 EDT
From: meyer@theory.LCS.MIT.EDU (Albert R. Meyer)
Message-Id: <8810200006.AA08961@theory.LCS.MIT.EDU>
To: JMC@sail.stanford.edu
Cc: mac25-registration@xx.lcs.mit.edu, wollan@theory.LCS.MIT.EDU
In-Reply-To: John McCarthy's message of 19 Oct 88 1526 PDT <1r4yTu@SAIL.Stanford.EDU>
Subject: no reservation needed
Date: 19 Oct 88 1526 PDT
From: John McCarthy <JMC@sail.stanford.edu>
[In reply to message sent Wed, 19 Oct 88 14:13:12 EDT.]
I tracked down Fredkin in Japan, and I will be staying with him,
so I won't need a hotel reservation after all.
---------
ok. Regards, A.
∂19-Oct-88 2024 @Gang-of-Four.Stanford.EDU:arg@lucid.com Talk on Parallel Processing and the Butterfly Computer
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Oct 88 20:24:24 PDT
Received: from LUCID.COM by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04623; Wed, 19 Oct 88 20:24:17 PDT
Received: from bhopal ([192.9.200.13]) by LUCID.COM id AA07514g; Wed, 19 Oct 88 20:23:48 PDT
Received: by bhopal id AA02813g; Wed, 19 Oct 88 20:22:15 PDT
Date: Wed, 19 Oct 88 20:22:15 PDT
From: Ron Goldman <arg@lucid.com>
Message-Id: <8810200322.AA02813@bhopal>
To: qlisp@go4.stanford.edu
Subject: Talk on Parallel Processing and the Butterfly Computer
Academic Information Resources will sponsor a talk on "The History of Parallel
Processing Architectures and the Butterfly Computer" on Tuesday, October 25
from 3:30 p.m. to 5:00 p.m. in Tresidder Oak Lounge East. The speaker will
be Dr. W. B. Barker of BBN Advanced Computers.
Barker's talk will offer an historical perspective on parallel
processing architectures using BBN's own experience as an
illustration. The talk will cover BBN's development of a bus-based
multiprocessor based on Lockheed Sue minicomputers and show the
migration of that architecutre to the Butterfly parallel processor.
The Butterfly currently supports an implementation of the Mach
operating system developed by Carnegie Mellon University. Programming
with Mach on the Butterfly as well as several other parallel
programming paradigms will be discussed. The current Butterfly will
be used as a point of departure to discuss future architectures that
implement new technologies to improve processor and interconnection
network performance.
Dr. Barker is BBN's Senior Vice President for Business Development.
He is responsible for the identification and development of strategic
opportunities for the company, principally in technology transfer from
BBN Laboratories to BBN's product activities. As design engineer on
the ARPANet, he was responsible for the installation of the world's
first packet switch. Barker was the principal architect of the first
commercial parallel processor, Pluribus, and holds two patents in
computer architecture.
∂20-Oct-88 0858 ELLIOTT%SLACVM.BITNET@forsythe.stanford.edu re: bicycling
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 20 Oct 88 08:58:15 PDT
Received: by Forsythe.Stanford.EDU; Thu, 20 Oct 88 08:59:02 PDT
Date: 20 Oct 88 08:57 PST
From: ELLIOTT%SLACVM.BITNET@Forsythe.Stanford.EDU
To: JMC @ SAIL.STANFORD.EDU
Subject: re: bicycling
Date: 20 October 1988, 08:57:24 PST
From: Bloom, Elliott (415)-926-2469 ELLIOTT at SLACVM
To: JMC at SAIL.STANFORD.EDU
Subject: re: bicycling
In-Reply-To: JMC AT SAIL.STANFORD.EDU -- 10/19/88 11:51
Yes.
∂20-Oct-88 1054 pimeet@vax.darpa.mil PI Meeting
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 20 Oct 88 10:54:39 PDT
Posted-Date: Thu 20 Oct 88 13:09:00-EDT
Received: by vax.darpa.mil (5.54/5.51)
id AA10914; Thu, 20 Oct 88 13:09:03 EDT
Date: Thu 20 Oct 88 13:09:00-EDT
From: Mark Pullen <PIMEET@vax.darpa.mil>
Subject: PI Meeting
To: makhoul@bbn.com, wilensky@ucbvax.berkeley.edu, gclhbx@gcuxb.att.com,
Forsdick@bbn.com, cbd@cs.utah.edu, tenenbaum@spar.slb.com,
kanade@ius3.ius.cs.cmu.edu, weems@cs.umass.edu,
ht.Kung@n.sp.cs.cmu.edu, MILLER@bbn.com, rick.rashid@cs.cmu.edu,
POSTEL@venera.isi.edu, Arvind@xx.lcs.mit.edu, BALZER@isi.edu,
mike@postgres.Berkeley.EDU, GURFIELD@isi.edu,
wedlake@shrike.austin.lockheed.com, jmc@sail.stanford.edu,
sincovec@prandtl.nas.nasa.gov
Message-Id: <593370540.0.PIMEET@VAX.DARPA.MIL>
Mail-System-Version: <VAX-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Gentlemen:
As speakers for the upcoming DARPA/ISTO 1988 Principal Investigators'
Meeting, you are requested to provide a one-page presentation abstract
which will be included in the published Meeting Notes. I apologize
for the short notice but I must send camera-ready copy to the printer
by next week. If you will simply create an outline of your presentation
and e-mail it (pimeet@vax.darpa.mil) by noon, Tuesday, Oct. 25, I will
be able to include it in the publication.
Thanks for your cooperation in this matter.
Juanita Walton
-------
∂20-Oct-88 1203 pehoushe@Gang-of-Four.Stanford.EDU How to ise NSTACK
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 12:03:16 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA06752; Thu, 20 Oct 88 12:03:00 PDT
Date: Thu, 20 Oct 88 12:03:00 PDT
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8810201903.AA06752@Gang-of-Four.Stanford.EDU>
To: qlisp@Gang-of-Four.Stanford.EDU
Subject: How to ise NSTACK
This is the /qlisp/ql-nusage.text file. It explains how to start up
and use the NSTACK scheduler for Qlisp (on GO4).
To start NSTACK, get into QLISP (not new-qlisp). Then (load
"/qlisp/ql-nstack") and then (NSTACK). It takes a minute or so to
start up.
Warnings:
Code must be recompiled, after NSTACK is loaded, to work in NSTACK.
Interpreted code is not supported. GC while in parallel is shaky at
best. An error while in parallel may be hazardous. To recover, if
the system is nice enough to prompt you, do a :A, to abort to
top-level, followed by an (NQL-INIT T). This recovery procedure is
not gauranteed to keep your Lisp intact, but it does usually work.
Caveat Emptor:
The system is fast, but Fragile. Spawning a task costs
about 1/3 of a floating point multiply; 2/3 of a closure creation; 3
single argument simple function calls, et cetera. QCATCH and QTHROW
are not yet supported.
The supported parallelism primitives are QLET, QDOTIMES,
TDOTIMES, #!(symbol form1 form2) and #?(symbol form1 form2).
A summary of each primitive:
(QLET prop ((var1 form1) ... (varn formn)) body)
The EAGER version of prop is not supported. Prop acts in the usual
way. If Prop is T, Processes are always created, regardless of serial
are paralell mode. When Prop is T, Processes are spawned to evaluate
form2 ... formn. These forms must be function calls; the arguments to
these function calls are evaluated at spawn time. form1 can be an arbitrary
lisp form. It is evaluated after spawning the other n-1 processes.
(QDOTIMES (var howmany &optional result-form) body)
(TDOTIMES (var howmany &optional result-form) body)
These are parallel forms of the serial iterative form, DOTIMES. As
long as the individual iterations don't depend on other iterations,
these parallelism control structures are quite useful. There is no
restriction on the bodies of these forms, other than independence.
They may be nested, but compilation tends to take a while if the
nested bodies are large (or small, for that matter). QDOTIMES uses
the dynamic predicate, while TDOTIMES always spawns a task. The
number of iterations that can be effectively handled is roughly 2↑400.
#! and #?:
;;; Always Parallel
(defun nfib (n)
(if (< n 2) n
#!(+ (nfib (- n 1))
(nfib (- n 2)))))
;;; May Be Parallel.
(defun bfib (n)
(if (< n 2) n
#?(+ (bfib (- n 2))
(bfib (- n 1)))))
Note: The current syntax requires that FORM2 be a function call.
The Top-Level entry to parallelism is called CPU.
The syntax: (CPU Form-With-Parallelism Integer), where Integer is the
number of times you want to run the form. If Integer is omitted, it
defaults to 10. Do not Ctrl-C out of Parallel mode.
For example,
> (cpu (nfib 25) 5)
ParaCpu: 839 Spawns:121392
ParaCpu: 838 Spawns:121392
ParaCpu: 843 Spawns:121392
ParaCpu: 882 Spawns:121392
ParaCpu: 840 Spawns:121392
#P:8 (NFIB 25)
CPU (min mean stddev): 838 848.4 16.9
Spawn (min mean stddev):121392 121392.0 0.0
NIL
> (cpu (bfib 25) 5)
ParaCpu: 368 Spawns: 3330
ParaCpu: 366 Spawns: 3026
ParaCpu: 358 Spawns: 2990
ParaCpu: 361 Spawns: 3365
ParaCpu: 354 Spawns: 2140
#P:8 (BFIB 25)
CPU (min mean stddev): 354 361.4 5.1
Spawn (min mean stddev): 2140 2970.2 442.3
NIL
>
Using the TIME function:
If you want to evaluate Parallel Code serially, from top-level, use
the system's TIME function. Code written with #! will still spawn
tasks, but all tasks run on 1 processor. Code written with #? will
not spawn any tasks when using the Time function.
MORE WARNINGS:
Don't CTRL-C while in parallel mode, Don't Garbage Collect in
Parallel mode (make frequent calls to GC from Top-Level if you
experiments CONS alot),
In general, any error while in parallel mode may cause irreparable
damage. If you get an error, and the system is nice enough to give a
prompt back, it may be best to (QUIT). If you don't get a prompt
back, do a "kill -9 qlisp-pid" from Unix.
There are other features available, such as cheap Locks, and
Condition-Memory...
If you have difficulties, let me know.
Dan Pehoushek
pehoushek@Gang-Of-Four
∂20-Oct-88 1229 @Gang-of-Four.Stanford.EDU,@IGNORANT.Stanford.EDU:rivin@GANG-OF-FOUR.Stanford.EDU How to ise NSTACK
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 12:29:30 PDT
Received: from Ignorant.Stanford.EDU by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA06801; Thu, 20 Oct 88 12:29:12 PDT
Date: Thu, 20 Oct 88 12:26 PDT
From: Igor Rivin <rivin@GANG-OF-FOUR.Stanford.EDU>
Subject: How to ise NSTACK
To: pehoushe@GANG-OF-FOUR.Stanford.EDU, qlisp@GANG-OF-FOUR.Stanford.EDU
In-Reply-To: <8810201903.AA06752@Gang-of-Four.Stanford.EDU>
Message-Id: <19881020192642.6.RIVIN@IGNORANT.Stanford.EDU>
Date: Thu, 20 Oct 88 12:03:00 PDT
From: Dan Pehoushek <pehoushe>
<stuff>
The number of iterations that can be effectively handled is roughly 2↑400.
<stuff>
I guess I'll see you after the next 2↑200 Big Bangs then,
dude...
∂20-Oct-88 1245 ME Boise
To: JJW@SAIL.Stanford.EDU, TD@SAIL.Stanford.EDU, DDM@SAIL.Stanford.EDU,
STU@SAIL.Stanford.EDU, JMC@SAIL.Stanford.EDU
SAIL is once again able to reach Boise. I assume someone (Joe Pallas?)
did something at the Boise end to fix it (gateway problem?).
∂20-Oct-88 1334 pehoushe@Gang-of-Four.Stanford.EDU How to use NSTACK
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 13:34:27 PDT
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA07038; Thu, 20 Oct 88 13:29:13 PDT
Date: Thu, 20 Oct 88 13:29:13 PDT
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8810202029.AA07038@Gang-of-Four.Stanford.EDU>
To: rivin@GANG-OF-FOUR.Stanford.EDU
Cc: qlisp@GANG-OF-FOUR.Stanford.EDU
In-Reply-To: Igor Rivin's message of Thu, 20 Oct 88 12:26 PDT <19881020192642.6.RIVIN@IGNORANT.Stanford.EDU>
Subject: How to use NSTACK
The Idea behind that phrase is that, when doing a balanced binary
computation, you won't run out of Process Shells (a fairly finite
resource) while spawning continuously.
This is an important Robustness feature of depth first scheduling. Dude!
-Dan
∂20-Oct-88 1455 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU:kapur@albanycs.albany.edu extension of submission deadline for RTA89
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 14:55:28 PDT
Received: from AI.AI.MIT.EDU by SCORE.STANFORD.EDU with TCP; Thu 20 Oct 88 14:54:58-PDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 20 Oct 88 17:50:49 EDT
Received: from albanycs.albany.edu (TCP 20063000404) by MC.LCS.MIT.EDU 20 Oct 88 14:40:30 EDT
Received: by albanycs.albany.edu (5.54/4.8)
id AA25962; Thu, 20 Oct 88 12:26:55 EDT
Date: Thu, 20 Oct 88 12:26:55 EDT
From: kapur@albanycs.albany.edu (Deepak Kapur)
Message-Id: <8810201626.AA25962@albanycs.albany.edu>
To: rewriting@crin.crin.fr, narrow@a.cs.uiuc.edu,
theorem-provers@mc.lcs.mit.edu
Subject: extension of submission deadline for RTA89
EXTENSION OF RTA-89 DEADLINE FOR PAPER SUBMISSION
CALL FOR PAPERS
Third International Conference on Rewriting Techniques and Applications
April 3-5, 1989, Chapel Hill, North Carolina, USA
The deadline for submission of papers for RTA-89 is extended to
October 27, 1988 for receipt of all submissions. However, late
submissions should not expect to receive notification of
acceptance/rejection by December 19, 1988. Camera-ready copies of all
acceptances are still due January 20, 1989.
Paper submission: 10 copies of a full draft paper of no more than 15
(fifteen) double-spaced pages should be RECEIVED by October 27, 1988
by the program chairman:
Nachum Dershowitz - RTA-89
University of Illinois
1304 West Springfield Ave.
Urbana, IL 61801-2987, USA
Bitnet: nachum@uiucvmd Internet: nachum@m.cs.uiuc.edu
∂20-Oct-88 1618 hoffman@csli.Stanford.EDU re: some Symbolic System Forums Announcements
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 16:18:18 PDT
Received: by csli.Stanford.EDU (4.0/4.7); Thu, 20 Oct 88 16:19:52 PDT
Date: Thu 20 Oct 88 16:19:51-PDT
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: re: some Symbolic System Forums Announcements
To: JMC@SAIL.STANFORD.EDU
Message-Id: <593392791.0.HOFFMAN@CSLI.Stanford.EDU>
In-Reply-To: <$5sko@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
I figured that out, belatedly. Thanks! Incidentally, I was delayed from
calling you by the Forum emergencies (emergencies due to the number of
people who had to be notified and attempts at rescheduling) about the Kant
lecture changes. If it interests you, the popular guess is that David Lewis
(who write alot on phil of mind) will be dealing with a new intuitive notion
of part which is not captured by subset or member of a class relation. I
will see you tomorrow. (Today I am trying to finish the SSP essay attempting
to define/characterize symbolic systems.)
thanks
reid
p.s. nothing back from Searle yet.
-------
∂20-Oct-88 1638 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU:kapur@albanycs.albany.edu extension of submission deadline for RTA89
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 16:38:33 PDT
Received: from AI.AI.MIT.EDU by SCORE.STANFORD.EDU with TCP; Thu 20 Oct 88 16:38:05-PDT
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 20 Oct 88 19:34:22 EDT
Received: from albanycs.albany.edu (TCP 20063000404) by MC.LCS.MIT.EDU 20 Oct 88 18:07:59 EDT
Received: by albanycs.albany.edu (5.54/4.8)
id AA25962; Thu, 20 Oct 88 12:26:55 EDT
Date: Thu, 20 Oct 88 12:26:55 EDT
From: kapur@albanycs.albany.edu (Deepak Kapur)
Message-Id: <8810201626.AA25962@albanycs.albany.edu>
To: rewriting@crin.crin.fr, narrow@a.cs.uiuc.edu,
theorem-provers@mc.lcs.mit.edu
Subject: extension of submission deadline for RTA89
EXTENSION OF RTA-89 DEADLINE FOR PAPER SUBMISSION
CALL FOR PAPERS
Third International Conference on Rewriting Techniques and Applications
April 3-5, 1989, Chapel Hill, North Carolina, USA
The deadline for submission of papers for RTA-89 is extended to
October 27, 1988 for receipt of all submissions. However, late
submissions should not expect to receive notification of
acceptance/rejection by December 19, 1988. Camera-ready copies of all
acceptances are still due January 20, 1989.
Paper submission: 10 copies of a full draft paper of no more than 15
(fifteen) double-spaced pages should be RECEIVED by October 27, 1988
by the program chairman:
Nachum Dershowitz - RTA-89
University of Illinois
1304 West Springfield Ave.
Urbana, IL 61801-2987, USA
Bitnet: nachum@uiucvmd Internet: nachum@m.cs.uiuc.edu
∂20-Oct-88 1650 rick@hanauma.STANFORD.EDU re: Amazon rain forest destruction
Received: from hanauma.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 20 Oct 88 16:50:02 PDT
Date: Thu, 20 Oct 88 16:52:59 pdt
From: Richard Ottolini <rick@hanauma.STANFORD.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: re: Amazon rain forest destruction
Cc: su-etc@score
You know that well-intentioned, but naive American intellectuals like to
stick their noses in other country's businesses. Foundations have been
buying forest land in Costa Rica and western Brazil. I just finished reading one
of Fossey's gorilla books. We are imposing social standards on
South Africa that we only adopted twenty years ago. Probably a remant of
missionarism-- `our ideology is better than yours'.
It is better to persuade others by example than interference.
∂20-Oct-88 1859 JK
∂20-Oct-88 1834 JMC
How is the 91 function doing?
-------------
Not well yet --- been consulting. Do you have an informal proof
anywhere that I can look at?
∂20-Oct-88 2100 JMC
zalta
∂21-Oct-88 0207 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM hyperbowly
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 21 Oct 88 02:07:36 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 328688; Fri 21-Oct-88 05:05:30 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73484; Fri 21-Oct-88 02:02:04 PDT
Date: Fri, 21 Oct 88 01:59 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: hyperbowly
To: rwg@RUSSIAN.SPA.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19881020104555.2.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19881021085923.4.RWG@TSUNAMI.SPA.Symbolics.COM>
Date: Thu, 20 Oct 88 03:45 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
∞
====
\ X Y X π X π Y
> ATAN(------------) = ATAN(-) - ATAN(TANH(-------) COT(-------)) .
/ 2 2 Y SQRT(2) SQRT(2)
==== 2 X - Y
K = 1 K + -------
2
This is always true mod π, but using vanilla ATAN (abs ≤ π/2), the
region of true equality is peculiar in shape. It has mirror symmetry
about both (x and y) axes. There's an infinite strip of thickness
2 sqrt(2) surrounding the x axis. Think of this as a table top, viewed
from the side. Resting on the origin is a stack of covered hyperboloidal
(?) bowls, sqrt(2) high, but having diameters increasing sublinearly
(sqrt?) with height. (The region of validity is the silhouette of this
tower of bowls and tabletop, reflected around the x axis.)
A surface plot of the difference between the two sides of the above
equation reminds me of certain eroded regions of Death Valley. (The
only symmetry axis is x = -y, unlike the 0 set!)
A more careful reexpression, using ATAN2s, simplifies the error surface
to something for which there is some hope of writing an "analytic"
description, thereby doing the sum for the whole plane.
There is a similar identity, with K-1/2 in place of K, so that
Sum atan(a/(k↑2+nk+b)) has a closed form for each integer n and all a and b.
k≥1
In fact, the general quadratic can be summed bilaterally:
∞
====
\ 2 X Y
> ATAN(------------------) = ATAN(COTH(π X) TAN(π (Y + f)))
/ 2 2 2
==== (K + f) + X - Y
K = - ∞
+ ATAN(COTH(π X) TAN(π (Y - f))) .
The unilateral identities come from the symmetry when f = 0 or -1/2. The
"Death Valley effect" is partly due to an error of -π for sign change in
the summand denominator (which causes the hyperbolic edges of the "bowls"),
and partly the discontinuities of the TANs on the right (which cause the flat
tops of the bowls). Since only the former discontinuities depend on X, there
is no way to align them with the latter, so the corrective integer multiple
of π will be a fairly complicated n(X,Y,f). The region Y↑2 < X↑2 + F↑2
intersect |Y|<1-|F|, where F is the (nearest integer) fraction part of f,
looks safe, though.
∂21-Oct-88 0748 pimeet@vax.darpa.mil PI Meeting
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Oct 88 07:48:03 PDT
Posted-Date: Fri 21 Oct 88 10:46:54-EDT
Received: by vax.darpa.mil (5.54/5.51)
id AA13810; Fri, 21 Oct 88 10:46:56 EDT
Date: Fri 21 Oct 88 10:46:54-EDT
From: Mark Pullen <PIMEET@vax.darpa.mil>
Subject: PI Meeting
To: jmc@sail.stanford.edu
Message-Id: <593448414.0.PIMEET@VAX.DARPA.MIL>
Mail-System-Version: <VAX-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Dr. McCarthy:
Your address was erroneously included in my list of speakers who will need
to submit an abstract of your presentation during the PI Meeting in Dallas.
Please disregard my earlier note to that effect. I trust you have not
wasted any time on this... If so, my apologies.
Juanita Walton
-------
∂21-Oct-88 0800 JMC
fuse in car
∂21-Oct-88 1019 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
FIRST ORDER THEORIES OF QUANTIFICATION
Arkady Rabinov (AIR@SAIL)
Stanford University
Friday, October 21, 3:15pm
MJH 301
Many applications of logic to AI require that propositions be reified, i.e.,
made elements of the domain of reasoning. This is needed, in particular, in
formalizing reasoning about knowledge and about preconditions of actions.
Difficulties arise when the quantificational structure of propositions is
essential. We show how ideas from combinatory logic can be used to deal with
this problem. This approach allows us to construct first order theories in
which lambda-abstraction and quantification can be easily expressed as terms
of the language.
∂21-Oct-88 1304 scherlis@vax.darpa.mil PI MEETING PROJECT SUMMARIES
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Oct 88 13:04:08 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA14603; Fri, 21 Oct 88 15:09:34 EDT
Posted-Date: Fri 21 Oct 88 15:05:44-EDT
Received: by sun45.darpa.mil (5.54/5.51)
id AA21515; Fri, 21 Oct 88 15:05:47 EDT
Date: Fri 21 Oct 88 15:05:44-EDT
From: William L. Scherlis <SCHERLIS@vax.darpa.mil>
Subject: PI MEETING PROJECT SUMMARIES
To: SW-PI@vax.darpa.mil
Cc: nfields@vax.darpa.mil
Message-Id: <593463944.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
To The Software and Algorithms PIs:
For the PI mmeting, we are preparing a booklet that will contain one
page summaries of the individual projects funded by the office.
Through various glitches of a bureaucratic and technical sort, the
solicitation for the summaries failed to reach the Software PI
community. Since time is now very tight, I propose the following: I
will send you (in the next hour) the summaries we concocted from your
recent reports for use here in the office. If you are happy with this
summary, please respond to that effect. If not, please send us a
revised version.
Also, please send us a list of TWO of your recent publications to
include. The publications should be of a general sort and contain
pointers to your other work.
The format for the page will be:
- Institution
- Title of Project
- PIs and other principals
- Project summary
- References
NOTE THAT ALL THIS MUST FIT ON ONE PAGE (60 lines of text).
Please send us your response by net mail. Follow the following
guidelines in order to help us process the text swiftly: (1) DO NOT
PUT LEADING SPACES ON ANY LINES OR EXTRA EMBEDDED SPACES IN THE LINES.
(2) SEPARATE PARAGRAPHS BY A BLANK LINE. (3) NO LEADING INDENTATION
ANYWHERE. (4) DO NOT INCLUDE ANY SCRIBE, TROFF, TEX, ETC DECORATIONS.
Use plain text. This message satisfies the constraints, for example.
Finally: We need your response by 5pm Monday!! (If we get no
response, we will use the summary we have on hand, which I think will
fairly represent the work.) We are sorry that there is such a tight
deadline. (Some explanation will be given in the message that
follows: See below.)
Responses should go to Nicole Fields (nfields@vax.darpa.mil).
I will be sending you shortly a more general note concerning the
recent half-hour reviews conducted by Jack Schwartz and our plans for
the PI meeting.
Bill
-------
∂21-Oct-88 1342 kar@polya.Stanford.EDU Applications AI comp
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Oct 88 13:42:06 PDT
Received: from LOCALHOST by polya.Stanford.EDU (5.59/25-eef) id AA04679; Fri, 21 Oct 88 13:42:17 PDT
Message-Id: <8810212042.AA04679@polya.Stanford.EDU>
To: jmc@sail
Subject: Applications AI comp
Date: Fri, 21 Oct 88 13:42:14 -0700
From: kar@polya.Stanford.EDU
------- Forwarded Message
From: nilsson@Tenaya.stanford.edu (Nils Nilsson)
Subject: Applications AI comp.
I think it would not be a good idea to use electronic mail to
communicate possible comp problems. Perhaps you could spread that
word. Thanks, -Nils
------- End of Forwarded Message
∂21-Oct-88 1434 nfields@vax.darpa.mil SUMMARY ENCLOSED
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Oct 88 14:32:51 PDT
Received: from sun38.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA15224; Fri, 21 Oct 88 17:32:25 EDT
Posted-Date: Fri 21 Oct 88 17:32:58-EDT
Received: by sun38.darpa.mil (5.54/5.51)
id AA15478; Fri, 21 Oct 88 17:33:03 EDT
Date: Fri 21 Oct 88 17:32:58-EDT
From: Nicole L. Fields <NFIELDS@vax.darpa.mil>
Subject: SUMMARY ENCLOSED
To: rpg@sail.stanford.edu, jmc@sail.stanford.edu
Message-Id: <593472778.0.NFIELDS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Project Title Common Prototyping Language
Performer(s) Stanford University
Technical Description
The goal of this research is to make initial steps in the design of a common
prototyping language (CPL). The contractor will develop a set of design
principles and use these to formulate an initial language and environment
specification. Stanford will make use of existing tools to develop preliminary
experimental prototypes to support further exploration, particularly within the
context of the CPL Working Group.
Principal Accomplishments
This is a new effort.
Expected Accomplishments
A set of design principles and an initial candidate design for CPL and its
supporting environment will be developed. An implementation strategy for the
environment and language will be developed, including operating system
requirements.
-------
∂21-Oct-88 1441 nfields@vax.darpa.mil summary enclosed
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Oct 88 14:41:23 PDT
Received: from sun38.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA15269; Fri, 21 Oct 88 17:41:00 EDT
Posted-Date: Fri 21 Oct 88 17:41:33-EDT
Received: by sun38.darpa.mil (5.54/5.51)
id AA15492; Fri, 21 Oct 88 17:41:37 EDT
Date: Fri 21 Oct 88 17:41:33-EDT
From: Nicole L. Fields <NFIELDS@vax.darpa.mil>
Subject: summary enclosed
To: clt@sail.stanford.edu, jmc@sail.stanford.edu
Message-Id: <593473294.0.NFIELDS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Project Title Programming and Proving with Higher Order Abstractions and
Reflection
Performer(s) Stanford University
Technical Description
The contractor will perform basic research in the mathematical theory of
computation to develop a semantic framework for functions, control structures,
assignment, process abstractions, and reflection. This framework will be used
to develop an experimental system for manipulating and reasoning about
programs.
Principal Accomplishments
Programs with state and streams. The theory of program equivalence has been
extended to include programs with memory (for example Lisp or Scheme programs
that use Rplacs). The theory has also been extended to include streams,
including #lazy" objects, and input/output stream abstractions. Memoizing
transformations and transformations between different types of streams that
preserve the sequence computed were defined. Methods of proving properties of
streams were developed.
Foundations. Steps were taken to facilitate use of conditional equational
theories for program semantics. Techniques based on combinators were developed
to facilitate efficient compilation of coroutines and other programs that use
function and control abstractions in structured ways.
Expected Accomplishments
Foundations. Establish theoretical results necessary to permit use of a
higher-order language for expressing properties of imperative programs.
Object-oriented systems. Apply theories developed earlier to a richer class of
programs, especially to object oriented programs, to obtain a better
understanding of program equivalence and transformation in this realm. Also,
extend the programming language to include process abstractions and
asynchronous message passing.
Prototyping and experimentation. Develop simple interactive program
transformation tools to facilitate the study of substantial examples.
-------
∂21-Oct-88 1543 CLT dinner with Susie, etc.
no, but I think you should take Hazel, since you aren't willing
to control the chaos enought that Timothy is not too destracted
to eat supper.
∂21-Oct-88 1711 MPS vacation
I will be gone until November 1. You can reach
me at 967-5767 if necessary.
Pat
∂22-Oct-88 1002 CLT Hazel
was concerned about hurricane Joan and its effects on
Nicuaragua. If there is anything in NS about it you
might print it out for her.
∂22-Oct-88 1008 RPG PI Meeting
To: nfields@vax.darp.mil
CC: JMC@SAIL.Stanford.EDU
The summary should be amended as follows. In addition, here are two
recent publications:
Technical Issues of Separation in Function Cells and Value Cells, Lisp and
Symbolic Computation, Vol. 1 No. 1, 1988.
Qlisp: Experience and New Directions, 1988 ACM Symposium on Parallel
Programming: Experience with Applications, Languages, and Systemsm New
Haven, Connecticut, 1988.
**************************************************************************
Project Title Common Prototyping Language
Performer(s) Stanford University
Technical Description
The goal of this research is to make initial steps in the design of a common
prototyping system (CPS). The contractor will develop a set of design
principles and use these to formulate an initial language and environment
specification. Stanford will make use of existing tools to develop preliminary
experimental prototypes to support further exploration, particularly within the
context defined by the CPS Working Group.
Principal Accomplishments
This is a new effort.
Expected Accomplishments
A set of design principles and an initial candidate design for CPL and its
supporting environment will be developed. An implementation strategy for the
environment and language will be developed, including operating system
requirements.
∂22-Oct-88 1011 RPG PI Meeting
To: nfields@VAX.DARPA.MIL
CC: JMC@SAIL.Stanford.EDU
The summary should be amended as follows. In addition, here are two
recent publications:
Technical Issues of Separation in Function Cells and Value Cells, Lisp and
Symbolic Computation, Vol. 1 No. 1, 1988.
Qlisp: Experience and New Directions, 1988 ACM Symposium on Parallel
Programming: Experience with Applications, Languages, and Systemsm New
Haven, Connecticut, 1988.
**************************************************************************
Project Title Common Prototyping Language
Performer(s) Stanford University
Technical Description
The goal of this research is to make initial steps in the design of a common
prototyping system (CPS). The contractor will develop a set of design
principles and use these to formulate an initial language and environment
specification. Stanford will make use of existing tools to develop preliminary
experimental prototypes to support further exploration, particularly within the
context defined by the CPS Working Group.
Principal Accomplishments
This is a new effort.
Expected Accomplishments
A set of design principles and an initial candidate design for CPL and its
supporting environment will be developed. An implementation strategy for the
environment and language will be developed, including operating system
requirements.
∂23-Oct-88 0842 barwise@russell.Stanford.EDU common knowledge
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Oct 88 08:42:35 PDT
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Sun, 23 Oct 88 08:44:56 PDT
To: jmc@sail
Subject: common knowledge
Date: Sun, 23 Oct 88 08:44:55 PDT
From: Jon Barwise <barwise@russell.Stanford.EDU>
John, I am now in the process of preparing the final
draft for the printer of my paper on common knowledge,
for the JSL. I would like to be sure to give credit
where due. In particular, I think you are responsible
for the observation that the iterate version does not
satisfy the crucial condition that common knowledge
implies common knowledge of common knowledge, on that
analysis. Is that right? Is there a reference I can
give? Jon
∂23-Oct-88 1559 beeson@ucscd.UCSC.EDU common sense in word problems
Received: from ucscd.UCSC.EDU ([128.114.129.2]) by SAIL.Stanford.EDU with TCP; 23 Oct 88 15:58:58 PDT
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA07070; Sun, 23 Oct 88 16:01:49 PDT
Date: Sun, 23 Oct 88 16:01:49 PDT
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8810232301.AA07070@ucscd.UCSC.EDU>
To: jmc@sail.stanford.edu
Subject: common sense in word problems
My software (which I call MATHPERT) does not attemps to solve
word problems. (I wrote some programs that did so, but it's too hard
to accept a wide enough variety.) Now however I'm testing MATHPERT on
real exams from real algebra classes, and they do contain word
problems. I pose the following as an exercise for your bright
undergraduates: give a complete formalization in predicate calculus
of the background knowledge and assumptions needed to solve the
following problem:
John can paint a fence in 4 hours and Jerry can paint it in 6 hours.
How long would it take if they work together?
After carrying out that formalization, one will understand why it is
that algebra students really don't work these problems by logical
analysis, but by pattern-matching. They can't work a problem about
painting fences until they have had a demonstration problem involving
painting fences. Even substituting another kind of task won't enable
them to generalize.
∂23-Oct-88 1804 RDZ@Score.Stanford.EDU Dinner?
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Oct 88 18:03:54 PDT
Date: Sun 23 Oct 88 18:02:13-PDT
From: Ramin Zabih <RDZ@Score.Stanford.EDU>
Subject: Dinner?
To: jmc@Sail.Stanford.EDU
Message-ID: <12440841277.10.RDZ@Score.Stanford.EDU>
Are you going to be interested in dinner this evening?
Ramin
-------
∂24-Oct-88 0923 VAL Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
A LOCAL FORMALIZATION OF INHERITANCE
Matthew Ginsberg (GINSBERG@POLYA)
Stanford University
Friday, October 28, 3:15pm
MJH 301
Existing work on formalizing inheritance hierarchies suffers from two
significant drawbacks. On the one hand, the formalization is often
nonlocal, in that the translation from the inheritance hierarchy into
the relevant declarative language requires a complete examination of
the topology of the graph being considered. On the other hand, the
formalization frequently uses an inference technique that is tailored
specifically to problems of this sort, and cannot be applied outside
of this very restricted area.
I present a formalization that avoids these difficulties. It is
local, and relies on a very simple generalization of existing
approaches to default reasoning. The generalization is based on a
simple formalization of the notion of causality.
∂24-Oct-88 0948 andy@cayuga.Stanford.EDU JFK's book? (was: Quayle)
Received: from cayuga.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Oct 88 09:48:30 PDT
Received: by cayuga.Stanford.EDU (5.51/inc-1.01)
id AA13601; Mon, 24 Oct 88 09:46:53 PDT
Date: Mon, 24 Oct 88 09:46:53 PDT
From: Andy Freeman <andy@cayuga.Stanford.EDU>
Message-Id: <8810241646.AA13601@cayuga.Stanford.EDU>
To: JMC@SAIL.STANFORD.EDU
Subject: JFK's book? (was: Quayle)
Newsgroups: su.etc
In-Reply-To: <86vP7@SAIL.Stanford.EDU>
Organization: Stanford University
Cc:
In article <86vP7@SAIL.Stanford.EDU> you write:
>Kennedy wrote a best-selling book. How many BBoard readers have
>read Kennedy's book?
I thought Kennedy wrote TWO books, *Profiles in Courage* and *Why
England Slept* (maybe "While" instead of "Why"). I've also heard that
one of them was actually ghost-written; I vaguely that the real
author's name was something like Salinger or Schlessinger (sic).
-andy
--
UUCP: {arpa gateways, decwrl, uunet, rutgers}!polya.stanford.edu!andy
ARPA: andy@polya.stanford.edu
(415) 329-1718/723-3088 home/cubicle
∂24-Oct-88 1005 CLT painter
I can't find the paper I wrote the painters number on
and it isn't in your phne file. Could you send it
in a msg. (Name is Frank Rosa).
∂24-Oct-88 1022 CLT notes
Not very useful since find can't find it
∂24-Oct-88 1024 CLT umbrella
I talked to Peter Friedland this morning. He said he
was potentially interested in taking over our contract
and suggested we (you, he, and I at least) talk about it
further in Dallas. To which I agreed.
∂24-Oct-88 1031 CLT pullen
I will send a msg harassing him about Qlisp
If I get no reply I will start calling.
Should I mention Qlisp for Okuno or just wait til Dallas?
∂24-Oct-88 1031 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSD-CF Employee List
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Oct 88 10:31:36 PDT
Received: from polya.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Mon 24 Oct 88 10:29:51-PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA06116; Mon, 24 Oct 88 10:31:21 PDT
Date: Mon, 24 Oct 88 10:31:21 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
Message-Id: <8810241731.AA06116@polya.Stanford.EDU>
To: Facil@Score
Subject: CSD-CF Employee List
As promised, here is the current employee list for CSD-CF
J. Ball Director
T. Dienstbier Assoc Director
L. Gotelli Admin Asst
J. Baldwin Office Asst
D. Markley Hardware Engineer
D. Coates Hardware Technician
M. Frost Systems Programmer SAIL WAITS
S. Grossman Systems Programmer SCORE TOPS20, SUN OS
F. Shakeri Systems Programmer POLYA ULTRIX, UNIX
D. Koronakos Programmer Analyst CF Acct, CS database
Part time/students
3 tape operators
1 wire person
1 APS (phototypesetter) operator
∂24-Oct-88 1452 RWF re: Quayle
[In reply to message rcvd 22-Oct-88 18:46-PT.]
I read Profiles in COurage many years ago. One suspects
ghosting, with the family money so readily available.
He earlier wrote Why England Slept.
∂24-Oct-88 1605 chandler@polya.Stanford.EDU I have something for you
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Oct 88 16:05:34 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA29254; Mon, 24 Oct 88 16:05:37 PDT
Date: Mon, 24 Oct 1988 16:05:26 PDT
From: "Joyce R. Chandler" <chandler@polya.stanford.edu>
To: jmc@sail
Subject: I have something for you
Message-Id: <CMM.0.87.593737526.chandler@polya.stanford.edu>
Professor Heller came by to drop something off for you. Pat is out and he
didn't want to just leave it....it's of a confidential nature. He asked if I
would see to it that you received it. Please let me know when you'll be in
and I'll stop it by for you.
∂24-Oct-88 1622 RLM@Score.Stanford.EDU Re: Quayle
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 24 Oct 88 16:22:29 PDT
Date: Mon 24 Oct 88 16:21:01-PDT
From: Robert L. Miller <RLM@Score.Stanford.EDU>
Subject: Re: Quayle
To: JMC@Sail.Stanford.EDU
cc: rlm@Score.Stanford.EDU
In-Reply-To: <86vP7@SAIL.Stanford.EDU>
Message-ID: <12441085000.15.RLM@Score.Stanford.EDU>
I take it Profiles In Courage was the best-seller and not While London Slept.
-------
∂25-Oct-88 0454 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM theology question
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 25 Oct 88 04:54:15 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 329534; Tue 25-Oct-88 07:52:53 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73716; Tue 25-Oct-88 04:48:57 PDT
Date: Tue, 25 Oct 88 04:46 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: theology question
To: macsyma-i@RUSSIAN.SPA.Symbolics.COM, math-fun@RUSSIAN.SPA.Symbolics.COM
cc: "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"R@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19881025114600.7.RWG@TSUNAMI.SPA.Symbolics.COM>
CLtL defines (ATAN z1 z2) to be an error if either arg is complex.
MACSYMA's ATAN2(z1,z2) is tolerant, but mute. What is a reasonable extension?
One idea is CARG(z2+%I*z1), which is consistent with the current behavior for
real z1, z2. (CARG := complex arg := phase.)
But unlike ATAN, this has the peculiarity to always be real. And it
doesn't preserve the identity ATAN2(SIN(z),COS(z)) = z.
Another approach might be to say that if scaling z1 and z2 by a positive
real is a noop, and scaling by a negative real adds π, then scaling by
1/sqrt(z1↑2+z2↑2) should add carg of this sqrt. (But which?). The scaled
z1 and z2 can now be regarded as the sin and cos of a complex angle, which
is the (rest of the) value of the atan2. This would give a familiar
ATAN2(z1,z2) = ATAN(z1/z2) + fudge. But exactly which fudge?
∂25-Oct-88 0955 STAGER@Score.Stanford.EDU Winter text orders
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 25 Oct 88 09:55:16 PDT
Date: Tue 25 Oct 88 09:47:07-PDT
From: Claire Stager <STAGER@Score.Stanford.EDU>
Subject: Winter text orders
To: jmc@Sail.Stanford.EDU, reid@decwrl.dec.com, gail@sol-margaret.Stanford.EDU
Office: CS-TAC 29, 723-6094
Message-ID: <12441275436.11.STAGER@Score.Stanford.EDU>
Please let me know what, if any, required and optional texts you'll be
needing ordered for your Winter Qtr. courses. Necessary ordering information
includes the following:
Author
Title
Publisher
Required or Optional
Estimated Enrollment
Please reply to me by next MONDAY, OCTOBER 31, to ensure that the Bookstore
has time to order your texts and get them onto their shelves by the beginning
of Winter Quarter.
Thanks.
Claire
-------
∂25-Oct-88 1633 betsy@russell.Stanford.EDU Office Space at CSLI
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 25 Oct 88 16:33:45 PDT
Received: by russell.Stanford.EDU (4.0/4.7); Tue, 25 Oct 88 16:36:04 PDT
Date: Tue 25 Oct 88 16:36:03-PDT
From: Betsy Macken <BETSY@CSLI.Stanford.EDU>
Subject: Office Space at CSLI
To: jmc@SAIL.STANFORD.EDU
Cc: ingrid@CSLI.Stanford.EDU, bach-hong@CSLI.Stanford.EDU
Message-Id: <593825763.0.BETSY@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
We'd be delighted to give you some space at CSLI. How about room 226?
That is a triple office currently in use regularly by Ed Zalta. Dave
Rumelhart and Herb Clark have also been given space in that office, but
they rarely use it. So you could choose either of the empty desks and one
of the bookshelves for your use. We will order a name plate for you, but
(due to problems I won't bore you with), getting name plates takes a while.
You can get a key to the office from Bach-Hong Tran in Ventura 13.
If you would like a terminal or other equipment, let me know and I'll
see what I can do.
Betsy
-------
∂25-Oct-88 1741 GLB date of my exam
To: sf@CSLI.Stanford.EDU, JK@SAIL.Stanford.EDU, CLT@SAIL.Stanford.EDU,
jcm@Polya.Stanford.EDU, JMC@SAIL.Stanford.EDU
I would fix the tentative date of November 22, at 2.15. Would that be
good?
∂25-Oct-88 1829 CLT inamori
Sasaki of Inamori called. 3 items
(1) They need ASAP
(a) script for acceptance speech (3-4 min)
(b) text for workshop lecture
(2) They would like you to be available for Newspaper interviews
at the Tokyo Prince Hotel 9-12am Nov 8
(3) They want to know what equipment you need for workshop lectures
Nov 11,12 (OH, slides ...)
Would like to telefax Friday am
(1) when to expect items (1)
(2) will you be available for the interviews
(3) what equipment is needed
∂26-Oct-88 1440 chandler@polya.Stanford.EDU Something for you...
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 26 Oct 88 14:39:54 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA09013; Wed, 26 Oct 88 14:39:52 PDT
Date: Wed, 26 Oct 1988 14:39:50 PDT
From: "Joyce R. Chandler" <chandler@polya.stanford.edu>
To: jmc@sail
Subject: Something for you...
Message-Id: <CMM.0.87.593905190.chandler@polya.stanford.edu>
I've got something to deliver personally to you.....please let me know when
you'll be in the office. Thanks.
jc
∂27-Oct-88 0652 harnad%confidence.Princeton.EDU@Princeton.EDU California talks
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 27 Oct 88 06:49:36 PDT
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.83)
id AA29475; Thu, 27 Oct 88 09:45:11 EDT
Received: by psycho.Princeton.EDU (3.2/1.62)
id AA02276; Thu, 27 Oct 88 09:41:07 EDT
Date: Thu, 27 Oct 88 09:41:07 EDT
From: harnad@confidence (Stevan Harnad)
Message-Id: <8810271341.AA02276@psycho.Princeton.EDU>
To: shepard@psych.stanford.edu
Subject: California talks
Cc: brian@psych.stanford.edu, bt@psych.stanford.edu, gluck@psych.stanford.edu,
jmc@sail.stanford.edu, leda@psych.stanford.edu,
wasow@russell.stanford.edu
Dear Roger:
I'm giving a series of talks on the West Coast next week (UCSD, UCSC,
Berkeley). I'll be talking at CSLI on Thursday morning at 10
(Nov 3). I'm free that afternoon; if there's any interest in the
psychology (or philosophy, or computer science) department, and the
notice of one week is not too short, I'd be happy to give one of my
talks there. (The CSLI talk is number I: "Minds, Machines and
Searle.")
The talks are described below, and I've included abstracts. Please let
me know whether there's any interest.
Best wishes,
Stevan
--------------
The three talks ("I. Minds, Machines and Searle," "II. Uncomplemented
Categories, or What Is It Like to Be a Bachelor?" and "III. Category
Induction and Representation") form a 3-part series or can each
be given independently. There is also the possibility of a hybrid
combination of all three.
In these papers I am trying to propose a generalization of the
phenomenon of categorical perception as a way of grounding higher-order
categorization and cognition, bottom up. This also turns out to be
connected with some important foundational problems in cognitive
science. I am also criticizing and proposing an alternative to the
prevailing "prototype" view that I know has some adherents
in California. So a discussion might be stimulating and useful.
The three abstracts follow below; the full texts are available too; the
first is to appear in the inaugural issue of the Journal of
Experimental and Theoretical Artificial intelligence; the second was my
presidential address to the Society for Philosophy and Psychology. The
third is a chapter in a book I edited recently: Categorical Perception:
The Groundwork of Cognition (Cambridge University Press 1987). There is
also a 4th related paper (no abstract available) called: "IV. The Origin of
Words: A Psychophysical Hypothesis," to be presented at a conference on
the evolution of language in Poland this winter.
--------------------------------------------------------
I. Minds, Machines and Searle
Searle's provocative "Chinese Room Argument" attempted to
show that the goals of "Strong AI" are unrealizable.
Proponents of Strong AI are supposed to believe that (i) the
mind is a computer program, (ii) the brain is irrelevant,
and (iii) the Turing Test is decisive. Searle's point is
that since the programmed symbol-manipulating instructions
of a computer capable of passing the Turing Test for
understanding Chinese could always be performed instead by a
person who could not understand Chinese, the computer can
hardly be said to understand Chinese. Such "simulated"
understanding, Searle argues, is not the same as real
understanding, which can only be accomplished by something
that "duplicates" the "causal powers" of the brain. In this
paper I make the following points:
1. Simulation versus Implementation: Searle fails to
distinguish between the simulation of a mechanism, which is
only the formal testing of a theory, and the implementation
of a mechanism, which does duplicate causal powers. Searle's
"simulation" only simulates simulation rather than
implementation. It can no more be expected to understand
than a simulated airplane can be expected to fly.
Nevertheless, a successful simulation must capture formally
all the relevant functional properties of a successful
implementation.
2. Theory-Testing versus Turing-Testing: Searle's argument
conflates theory-testing and Turing-Testing. Computer
simulations formally encode and test models for human
perceptuomotor and cognitive performance capacities; they
are the medium in which the empirical and theoretical work
is done. The Turing Test is an informal and open-ended test
of whether or not people can discriminate the performance of
the implemented simulation from that of a real human being.
In a sense, we are Turing-Testing one another all the time,
in our everyday solutions to the "other minds" problem.
3. The Convergence Argument: Searle fails to take
underdetermination into account. All scientific theories are
underdetermined by their data; i.e., the data are compatible
with more than one theory. But as the data domain grows, the
degrees of freedom for alternative (equiparametric) theories
shrink. This "convergence" constraint applies to AI's "toy"
linguistic and robotic models as well, as they approach the
capacity to pass the Total (asympototic) Turing Test. Toy
models are not modules.
4. Brain Modeling versus Mind Modeling: Searle also fails
to note that the brain itself can be understood only through
theoretical modeling, and that the boundary between brain
performance and body performance becomes arbitrary as one
converges on an asymptotic model of total human performance
capacity.
5. The Modularity Assumption: Searle implicitly adopts a
strong, untested "modularity" assumption to the effect that
certain functional parts of human cognitive performance
capacity (such as language) can be be successfully modeled
independently of the rest (such as perceptuomotor or
"robotic" capacity). This assumption may be false for models
approaching the power and generality needed to pass the
Total Turing Test.
6. The Teletype versus the Robot Turing Test: Foundational
issues in cognitive science depend critically on the truth
or falsity of such modularity assumptions. For example, the
"teletype" (linguistic) version of the Turing Test could in
principle (though not necessarily in practice) be
implemented by formal symbol-manipulation alone (symbols in,
symbols out), whereas the robot version necessarily calls
for full causal powers of interaction with the outside world
(seeing, doing AND linguistic understanding).
7. The Transducer/Effector Argument: Prior "robot" replies
to Searle have not been principled ones. They have added on
robotic requirements as an arbitrary extra constraint. A
principled "transducer/effector" counterargument, however,
can be based on the logical fact that transduction is
necessarily nonsymbolic, drawing on analog and analog-to-
digital functions that can only be simulated, but not
implemented, symbolically.
8. Robotics and Causality: Searle's argument hence fails
logically for the robot version of the Turing Test, for in
simulating it he would either have to USE its transducers
and effectors (in which case he would not be simulating all
of its functions) or he would have to BE its transducers and
effectors, in which case he would indeed be duplicating
their causal powers (of seeing and doing).
9. Symbolic Functionalism versus Robotic Functionalism: If
symbol-manipulation ("symbolic functionalism") cannot in
principle accomplish the functions of the transducer and
effector surfaces, then there is no reason why every
function in between has to be symbolic either. Nonsymbolic
function may be essential to implementing minds and may be a
crucial constituent of the functional substrate of mental
states ("robotic functionalism"): In order to work as
hypothesized, the functionalist's "brain-in-a-vat" may have
to be more than just an isolated symbolic "understanding"
module -- perhaps even hybrid analog/symbolic all the way
through, as the real brain is.
10. "Strong" versus "Weak" AI: Finally, it is not at all
clear that Searle's "Strong AI"/"Weak AI" distinction
captures all the possibilities, or is even representative of
the views of most cognitive scientists. In any case, most of
AI is not concerned with modeling the mind but with making
machines do intelligent things.
Hence, most of Searle's argument turns out to rest on
unanswered questions about the modularity of language and
the scope of the symbolic approach to modeling cognition. If
the modularity assumption turns out to be false, then a
top-down symbol-manipulative approach to explaining the mind
may be completely misguided because its symbols (and their
interpretations) remain ungrounded -- not for Searle's
reasons (since Searle's argument shares the cognitive
modularity assumption with "Strong AI"), but because of the
transdsucer/effector argument (and its ramifications for the
kind of hybrid, bottom-up processing that may then turn out
to be optimal, or even essential, in between transducers and
effectors). What is undeniable is that a successful theory
of cognition will have to be computable (simulable), if not
exclusively computational (symbol-manipulative). Perhaps
this is what Searle means (or ought to mean) by "Weak AI."
------------------------------------------------------------
II. Uncomplemented Categories or
What is it Like to Be a Bachelor?
To be able to categorize and identify objects, both concrete
and abstract, we must form an internal representation that
can sort the members from the nonmembers of the category. To
do this, we must first sample instances of members and
nonmembers and pick out the features that the members share
and that will reliably distinguish them from the nonmembers.
The set of nonmembers is called the "complement" of the
category and there are some constraints on what it can be:
The complement of a category must be bounded (it cannot
consist of everything else in the universe, but only those
things we might mistake for members, otherwise the search
for distinguishing features could go on for ever) and it
cannot be empty (otherwise there would be no way to
determine the distinguishing features of the members).
Category representations are "context-dependent" (a
different complement of confusable alternatives requires
different distinguishing features) and "approximate" (a
context can always be widened, which may require finding
new, more general features). If the complement of a category
is not available, sometimes it can be furnished by
extrapolation and analogy, as perhaps occurs with (1) the
"impoverished" grammatical input of the language-learning
child and (2) the experiential category "what it's like to
be a bachelor" (for someone who has never experienced its
complement). Sometimes this strategy cannot work, however,
as in the case of (3) experiential categories such as "what
it's like to be awake, to be aware, to be alive, to be" and
(4) epistemic and existential categories such as "is
conceivable" and "exists," and perhaps even some kinds of
counterfactuals. Certain longstanding philosophical problems
associated with 3 and 4 may be related in part to the fact
that such categories are not just uncomplemented but
uncomplementable, and hence that our representations for
them, if we have any, must be defective. Some of the self-
denial paradoxes (e.g., Russell's paradox) may also arise in
part because of problems of noncomplementability.
------------------------------------------------------------
III. Category Induction and Representation
Categorization is a very basic cognitive activity. It is
involved in any task that calls for differential responding,
from operant discrimination to pattern recognition to naming
and describing objects and states-of-affairs. Explanations
of categorization range from nativist theories denying that
any nontrivial categories are acquired by learning to
inductivist theories claiming that most categories are
learned.
"Categorical perception" (CP) is the name given to a
suggestive perceptual phenomenon that may serve as a useful
model for categorization in general: For certain perceptual
categories, within-category differences look much smaller
than between-category differences even when they are of the
same size physically. For example, in color perception,
differences between reds and differences between yellows
look much smaller than equal-sized differences that cross
the red/yellow boundary; the same is true of the phoneme
categories /ba/ and /da/. Indeed, the effect of the category
boundary is not merely quantitative, but qualitative.
There have been two theories to explain CP effects. The
"Whorf Hypothesis" explains color boundary effects by
proposing that language somehow determines our view of
reality. The "motor theory of speech perception" explains
phoneme boundary effects by attributing them to the patterns
of articulation required for pronunciation. Both theories
seem to raise more questions than they answer, for example:
(i) How general and pervasive are CP effects? Do they occur
in other modalities besides speech-sounds and color? (ii)
Are CP effects inborn or can they be generated by learning
(and if so, how)? (iii) How are categories internally
represented? How does this representation generate
successful categorization and the CP boundary effect?
Some of the answers to these questions will have to come
from ongoing research, but the existing data do suggest a
provisional model for category formation and category
representation. According to this model, CP provides our
basic or elementary categories. In acquiring a category we
learn to label or identify positive and negative instances
from a sample of confusable alternatives. Two kinds of
internal representation are built up in this learning by
acquaintance: (1) an iconic representation that subserves
our similarity judgments and (2) an analog/digital feature-
filter that picks out the invariant information allowing us
to categorize the instances correctly. This second,
categorical representation is associated with the category
name. Category names then serve as the atomic symbols for a
third representational system, the (3) symbolic
representations that underlie language and that make it
possible for us to learn by description.
This model provides no particular or general solution to the
problem of inductive learning, only a conceptual framework;
but it does have some substantive implications, for example,
(a) the "cognitive identity of (current) indiscriminables":
Categories and their representations can only be provisional
and approximate, relative to the alternatives encountered to
date, rather than "exact." There is also (b) no such thing
as an absolute "feature," only those features that are
invariant within a particular context of confusable
alternatives. Contrary to prevailing "prototype" views,
however, (c) such provisionally invariant features must
underlie successful categorization, and must be "sufficient"
(at least in the "satisficing" sense) to subserve reliable
performance with all-or-none, bounded categories, as in CP.
Finally, the model brings out some basic limitations of the
"symbol-manipulative" approach to modeling cognition,
showing how (d) symbol meanings must be functionally
anchored in nonsymbolic, "shape-preserving" representations
-- iconic and categorical ones. Otherwise, all symbol
interpretations are ungrounded and indeterminate. This
amounts to a principled call for a psychophysical (rather
than a neural) "bottom-up" approach to cognition.
∂27-Oct-88 1047 littell@polya.Stanford.EDU Alex Gorbis
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 27 Oct 88 10:47:43 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA18959; Thu, 27 Oct 88 10:47:41 PDT
Date: Thu, 27 Oct 88 10:47:41 PDT
From: Angelina M. Littell <littell@polya.Stanford.EDU>
Message-Id: <8810271747.AA18959@polya.Stanford.EDU>
To: jmc@sail
Cc: littell@polya.Stanford.EDU, mps@sail
Subject: Alex Gorbis
Prof. McCarthy,
Alex came by my office and said he had an RAship with you for this
academic year for 50% time. Please confirm and let me know what
account his salary should be charged to.
Thank you.
--Angie
∂27-Oct-88 1057 goldberg@polya.Stanford.EDU Industrial Lecturers
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 27 Oct 88 10:57:13 PDT
Received: by polya.Stanford.EDU (5.59/25-eef) id AA19830; Thu, 27 Oct 88 10:57:13 PDT
Date: Thu 27 Oct 88 10:57:06-PDT
From: Andrew V. Goldberg <GOLDBERG@Polya.Stanford.EDU>
Subject: Industrial Lecturers
To: jmc@sail
Message-Id: <593978226.0.GOLDBERG@Polya.Stanford.EDU>
Mail-System-Version: <VAX-MM(229)+TOPSLIB(128)@Polya.Stanford.EDU>
John,
Are you going to be the person who recommends Industrial Lecturers this
year, or do you whish to pass this task t the visiting professor committee?
--Andy
-------
∂27-Oct-88 1438 scherlis@vax.darpa.mil REVIEWS, PI MEETING, PLANS
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 27 Oct 88 14:38:27 PDT
Received: from sun45.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA07600; Thu, 27 Oct 88 16:11:11 EDT
Posted-Date: Thu 27 Oct 88 16:06:35-EDT
Received: by sun45.darpa.mil (5.54/5.51)
id AA25787; Thu, 27 Oct 88 16:06:36 EDT
Date: Thu 27 Oct 88 16:06:35-EDT
From: William L. Scherlis <SCHERLIS@vax.darpa.mil>
Subject: REVIEWS, PI MEETING, PLANS
To: SW-PI@vax.darpa.mil
Cc: squires@vax.darpa.mil, scherlis@vax.darpa.mil, boesch@vax.darpa.mil
Message-Id: <593985995.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
To the Software and Algorithms PIs:
This note contains various bits of news and comment. It also suggests
some topics for our community session at the PI meeting. Please
comment on these suggested topics! The real value of the PI meeting
will the results of our community discussions.
THE BRIEFINGS. I thank you all again for coming here to give the
half-hour briefings to Jack Schwartz. I recognize that the briefings
were an imposition on you, especially when there is no clear model
for what they are intended to accomplish. To clarify: Jack
instructed me to set up two days of these briefings so that he could
improve his technical understanding of the projects and have some
personal interaction with the PIs. As I understand it, he is using
the knowledge he gains to help him in assessing both the overall
program and the individual efforts for his own planning and
budgeting.
The "up-side" scenario is that we are trying to scale up the program
in a way that will accelerate the development of our ideas into a form
that can be meaningfully transitioned. The only way to do this is to
show that the new program will have meaningful structure, a coherent
plan, technical depth, demonstrable accomplishment, and a clear notion
of what are the products and who will receive them.
Our current detailed thinking on the future of the software program
(really only assembled as a visible program in ISTO two years ago)
will be discussed in our community session at the PI meeting. (See
[1] below.)
Since the new program must evolve from the current set of efforts, the
first step is to evaluate the current efforts with respect to these
qualities. Jack wanted a direct personal sense of this before
proceeding. I believe that this, along with the more usual budget
management decisions, motivated him to request the briefings. These
briefings also occurred for several other program areas in the office.
THE PI MEETING. There has been some discussion in the community and
internally in the ISTO concerning the PI meeting. It will go ahead
as planned, and you should be registered by now. There will be a
number of plenary sessions featuring talks by esteemed personages.
There will also be several sessions to discuss new ideas and to get
together as a community to discuss issues relevant to us. These
sessions will occur Wednesday and Thursday afternoons, if I
understand the schedule correctly. It is from these sessions that I
hope to obtain some accomplishment from this meeting. There are a
number of topics that we can cover:
[1] Plans and directions for the ISTO Software Technology program:
The program will be changing and, hopefully, improving in significant
ways in the next two years. The planned changes involve both (1) more
specific technical focal points and (2) some alteration in the way we
do business with the community. The goal is to obtain an improved
level of accomplishment while retaining essential exploratory
activity.
There will be time at the PI Meeting for us (Squires, Boesch, and
myself) to describe our current view of the overall software activity.
DARPA is usually an information sink, not a source. But in this case,
we are missing opportunities because the community doesn't have a very
good idea of how we interpret events and what sort of technical plans
we are formulating. This will be especially important as the software
program begins to take more major evolutionary steps.
As you know, the DARPA approach is to focus on specific objectives
(even if they are of a very long term character) and build programs
in which communities form that can build on shared results, including
engineering results. This means choosing a smaller number of areas
that we can pursue in greater depth, excluding other areas. As new
ideas come forward, we adjust our programs to respond, but we do not
strive for the broad coverage or "completeness" that NSF seeks. This
makes it especially important for us to develop plans that reflect
sound, and yet progressive, thinking in the research community.
As our scheming progresses, therefore, we try continually to seek
honest critical reactions from the community.
Items for discussion are:
(1.1) new technical directions for ISTO to consider, and
(1.2) means by which we can improve our externally visible
accomplishment (in order to preserve the program and
enable growth).
[2] Industry collaboration:
ISTO results generally find their way into Defense applications
through the commercial base, as opposed to going directly into DoD or
directly through Defense industry. This has important advantages for
Defense, for the research community, and for the economy. The US
counterpart to Esprit and MITI can be considered to be the informal
network of collaborative relationships that exist involving the
research community, industry, and the funding agencies. An obvious
question is how industry and the community can work together to
improve the efficiency of this mechanism.
On the DoD side, plans of a very general sort are being developed to
address this issue, particularly concerning the creation of incentives
for all sides to participate. For example, flexibility in the data
rights area has enabled larger scale industry involvement in many
areas: the government has often relinquished data rights in return for
concessionary access to products. Senior people in DoD now
increasingly recognize both the role of commercial industry in Defense
and the impact of Defense investment in supporting innovation outside
of Defense.
Within ISTO (and with much informal help from the community), a more
specific technology-focused plan (called SST) is being developed to
address this issue. It is a technology plan (as opposed to a
management plan) and has a distinct technical vision. Most of the
ISTO software activities contribute in one way or another to this
integrating vision, through the usual mixed strategy of investing in
multiple approaches.
Items for discussion are:
(2.1) means by which we can improve both the extent to which our
community collaborates with industry and the quality
of the interactions, and
(2.2) improvements to the ISTO plan.
[3] Technology transfer:
We are under pressure to make our accomplishments more effectively
transferable, including (1) internally in the community, in order to
improve our overall productivity (e.g., even theoreticians need to
know who their "customers" are, even if they are only other
theoreticians), and (2) out from the community, in order for us to
more effectively contribute to the practicing community, and (3) into
the community from practice, in order for us to be more attuned to the
practitioners' perception of their actual needs. The challenge is to
do this without sacrificing the exploratory character of much of our
effort.
Items for discussion are:
(3.1) ways in which we can improve the varous technology
transfer relationships, particularly within the community.
[4] Engineering and components:
Although not all transfer is in the form of engineering results,
special attention must be devoted to this area, since producing an
engineered system is often the most convincing way to sell and convey
an idea. But often the transfer, even within the community, is only
in the form of black box "concept demonstration" systems.
At DARPA we have had bad experiences in the (let us say remote) past
in which we have made huge investments in environment work that
produces, for example, a single monolithic prototype system. The
risk is that the resulting system (1) is not of real interest in the
world of practice because it appears to them as a crufty research
prototype without a clear evolutionary growth path and without a
strong community consensus behind it, and (2) is not of real interest
in the research community because they are not as interested in the
black box functionality as they are in the specific components of the
system they can exploit for their own purposes.
Components can include, for example, (1) specific language notions
and abstraction mechanisms, (2) abstract interface designs, (3)
engineered software components, (4) algorithms and theorems, and so
on, and so on.
This sharing of components is part of what leads to consensus in the
research community concerning both good technical ideas (e.g.,
abstractions for windows and servers) and actual sharable engineering
infrastructure (e.g., X). This consensus formation is an important
element of success in transferring technology. So our goal is to
move towards an interchange of components rather than of black box
systems. This often creates collaborations involving multiple
researchers and industry.
ISTO has been working with the research community to develop an
attitude towards engineering that involves more aggressive
conventionalization and sharing of interfaces and components, along
with explicit investment in more effective support for our own
internal prototyping (or exploratory programming). Explicit
incentives are being put into place in our selection and management
processes.
Obviously, it is easy to push this too far and damage our efforts.
But there is a widespread impression in the community and in ISTO
that we can do better.
Items for discussion are:
(4.1) how to enhance sharing in the community (e.g., through
explicit DARPA investment in a continually evolving
"software laboratory" along with distribution
mechanisms).
[5] Consensus and collaboration:
How can we work together as a community in order to understand in what
technical areas we are in consensus and which specific issues need to
be resolved? Often consensus exists, but is very hard to detect due
to site-specific terminology and culture. Often vast diffferences in
approach arise due to a difference in "operating point" -- the nature
and time of pay-off that is expected. It is fair to say, I think,
that differences of the former sort should be discouraged, while
differences of the latter sort are fine, as long as we are honest
about the operating points.
Items for discussion are:
(5.2) collaboration potential with other software research
efforts, including NSF efforts, STARS and SEI, ONR,
NASA, etc.
[6] Validation:
Validation (in the sense of assessing consistency of product, or
planned product, with actual need) is important in order to ensure we
are working on the right problems and in order to improve links among
the various factions of the community. (For example, those focusing
on trust in operating systems need to ensure that real operating
systems hackers and researchers are involved from time to time in
order to ensure that the right questions are being addressed and in
order to help feed back an awareness of the trust-related issues to
the OS community.)
Again, this can be pushed too far with obvious negative results, but,
again, there is a perception that we can be doing better than we are.
Items for discussion are:
(6.1) opportunities for validation workshops in our various
technical areas.
[7] Collaborations overseas:
It is occasionally possible for DARPA to invest in joint work with
non-US research groups. In many technical areas, there is much to be
gained on both sides through collaborations with Europeans, Japanese,
and others.
Items for discussion are:
(7.1) areas of technical interest for overseas collaboration
that would benefit from ISTO investment.
THE REQUEST. Please send me comments NOW concerning the selection of
these six topics and the discussion items mentioned for each.
Comments are solicited concerning these topics. If you feel that a
particular topic is irrelevant or inappropriate, please say so. If
you feel it is a perfectly good topic, but would not yield productive
discussion for whatever reason, say that too. If a topic is
insufficiently specific, please suggest improvements. Also send
suggestions for new topics.
I'll produce a draft agenda on the basis of your comments.
If you want to include the whole community in your note, say so and
I'll forward it onward. Otherwise, I won't redistribute.
Also, I welcome any private comments or sugggestions that respond
directly to the topics, including the technical topics (1.1) and
(7.1).
Thanks!
Bill
-------
∂27-Oct-88 1740 JK the 91 function
Done. See functx.lsp[ekl,jk]. There were no theoretical problems
with this approach. However, EKL is very awkward to use in case
of recursive expansions that should bottom out and also for
arithmetic facts.
∂27-Oct-88 1852 CLT msg
Please call Japan Travel Bureau
∂27-Oct-88 2341 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM theology question
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 27 Oct 88 23:40:54 PDT
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 330387; Fri 28-Oct-88 02:39:10 EDT
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 73820; Wed 26-Oct-88 02:56:17 PDT
Date: Wed, 26 Oct 88 02:53 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: theology question
To: rwg@RUSSIAN.SPA.Symbolics.COM
cc: macsyma-i@RUSSIAN.SPA.Symbolics.COM, math-fun@RUSSIAN.SPA.Symbolics.COM,
"dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"R@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19881025114600.7.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19881026095316.2.RWG@TSUNAMI.SPA.Symbolics.COM>
Date: Tue, 25 Oct 88 04:46 PDT
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
CLtL defines (ATAN z1 z2) to be an error if either arg is complex.
MACSYMA's ATAN2(z1,z2) is tolerant, but mute. What is a reasonable extension?
One idea is CARG(z2+%I*z1), which is consistent with the current behavior for
real z1, z2. (CARG := complex arg := phase.)
But unlike ATAN, this has the peculiarity to always be real. And it
doesn't preserve the identity ATAN2(SIN(z),COS(z)) = z.
Another approach might be to say that if scaling z1 and z2 by a positive
real is a noop, and scaling by a negative real adds π, then scaling by
1/sqrt(z1↑2+z2↑2) should add carg of this sqrt.
Nah.
(But which?). The scaled
z1 and z2 can now be regarded as the sin and cos of a complex angle, which
is the (rest of the) value of the atan2. This would give a familiar
ATAN2(z1,z2) = ATAN(z1/z2) + fudge. But exactly which fudge?
Salamin convinced me that ATAN2 should be analytic, which means fudge gotta be
an integer times π. Now here comes a swindle that looks too good to be true.
As near as I can tell,
ATAN2(Y,X) = 2 ATAN((SQRT(X↑2+Y↑2)-X)/Y),
for all real (X,Y) ≠ (0,0), except you have to take a limit for X>0, Y→0.
(Just to get 0 !)
So just use this for Y and X complex.
Diversion: define g(x) := (sqrt(x↑2+y↑2)-x)/y. Then g(x) = 1/g(-x). It
would seem that the most general such g would be f(x)/f(-x), for arbitrary f.
Then there must be an f satisfying g(x) = f(x)/f(-x). What is it?
Answer: If you insist on giving up, delete every other character from
xg+(gxg)-↑)(41↑/42x)!
I wonder how many high school math students get problems like this. I wish
I had.
∂28-Oct-88 1128 CLT goto
Is here. He is busy NTTing this week.
I suggested we have dinner with he and Richard next
Tuesday.
∂28-Oct-88 1213 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
A LOCAL FORMALIZATION OF INHERITANCE
Matthew Ginsberg (GINSBERG@POLYA)
Stanford University
Friday, October 28, 3:15pm
MJH 301
Existing work on formalizing inheritance hierarchies suffers from two
significant drawbacks. On the one hand, the formalization is often
nonlocal, in that the translation from the inheritance hierarchy into
the relevant declarative language requires a complete examination of
the topology of the graph being considered. On the other hand, the
formalization frequently uses an inference technique that is tailored
specifically to problems of this sort, and cannot be applied outside
of this very restricted area.
I present a formalization that avoids these difficulties. It is
local, and relies on a very simple generalization of existing
approaches to default reasoning. The generalization is based on a
simple formalization of the notion of causality.
∂28-Oct-88 1634 CLT loan
You need to sign the 3rd part of a 3rd party agreement.
Matt called to see if you would rather go to the
bank or the Stanford realestate office to sign.
I told him you would prefer the latter. Ok?
He will bring it to Stanford sometime Monday.
∂28-Oct-88 1730 pimeet@vax.darpa.mil PI Meeting - Speech
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 28 Oct 88 17:30:49 PDT
Posted-Date: Fri 28 Oct 88 20:30:05-EDT
Received: by vax.darpa.mil (5.54/5.51)
id AA11388; Fri, 28 Oct 88 20:30:07 EDT
Date: Fri 28 Oct 88 20:30:05-EDT
From: Mark Pullen <PIMEET@vax.darpa.mil>
Subject: PI Meeting - Speech
To: jmc@sail.stanford.edu
Cc: pullen@vax.darpa.mil
Message-Id: <594088205.0.PIMEET@VAX.DARPA.MIL>
Mail-System-Version: <VAX-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Dr. McCarthy:
If at all possible, could you please provide me with a title for your
speech on Wednesday at the PI Meeting in Dallas, by reply mail? I am
looking at printing deadling of Monday. I apologize for the short
notice. Thanks for your help.
Juanita Walton, Meeting Coordinator
-------
∂29-Oct-88 0918 CLT today
I will need Timothys car seat to pick him up.
Can you put it in my car after you drop them off?
Also Hazel's bag, which we forgot to put in my car
when I left. The car is in the 40 min zone just
across from mjax.
ps. No additional expeditions today. No point
in getting Timothy so exhausted he can't enjoy the
party.
∂29-Oct-88 1322 DEK
"I would kind of like to be involved with planning the new computer
systems and getting things set up in the new CS buildings." -- Joe Weening
(something he just said to me)
(If there's a way we can keep him around, our computers will surely
run much better than if there isn't...!)
∂30-Oct-88 0359 JSW Electronic mail
I'm interested in your proposed article on electronic mail systems, and
would be glad to offer help such as trying to predict the networking
community's responses.
Here are some of my own views:
1. We need to separate the viewpoint of those on the Arpanet before 1983
(the year of the TCP transition) to others. For us, the set of reachable
destinations has gone up tremendously, but the ease of getting to many of
these and the reliability of delivery is now quite variable. For others,
the situation has improved in every way since they have gone from no email
at all to approximately the same performance level of Arpanet users.
(This is a bit of an exaggeration since there were some non-Arpanet email
networks before 1983.)
2. In discussing Dialnet vs. networks, separation of the hardware and
software issues is important. I don't think you can argue that voice phone
lines are sufficient, unless you want to add dozens of phone lines to
systems that distribute large mailing lists. A single phone line simply
can't handle the bandwidth of these lists. We could pay phone companies
to distribute the lists for us, however.
3. The important issue from users' point of view is managing names and
addresses. (The current nomenclature is a source of some confusion;
"jmc@sail.stanford.edu" is called an address; "jmc" and "sail.stanford.edu"
are names, but the host name is then converted to 10.0.0.11 which is also
called an address. Maybe it's best to call user@host syntax a mailbox.)
The domain name system has been a massive experiment that I think has been
fairly successful. When you mail to a mailbox user@foo.bar.edu it almost
always succeeds (at least in my own experience). The problems arise when
you try to mix names with "%" and "!" and other strange syntaxes, that
were never thought out very well and weren't designed to mix together.
4. There is a lot of ongoing effort underway toward providing solutions to
these problems. The main thing to look at (and possibly worry about) is
the X.400 standard and the ISO protocols. These have the support of most
international phone companies (I'm not sure where the U.S. phone companies
stand), and are being pushed slowly though steadily toward implementation.
This is all on a much larger scale than the academic research community.
I think we are viewed by them as important for pioneering the networking
and email systems, but in the end just another set of customers.
∂30-Oct-88 1558 A.ERIC@GSB-HOW.Stanford.EDU Macintosh virus information
Received: from GSB-HOW.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Oct 88 15:58:48 PST
Date: Sun 30 Oct 88 15:58:17-PST
From: Eric M. Berg <A.Eric@GSB-HOW.Stanford.EDU>
Subject: Macintosh virus information
To: jmc@Sail.Stanford.EDU
Organization: Price Waterhouse Technology Centre, Menlo Park, CA
Phone-#s: 415/322-0606 (PWTC), 415/329-9940 (home)
Message-ID: <12442664649.139.A.ERIC@GSB-HOW.Stanford.EDU>
The other night at dinner you asked about Macintosh viruses. This is
the most complete explanation I've seen so far. /Eric
---------------
Date: Fri, 28 Oct 88 09:16:09 PDT
From: "Jon Dick" <AS.JHD@Forsythe.Stanford.EDU>
Subject: Virus info found on AppleLink
Computer "Viruses"
This article last reviewed: 15 April 1988
GENERAL ISSUES
What is a virus?
A virus is a program with two distinct functions:
- It spreads itself from machine to machine (self-reproducing code).
This includes the actual infection of other systems as well as the
stashing away of code into as many "carriers" as possible.
- It implements the "symptoms" planned by the perpetrator of the
virus. This could be any number of things, up to and including
erasing a disk on a specific date.
A Bit of History
Computer viruses have been around for almost as long as computers.
John Van Neumann, the father of the modern computer, toyed with the
idea of self-reproducing computer code as early as 1948. In the
late 1970s, there was even a training ground for the writing of
viruses. It was a program called Core Wars that implemented an
artificial environment pitting two virus programs against each
other.
Viruses Are Not Unique to the Macintosh
The Macintosh is not the only system to be plagued by viruses.
Mainframe and minicomputers are also targets for virus programmers.
One of the more recent mainframe incidents was the virus that
invaded IBM's mail system and brought it to its knees for a couple
of days. IBM PC users have been experiencing viruses for several
years now. The most common method of attack is through the
COMMAND.COM file. The Macintosh community has been lucky to have
gone so long without virus programming becoming the thing to do.
Not All Viruses Are Meant To Be Damaging, But...
Viruses are not all meant to be damaging. The programmer may just
want to prove he can do it and have the satisfaction of reading
about it in magazines and on the BBS network. Sometimes, these
viruses can cause problems anyway. For example, the virus that has
prompted this series of articles was meant to be benign except in
specific cases. However, it takes up memory and processing time and
has caused random side effects such as printing problems and system
crashes.
Don't Panic; Don't Overreact
If you think that you have a virus, it's important to not overreact.
It is important to take a step back and evaluate the situation
calmly. Once you know that you have a virus and what it has
infected, it is a relatively easy thing to combat. This document
contains enough information for you to deal with most viruses.
Unix Viruses
In all of this, there has not been much discussion of Unix viruses,
but they do exist, and the spread of public domain software is
almost as great in the Unix world as it is in the microcomputer
world.
THE GREAT VIRUS HUNT
When Do You Suspect You Might Have a Virus?
When your computer begins to do things out of the ordinary, or when
it stops being able to do things it has always done in the past.
The problem with this is that corrupted system files can lead to
similar symptoms even though a virus isn't involved. When problems
occur, they are much more likely to be the result of non-virus
difficulties. When you have ruled out the standard problem areas,
you should look into the possibility that your system has been
infected by a virus.
What to Look For If You Think You Have a Virus
Look for invisible files in your System folder that don't belong
there. Unless you specifically have an application that creates
invisible files in the System folder, every invisible file in the
System folder should be suspect. Also, a general check of all the
files in your System folder for resources that don't belong in those
files is well worth the effort.
Files and Resources a Virus Might Infect
- Any and all applications
- HyperCard Stacks (the MacMag virus was spread via a HyperCard
stack)
- Files in the System folder, including:
System
Finder
Note Pad file
Scrapbook file
Clipboard file
Easy Access
Sound
Mouse
Startup Device
Monitors
Color
General
Keyboard
LaserWriter
ImageWriter
AppleTalk ImageWriter
ImageWriter LQ
In other words, all system files.
Files a Virus Might Damage Inadvertently
- Any file on an infected volume or system, including system files,
documents, applications, etc.
Public Domain Issues
Most viruses spread via public bulletin board systems and are hidden
in public domain programs. "Sexy Ladies," a program distributed at
a MacWorld Expo in San Francisco, erased whatever hard disk or
floppy disk it was on when it was launched.
Network Issues
The use of networks can easily enhance the spread of a virus.
Different scenarios are possible, with the simplest being a public
domain folder on a server that everyone gets the latest neat stuff
from. Also, shared applications residing on a server could become
infected, which would then infect every machine that those
applications were run on.
TECHNICAL ISSUES
How Viruses Propagate
Viruses can propagate by a variety of methods. The most common way
for a Macintosh virus to replicate itself is to have an INIT that
installs a background (VBL) task that checks for specific
occurrences, such as a disk insertion, and then copies itself
somewhere to that disk.
VBL Tasks
The Macintosh has always had a limited form of background processing
available to it through the use of the Vertical BLanking queue.
Every time the screen on a Macintosh (except for a Macintosh II) is
refreshed, any routines installed in the queue are executed. The
Macintosh II has a dummy VBL queue for compatibility reasons since
the advent of a variety of screens has led to different vertical
retrace periods.
VBL tasks can be installed in the queue by any program. The program
has to load a routine into a section of memory and install the
routine into the VBL queue by calling the Vinstall ROM routine. It
is the responsibility of the installing program to make sure the
segment of memory containing the routine remains available even
after the program has exited. Each VBL task has a specified time
period it should be left "asleep" before it is called. Every time
the routine is executed, a counter is decremented for that routine.
When that counter reaches zero, the routine is deleted from the
queue unless the routine itself resets the counter.
Lengthy VBL tasks such as the one that might be used to replicate a
virus can interfere with the normal operation of the Macintosh by
interrupting processes that shouldn't be interrupted. A perfect
example of this is printing to a LaserWriter over an AppleTalk
network. If a VBL task takes too long in its execution, the
printing process could terminate abnormally and leave the machine's
connection to the network in an unstable state.
For the purposes of a virus, an INIT is most likely to be the
culprit responsible for installing a VBL task.
INITs
INITs are routines that are run when the Macintosh is booted. For
the most part, they have full access to all of the commands normally
available to a standard Macintosh program. The major difference is
that the low memory globals have not been set up yet, so any INIT
needing access to structures normally stored in low memory must
create its own.
INITs in the System file:
When a Macintosh boots, the INITs in the System file in the
"blessed" folder are the first code to be executed. These INITs
should generally be Apple INITs only -- any non-Apple INITs should
be considered suspect.
The INIT 31 mechanism:
A special INIT in the System file, INIT 31, was created to allow for
the execution of non-Apple INITs without having them installed in
the System file itself. When all of the other INITs in the System
file have been executed, INIT 31 walks through the System folder
looking for files of types INIT, RDEV, cdev, and executes any INIT
resources it finds in these files. The order in which the files get
loaded is alphabetical. Needless to say, a simple way for hiding
parts of a virus is to drop INITs into legitimate files already
existing in the System folder with these file types.
CDEVs
The file type cdev indicates a file containing a Control Panel
device. When the Control Panel is loaded, it walks through the
directory of the System folder looking for any files of type 'cdev'.
When it finds a file of this type, it loads the ICN# of that file
(assuming it has one) into the list of icons shown on the left side
of the Control Panel. When you click on the icon of the cdev in the
Control Panel, the code in the cdev resource in the file of type
'cdev' is executed. A virus could easily use this mechanism as a
way to infect a system, install a VBL task, etc.
Many cdev files have INITs in them with the cdev controlling the
settings that the INIT will use when it is installed. A good
example of this is the settings for a screen blanker. The INIT
actually installs the VBL task, but the cdev controls when dimming
occurs. None of the standard Apple system cdev files have INITs in
them, but there is nothing to prevent a virus installing an INIT in
these files as a way of hiding its code.
DRVRs
DRVR resources typically can have one of two functions: they can be
the code for a desk accessory, or the code for drivers necessary for
the system to perform some function such as printing. Once again,
the key word here is 'code'. Whenever code is involved, the
potential arises for the perpetrator of a virus to take advantage of
it.
Just as with cdevs, when a DRVR gets opened, either by the choosing
of a desk accessory or by the system, code is executed at that
point. This is the stage at which a virus might fulfill its
purpose.
CODE Resources
Each application has at least two CODE resources. The first of
these CODE resources has an id of 0 and contains what is known as
the jump table. This table provides the basic information necessary
for various parts of a program to call routines in other CODE
segments. The current rage in viruses is to modify the CODE ID = 0
resource of an application so that a CODE segment it installs in the
application gets called before the application is actually run.
This CODE segment could go out and check if the virus has infected
the current system, and if it hasn't, install itself. All the
perpetrator of a virus has to do at this point is upload a copy of
an infected application to a BBS, and it spreads across the world.
Applications that allow external procedures:
Viruses could take advantage of the external procedures that are
allowed by some applications. The perfect example of this is
HyperCard, with its XCMDs and XFCNs. This is how the MacMag virus
was transmitted.
KNOWN VIRUSES
The Scores Virus
----------------
You can be almost positive your system has been infected by the
Scores virus if the icons of your Note Pad file and Scrapbook file
look like document icons instead of system icons. Launch ResEdit
and look in your System folder. If you see files called "Desktop"
and "Scores" you can be 99% sure that you have the Scores virus.
How Scores Spreads and What It Does
The Scores virus is relatively harmless. The initial infection is
caused by an application with a modified CODE ID = 0 resource, and
an additional CODE resource (first unused ID number plus 1). When
the 'carrier' application is launched, the CODE ID = 0 resource runs
the virus installer code. This code checks for previous
installation of the Scores virus. If the virus is not there, the
virus files are installed. The virus consists of three INITs, one
atpl, and one DATA resource found in the files listed below:
FILE TYPE CREATOR RESOURCES SIZE
---------------------------------------------------------------------
Desktop (invisible) INIT FNDR atpl ID = 128 2410 bytes
DATA ID = -4001 7026 bytes
INIT ID = 10 1020 bytes
Note Pad File INIT ZSYS INIT ID = 6 772 bytes
Scores (invisible) RDEV ZSYS atpl ID = 128 2410 bytes
DATA ID = -4001 7026 bytes
INIT ID = 10 1020 bytes
Scrapbook File RDEV ZSYS INIT ID = 6 772 bytes
ID = 17 480 bytes
System File ZSYS MACS atpl ID = 128 2410 bytes
DATA ID = -4001 7026 bytes
INIT ID = 6 772 bytes
INIT ID = 10 1020 bytes
INIT ID = 17 480 bytes
---------------------------------------------------------------------
If the Note Pad and Scrapbook files do not exist, they are created.
If they exist, the type and creator of the files are altered to
those listed above, and the corresponding resources are added to the
files. The files still appear to function normally with the Note
Pad and Scrapbook DAs, but their icons change to document icons.
The Desktop and Scores files are invisible, and are created during
the infection process.
The next time the infected system is rebooted, the INITs are loaded
into memory and are ready to infect other applications. The INITs
install a VBL task that actually modifies and installs resources
into an application. After an application has been launched, an
internal timer is started. Somewhere between two and three minutes
later, the open application is infected and becomes a carrier. A
new CODE resource is added to the infected application, and the
application's CODE ID = 0 resource is modified to execute the new
CODE resource first, then continues with the application.
To determine if an application is infected, examine the CODE ID = 0
resource. If the eleventh word of the resource (third word on the
third line in the ResEdit listing) is NOT "0001", the application is
suspect. If the third word is something other than "0001", convert
the value to its decimal equivalent (the numbers are in
hexadecimal). Then determine the resource number of the CODE
resource at the top of the ResEdit resource list. If these numbers
are the same, the application is probably infected, and should be
replaced. Some applications will appear to be infected even though
they are not. If the eleventh word of CODE ID = 0 is not 1, check
the tenth word; if it is '4EED' the application is most likely not
infected.
How to Get Rid of the Scores Virus
It is not hard to remove this virus from a system, but it may take
some time. Here's how:
1. Use Font/DA Mover to copy all fonts and DAs that you do not have
backups of to font and DA suitcase files (this virus does not
attach itself to DAs).
2. Start the system from a locked, not infected, floppy disk.
3. Throw away the System folder on the infected disk.
4. Use ResEdit to identify all suspect applications on the infected
disk.
5. Make a list of all suspect applications.
6. Throw all suspect applications in the trash, and empty the
trash.
7. Reinstall the system software from a known good System Tools
installer disk.
8. Using locked masters, recopy any applications that were deleted
from the infected disk (it is important to verify that the
master disks have not been infected).
9. You're all done.
The nVIR Virus
--------------
How the nVIR Virus Spreads and What It Does
The nVIR virus is similar to the Scores virus in many ways. It does
not appear to have malicious intent and is relatively harmless.
Initial infection of a system is also caused by an application with
a modified CODE ID = 0 resource. When a nVir carrier application is
launched, the virus' code segment is executed first. This code
checks for its INIT in the System File, and if it doesn't find it,
the code copies the INIT there. Along with the INIT resource, eight
'nVIR' resources (0-7) are added to the System file.
The next time the system is restarted, the INIT ID = 32 is loaded
into memory and tries to infect every application that is launched.
The nVir virus adds a CODE ID = 256 resource and modifies the CODE
ID = 0 so that the nVir code is executed first.
Again, infection of an application is determined by examination of
the CODE ID = 0 resource. If the eleventh word of the resource
(third word on the third line in the ResEdit listing) is NOT "0001",
the application is suspect. If the third word is something other
than "0001", convert the value to its decimal equivalent (the
numbers are in hexadecimal). Then determine the resource number of
the CODE resource at the top of the ResEdit resource list. If these
numbers are the same, the application is probably infected, and
should be replaced. Some applications will appear to be infected
even though they are not. If the eleventh word of CODE ID = 0 is
not 1, check the tenth word; if it is '4EED' the application is most
likely not infected. The tenth word normally contains '3F3C'.
When launching an infected application, there is a one in sixteen
chance that you will hear a short system beep. We have been told
that if MacinTalk is installed you will hear the words "don't
panic".
How to Get Rid of the nVIR Virus
Remove the nVIR virus the same way you remove the Scores virus
except you do not need to throw away all of the files in the System
Folder; just throw away the System file.
The MacMag Virus
----------------
We don't have much information regarding the MacMag virus. It was
apparently uploaded to CompuServe, inside a HyperCard stack, in the
form of an XCMD, and it installed an INIT ID = 6 with a name of
'RR'. Its sole purpose in life was to display a "universal message
of peace" on your computer on March 2, 1988. The virus removed
itself after displaying this message and should be of little concern
now.
SAFEGUARDING YOUR SYSTEMS
What Makes Our System Susceptible to Viruses
The various mechanisms described in part 2 of this article make our
system easy to infiltrate by a virus. Remember that it is those
same mechanisms that add to the flexibility and "look and feel" of
the Macintosh. For instance, the INIT mechanism is used by mail
systems to load their code in. AppleShare uses the INIT mechanism
to mount network volumes at boot time.
Why Vaccine Works in This Case, But Is Easy to Bypass
Vaccine, a public domain INIT written to block viruses, does a good
job of alerting you when the three known viruses are trying to
infect your system. The problem with Vaccine: once a cure is found
for one set of viruses, a new strain may appear that knows how to
bypass the existing defenses.
Some Suggestions
- Lock your master diskettes
Always keep original "Master" disks locked. This prevents a virus
from spreading to your original disks. Our disk locking mechanism
is hardware based -- viruses can't infect locked disks!
- Protect your networks
Network administrators should not allow just anyone to put software
on the server. Applications on a network server should come only
from known good masters.
- Be wary of public domain software
Public domain software should be checked quite thoroughly on a
floppy- based system for any infections before being copied to a
hard disk based system. This will also protect you from any "Trojan
Horse" programs such as "Sexy Ladies."
- Quarantine infected systems
If you identify a system as being infected with a virus, immediately
isolate (quarantine) it from other systems. This means
disconnecting it from any network and not allowing anyone to take
any files from the exposed system to another system. Once the
system has been 'disinfected,' you can allow the files to be copied
or moved.
- Use ResEdit
ResEdit is a good tool to look for viruses on your disks. There is
very little that can be hidden from ResEdit, so you can use it to
remove troublemaking files and resources.
-------
∂31-Oct-88 0807 Mailer re: those whales
Received: from SUMEX-AIM.Stanford.EDU by SAIL.Stanford.EDU with TCP; 31 Oct 88 08:07:16 PST
Date: Mon, 31 Oct 88 08:07:09 PST
From: Mark Crispin <Crispin@SUMEX-AIM.Stanford.EDU>
Subject: re: those whales
To: JMC@SAIL.Stanford.EDU
cc: SU-etc@SAIL.Stanford.EDU
In-Reply-To: <PpCIO@SAIL.Stanford.EDU>
Postal-Address: P.O. Box 2652; Seattle, WA 98111-2652 USA
Message-ID: <12442841023.14.CRISPIN@SUMEX-AIM.Stanford.EDU>
Perhaps the Bell Labs whale meat wasn't prepared properly. I've eaten
whale in Japan, both cooked and as sashimi. It's tender and quite tasty.
I know of two places in Tokyo where you can get whale; Kujira-Ya near
Shibuya station and a new place in the Kabuki-cho entertainment district in
Shinjuku.
As for which individuals in the presidential campaign to dispatch, I
would say that the current offering of mediocrities is perfect. It reflects
the general level of mediocrity that America and the American people have
worked so hard to attain in the past two decades.
-------
∂31-Oct-88 0856 VAL Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
ARGUMENT SYSTEMS:
A UNIFORM BASIS FOR NONMONOTONIC REASONING
Fangzhen Lin (lin@polya)
Yoav Shoham (shoham@score)
Stanford University
Friday, November 4, 3:15pm
MJH 301
We introduce argument systems. The key notions are inference rules, arguments,
argument structures, and completeness conditions. Inference rules are
primitive relations between premises and conclusions. Arguments are trees of
inference rules and are used to establish propositions. Argument structures
are aggregations of arguments and must satisfy some conditions in order to be
viewed as a candidate of supporting a set of beliefs held by a rational agent.
Completeness conditions capture when our knowledge is complete.
All of the notions are simple and natural. It turns out that all of the major
existing nonmonotonic logics can be viewed as special argument systems. By
showing this we are able to unveil some common features of existing
nonmonotonic logics. For example, all of them are presumptive. It also
suggests that a generalized "negation as failure" rule may be useful in
implementing these logics.
∂31-Oct-88 1033 RPG
∂30-Oct-88 2309 JMC randomness
"An even worse example can be found in the 1985 LISP text by Gabriel [10]
which uses a = 17 and m = 251. Again, the multiplier is not a primitive
root of the modulus and the resulting period in this case is just 125."
Yes, I saw this. The interesting thing is that the book explains that
the generator is designed to have the same behavior in all Lisp implementations
with 16 bit 2's complement arithmetic and is called exactly 100 times.
In fact, I mention that the alternative I considered was a list of 100 numbers.
I'm debating whether to answer them.
∂31-Oct-88 1209 RPG
∂31-Oct-88 1200 JMC reply to message
[In reply to message rcvd 31-Oct-88 10:33-PT.]
I suggest you don't bother. If I understand you correctly, since the
LISP benchmarks aren't genuine Monte Carlo problems, it doesn't matter
if the random number generator isn't too good. It might be worthwhile,
however, to take the advice in the article for the random number generator
in Lucid's Common Lisp.
I think we do take their advice already. In any event, our random guy
is reading the article to see what it says.
∂01-Nov-88 1055 GC.TLX@Forsythe.Stanford.EDU MSG RCVD FOR J. MCCARTHY
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 1 Nov 88 10:55:42 PST
Date: Tue, 1 Nov 88 10:55:44 PST
To: jmc@sail
From: "MESSAGE CENTR 3-4081" <GC.TLX@Forsythe.Stanford.EDU>
Subject: MSG RCVD FOR J. MCCARTHY
RX-TLX 0501 PST 11/01/88
RCA NOV 01 0659
STANFRD STNU A
727442 UNIVED
TLX REF4267 88-11-01 12:00
TO : PROF. JOHN MCCARTHY
DEPT OF COMPUTER SCIENCE
STANFORD UNIVERSITY
DEAR PROF. MCCARTHY,
I AM A RESEARCH FELLOW IN THE PROGRAMME ON INFORMATION AND
COMMUNICATION TECHNOLOGY AT THE UNIVERSITY OF EDINBURGH, WHERE
I AM INVOLVED IN A RESEARCH PROJECT ON THE DEVELOPMENT OF PARALLEL
COMPUTING. THE PROJECT IS FUNDED BY THE BRITISH ECONOMIC AND
RESEARCH COUNCIL, AND THE OBJECT IS TO WRITE IN 'REAL TIME' THE
HISTORY OF PARALLEL COMPUTERS AND PARALLEL PROGRAMMING. THE
OBJECT IS NOT TO EVALUATE THE DIFFERENT PROJECTS OF APPROACHES,
BUT TO WRITE THE HISTORY OF THESE DEVELOPMENTS AS THEY HAPPEN.
I SHALL BE IN CALIFORNIA AND OREGON DURING THE MONTHS OF NOVEMBER
AND DECEMBER, TALKING TO PEOPLE INVOLVED IN ONE WAY OR ANOTHER WITH
PARALLEL COMPUTING AND VISITING DIFFERENT PROJECTS. I SHALL
BE IN THE PALO ALTO AREA FROM 28TH NOVEMBER TO 5TH DECEMBER, AND
WOULD BE EXTREMELY INTERESTED IN TALKING TO YOU ABOUT YOUR WORK.
BROADLY SPEAKING, I AM INTERESTED IN LEARNING MORE ABOUT YOUR WORK ON
THE DEVELOPMENT OF QLISP, THE SPECIFIC MERITS OF THE APPROACH BEING
ADOPTED, THE PROGRESS OF THE WORK AND THE MAJOR PROBLEMS ENCOUNTERED.
I WOULD ALSO BE INTERESTED IN DISCUSSING THE RELATION BETWEEN CURRENT
CONTROVERSIES IN THE AREA OF PARALLEL PROGRAMMING AND EARLIER
CONTROVERSIES BETWEEN DIFFERENT APPROACHES IN SEQUENTIAL PROGRAMMING.
THE PRODUCT OF THE RESEARCH WILL BE A NUMBER OF REPORTS AND A BOOK
ON THE DEVELOPMENT OF PARALLELISM. OF COURSE, I WOULD NOT QUOTE
FROM YOU WITHOUT YOUR PERMISSION, AND ANY PASSAGE DESCRIBING YOUR
WORK WOULD BE SENT TO YOU FOR COMMENT BEFORE PUBLICATION.
IF YOU HAVE NO OBJECTION, I SHALL CONTACT YOU AS SOON AS I ARRIVE IN
CALIFORNIA, TO SEE IF YOU CAN SPARE THE TIME TO SEE ME AND TO ARRANGE
A TIME WHEN WE COULD MEET.
I SHALL CONTACT YOU IN A COUPLE OF WEEKS' TIME AND HOPE TO HAVE THE
PLEASURE OF MEETING YOU.
YOURS SINCERELY
DR. ELOINA PELAEZ
RESEARCH CENTRE FOR SOCIAL SCIENCES
UNIVERSITY OF EDINBURGH.
STANFRD STNU A
727442 UNIVED
To: JMC@SAIL
∂01-Nov-88 1201 iris@cive.STANFORD.EDU getting ahold of Dr. John Sowa
Received: from cive.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 12:01:18 PST
Received: by cive.STANFORD.EDU (1.2/Ultrix2.0-B)
id AA02790; Tue, 1 Nov 88 12:01:55 pst
Date: Tue, 1 Nov 88 12:01:55 pst
From: iris@cive.STANFORD.EDU (Iris Tommelein)
Message-Id: <8811012001.AA02790@cive.STANFORD.EDU>
To: jmc@sail
Subject: getting ahold of Dr. John Sowa
Cc: iris@cive.STANFORD.EDU
Professor McCarthy,
I am trying to contact Dr. Sowa by e-mail.
He taught CS309A here in the fall of last year, and I hear
that you might have arranged for that, so that you may know
his current e-mail address.
Would you be able to give dr. Sowa's e-mail to me,
or would you know how I could reach him my mail or phone?
Thank you very much,
Iris
∂01-Nov-88 1207 bill@gatech.edu Need information on Dany Guindi
Received: from gatech.edu by SAIL.Stanford.EDU with TCP; 1 Nov 88 12:06:11 PST
Received: by gatech.edu (5.58/GATECH-8.0)
id AA01605 for ; Tue, 1 Nov 88 14:55:21 EST
Date: Tue, 1 Nov 88 14:55:21 EST
From: bill@gatech.edu (Bill Appelbe)
Message-Id: <8811011955.AA01605@gatech.edu>
To: jmc@sail.stanford.edu
Subject: Need information on Dany Guindi
Dear John:
Dany Guindi, whom we are informally considering for an academic
appointment at Georgia Tech., has listed you as the advisor on his PhD
committee at Stanford. Before we proceed further, would you be willing
to give an informal opinion (either by E-Mail, or phone) on his academic
progress and research potential?
Cheers -- Bill
Bill Appelbe, Hiring Committee Chair
Associate Professor, School of ICS, Georgia Tech.
(404)-894-6187
(bill@gatech.edu)
∂01-Nov-88 1217 rivin@polya.Stanford.EDU preprints
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 12:17:38 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA08875; Tue, 1 Nov 88 12:17:40 PDT
Date: Tue, 1 Nov 88 12:17:40 PDT
From: Igor Rivin <rivin@polya.Stanford.EDU>
Message-Id: <8811012017.AA08875@polya.Stanford.EDU>
To: jmc@sail
Subject: preprints
Cc: rivin@polya.Stanford.EDU
I will arrange to havwe some preprints transported to your door. I feel
rather silly, not having talked to you about all this work earlier, but
better late than never, they say. I am out east interviewing with places,
but will be ba back soon. If you have any questions/comments, I am of course
always online.
Thanks,
Igor.
∂01-Nov-88 1316 VAL Ablex
[In reply to message rcvd 31-Oct-88 21:34-PT.]
Barbara Bernstein, Sales and Promotions Manager, (201)767-8450.
Production Manager, Carol Davidson, probably Spring
Formalizing Common Sense
∂01-Nov-88 1327 BYRD@SUMEX-AIM.Stanford.EDU Re: anti-racism
Received: from SUMEX-AIM.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 13:26:06 PST
Date: Tue, 1 Nov 88 13:25:44 PST
From: Greg Byrd <Byrd@SUMEX-AIM.Stanford.EDU>
Subject: Re: anti-racism
To: JMC@SAIL.Stanford.EDU
cc: Byrd@SUMEX-AIM.Stanford.EDU, su-etc@SUMEX-AIM.Stanford.EDU
In-Reply-To: <1Hqq#u@SAIL.Stanford.EDU>
Message-ID: <12443161164.32.BYRD@SUMEX-AIM.Stanford.EDU>
The right, led at present by Reagan Republicans, has shown no capacity
for distinguishing opposition to its own politcal agenda from a lack
of patriotism.
...Greg
-------
∂01-Nov-88 1500 G.GSB@MACBETH.STANFORD.EDU Re: anti-racism
Received: from MACBETH.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 15:00:31 PST
Date: Tue 1 Nov 88 14:55:12-PST
From: Warren Redlich <G.GSB@MACBETH.STANFORD.EDU>
Subject: Re: anti-racism
To: JMC@SAIL.STANFORD.EDU
In-Reply-To: <1Hqq#u@SAIL.Stanford.EDU>
Message-ID: <12443177453.86.G.GSB@MACBETH.STANFORD.EDU>
< I fear that we
< will have required courses on ``other cultures'' which will
< amount to indoctrination with the students pressured into
< parrotting the views of the professors in order to get good
< grades.
In what way would that be different, and worse than what we
have now? Would it be worse to have students indoctrinated in the
views of those professors than in the views of professors of current
culture courses? If you don't think that this happens under the
current system, why do you think that it will in the system proposed?
WAR
-------
∂01-Nov-88 1510 BLUMENTHAL@A.ISI.EDU Final Touches
Received: from A.ISI.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 15:10:12 PST
Date: Tue 1 Nov 88 17:20:47-EST
From: Marjory Blumenthal <BLUMENTHAL@A.ISI.EDU>
Subject: Final Touches
To: duane.adams@cs.cmu.edu, dongarra@MCS.ANL.GOV,
gannon%rdvax.dec@decwrl.dec.com, gossard@cadlab2.mit.edu,
hearn@rand-unix.arpa, jlh@vsop.stanford.edu, jmc@sail.stanford.edu,
mchenry%guvax.bitnet@cunyvm.cuny.edu, ouster@ginger.berkeley.edu,
ralston@mcc.com, thornj@max.acs.washington.edu, CWeissman@dockmaster.arpa,
troywil@ibm.com
cc: blumenthal@A.ISI.EDU
Message-ID: <12443171185.40.BLUMENTHAL@A.ISI.EDU>
I have received a number of positive responses to the conclusions
material distributed last week. If you have any other responses,
I must have them by noon on Friday, and the earlier the better. What
follows are (1) a brief executive summary, that can be printed on
two pages, single spaced, and (2) an alternative cut on the
conclusions. The latter is based on a conversation yesterday
with a committee member who wanted a greater action orientation.
The fact that the committee did not study bureaucratic aspects of
the control problem, per se, limits what it can say but please
review the following as a more focused alternative.
Again, any feedback is needed ASAP.
Regards...Marjory
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -1-
11/01/1988
EXECUTIVE SUMMARY
_________________
The United States and other countries of the Coordinating Committee
for Multilateral Export Control (CoCom) are maintaining a substantial
qualitative and quantitative lead in computer technology over the USSR and
other countries of the Council for Mutual Economic Assistance (CMEA). In
many areas the lead is on the order of five to ten years of more. Export
controls have thus far contributed to this gap, but not only will current
technological progress make controls harder to enforce, technological and
market developments also combine to make a case for a more focused and
flexible control regime. This report, commissioned by the Department of
State, presents recommendations for achieving such a regime.
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -2-
11/01/1988
Technical trends make control more difficult and computer technology
_____________________________________________________________________
continues to develop rapidly. Steady reduction in computer and component
____________________________
size makes increasingly sophisticated hardware both more portable and easy
to hide. Professional workstations and other small computers will soon be
offering performance capabilities for some applications that were
previously available only in conventional large supercomputers. Small,
high-power computers have already boosted the lead of CoCom countries in
scientific computing, but the proliferation of ever more powerful small
systems will undermine control efforts. This situation could be
aggravated by the movement toward parallel processors which can be built
using small low-technology computers in large numbers. The value of
parallel processors depends on the development of appropriate software,
however, and CoCom countries will continue to lead in that area.
Software is taking on growing importance in computer systems and it is
inherently easy to acquire. Scientific software is particularly difficult
to control because it is so widely available in the scientific community.
Software development tools are, along with computer manufacturing systems,
making possible technologies critical to CoCom's computing lead.
More and increasingly powerful hardware and software will become
commodities. Commodity products are available in high volume and at low
cost, they may be available in multiple and substitutable forms, and they
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -3-
11/01/1988
tend to be small and easy to transport. These attributes make commodities
vital to the economic health of the computer industry, but also
effectively, uncontrollable.
International computer networking is flourishing among businesses and
researchers and is an implicit conduit for technology exports. Because
computer networks are used to communicate technical information about
software and hardware, computer networking probably represents the fastest
growing gap between development and decision in current export control
strategy.
Technologies are more than products and knowhow is the key to using
____________________________________________________________________
them. The computing advantage is fundamental to CoCom countries. CMEA
____
countries are weak in computer manufacturing because they lack both
adequate equipment and the knowhow necessary for volume production of
high-quality products. Similar problems plague CMEA scientific
computing. Nevertheless, the knowhow of the West may not be sufficiently
protected by export controls, and increasing standardization of computer
technologies will make it even harder to control.
Commercial Vitality is Essential to Technical Vitality. The
______________________________________________________
development of computer technologies and their commercial markets in CoCom
countries are closely intertwined, and this relationship is fundamental to
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -4-
11/01/1988
the CoCom technical lead. CoCom commercial computer technology is best
overall, followed by CoCom military computer technology, then CMEA
military computer technology, with CMEA commercial computer technology
last overall.
Unlike past decades, CoCom military establishments now have more to
gain from than give to the commercial computing technology base.
Consequently, U.S. policymakers must be concerned with the impact of
control options on the domestic computer sector as well as their impact on
the CoCom lead overall.
The United States cannot afford to be complacent about its computer
technology strengths or base export control decisions on an assumption of
invincible lead. As the computer market becomes increasingly global, U.S.
firms face increasing foreign competition, most from firms operating with
fewer export barriers under the same CoCom guidelines. Tighter U.S.
controls may reflect the absence of a fully-effective multilateral control
effort, but there is a risk that in the computer arena, the United States
may lock the proverbial barn door after the horse has escaped. And if the
U.S. competitive position in computer technologies erodes, tighter U.S.
controls will only find the United States devoting too much effort to
controlling technologies available in equal or better form elsewhere.
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -5-
11/01/1988
Non-CoCom Countries are also Expanding the Supply of Computer
______________________________________________________________
Technology. The growing production and use of computer technology among
__________
countries, outside of both CoCom and CMEA,is another reason why computer
technology is becoming much more readily available and therefore harder
for CoCom to control around the world. Newly-industrializing countries,
especially those in Asia, are a major source of technology that the
traditional CoCom-CMEA dichotomy fails to capture.
CMEA Prospects in computer technology and Improving Strengthening but
______________________________________________________________________
still Clouded. CMEA weaknesses reflect three factors in particular:
_____________
export controls, Western firms' perception that market opportunities are
limited in CMEA, and self-imposed constraints in CMEA countries. Change
in each factor will facilitate absorption of new computer technology in
CMEA countries--and each factor is changing in ways that will promote
technology transfer. But while this prospect creates more pressure on
export controls to hold the line, there are countervailing pressures to
keep export controls in harmony with the technological and international
facts of life. The two need not be inconsistent.
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -6-
11/01/1988
Recommendations
_______________
Recommendation 1: Computer technologies captured on the list of CoCom
________________
controlled products should have flexible definitions that account for
technology change, market developments, and variations among technologies
that might be colloquially labeled the same.
Recommendation 2: All computer technologies should not be treated alike.
________________
Categories should be distinguished based on potential benefit to CMEA and
factors affecting the usefulness of different systems within categories.
Recommendation 3: A computer technology should be treated in export
________________
control regimes as a commodity and recognized as effectively
uncontrollable if either it is readily available from foreign sources
outside CoCom control or if certain other factors obtain.
Recommendation 4: Network security should be improved to prevent computer
________________
networks from becoming a channel for significant covert technology
transfer while still allowing CoCom countries to benefit from networks.
Recommendation 5: Further study should be undertaken to address several
________________
computer network-related issues, including access to CoCom research
networks, transborder flow of technical information, and the application
of the International Traffic in Arms Regulations to network security
products for commercial applications.
Recommendation 6: The U.S. government should nurture the U.S. lead in
________________
computer networking technology, and part of that effort should include
exploring options for easing licensing for export of noncritical network
technologies to other CoCom nations.
Recommendation 7: The U.S. export control strategy should account for both
________________
technical factors making control more difficult and commercial factors
making control more burdensome, and focus on technologies of compelling
military importance that could enable substantial advances in CMEA
countries or that represent key leverage points in the increasingly
interdependent world of computer technology. Chapter 7 lists priorities.
Recommendation 8: Options for restricting the distribution of source code
________________
with software outside of the United States should be explored.[encryp?]
Recommendation 9: Software should be divided into three principal classes
________________
for control purposes. Software with a compelling and direct military
usefulness should be tightly controlled; some degree of control should be
provided for software tools that could build software in the first class;
but other software should be traded freely among CoCom nations.
Recommendation 10: Better and ongoing monitoring of computer technology
_________________
and market developments worldwide should be instituted to improve export
control decisionmaking and to better inform U.S. technologists.
Recommendation 11: Computer technology trends should be evaluated
_________________
periodically, at least every three years.
Recommendation 12: Review of definitions and categorizations of controlled
_________________
technologies should be more timely, rapid, and expert.
PROPRIETARY INFORMATION DO NOT QUOTE, OR CITE -7-
11/01/1988
INSERT FOR PCS
While the RISC architecture discussed in the section on workstations
may eventually be applied to PCs, this is not likely to happen until the
early 1990s. The enormous base of software developed for non-RISC,
complex instruction set computing (CISC) architectures and microprocessors
and their associated operating systems (e.g., MS/DOS) and the absence of a
comparable resource for PC-level RISC machines will slow the transition.
sorry about the above paragraph!
conrvw.doc
Recommendation #1: Definitions for computer technologies on the list of
_________________
controlled products should be made more flexible to account for technology
_
change, market developments, and variations among technologies that might
be colloquially labeled the same.
A key example of the need for flexibility involves supercomputers, which
are subject to case-by-case export restrictions. Defining supercomputers
is so controversial that recent trade legislation called for an official
definition of supercomputers for purposes of export control. At issue is
the capability--the physical representation may vary and will change
relatively quickly. The committee believes that a relative approach to
categorizing these machines (e.g., the n percent most powerful as measured
_
by generally accepted benchmark tests and/or as used in solving critical
classes of problems) will work better than any absolute definition or
label. Although technical progress will alter decisions about which
machines fall into such a category, the committee does not recommend
automatic decontrol after some period of time for those machines that are
superceded. Decontrol decisions must take into account both the advancing
level of CMEA technology and, as discussed below, whether the technology
has become a commodity. On the other hand, the committee recommends
relaxing controls on trade and access among CoCom [and other non-CMEA]
countries for obsolescing high-performance computers.
Recommendation #2: Delete.[all not alike]
_________________
Recommendation #3: The U.S. government should establish and publish a
_________________
list of computer technologies that are commodities, and it should
promulgate a policy that exempts such commodities from controls for trade
among CoCom nations. A computer technology should be identified and
treated in export control regimes as a commodity if the technology is
readily available from foreign sources outside CoCom control (a condition
that currently is cause for reducing export restrictions, although this is
often not done in a timely manner), or if other factors (e.g., high
volume, low price, small size, ready availability of substitutes) make the
technology effectively uncontrollable. To implement this recommendation
the government must establish a mechanism to identify the point at which a
technology becomes a commodity; this is critical in allowing U.S.
manufacturers to enter the market early.
Recommendation #4: The U.S. government should formulate a policy for
_________________
preventing network technology from becoming a channel for significant
covert technology transfer and to protect the computational resources of
CoCom countries. While in many cases the necessary technology exists,
putting it to use may require further study or change in existing policy.
For example:
[Recommendation #5: Deleted as separate item and fold examples into new
_________________
4]
Recommendation #6: Delete as implicit in #3.
_________________
Recommendation #7: The Department of State should work with other
_________________
agencies to cus export control effort on computer technologies of
compelling military importance that could enable CMEA countries to make
substantial gains in their technology base, or those that represent key
leverage points in the increasingly interdependent world of computer
technology.
In addition to military-specific technology (e.g., VHSIC or on-board
fire control systems), priority items should include: [as before]
Recommendation #8: Delete as weakly supported.
_________________
Recommendation #9: Same [software in 3 classes]
_________________
Recommendation 9a [NEW]: The Department of State should promote the
_______________________
integration of key Asian NICs (including Singapore, Hong Kong, Taiwan,
and Korea) into the CoCom control program.
Recommendation #10: The U.S. government should greatly increase its
__________________
investment in the monitoring of computer technology development and
associated market trends around the world. Although the intelligence
community monitors developments in CMEA countries, the committee
recommends that more comprehensive attention (i.e., addressing
commercial as well as military applications) be paid on an ongoing
basis to developments around the world, especially in non-CMEA,
non-CoCom nations (e.g., newly-industrializing countries in Asia and
Latin America).
Although the Department of Commerce monitors foreign availability and
performs competitive assessments, its limited resources appear to be
stretched quite thin, and it has only issued reports on selected
technologies on a very infrequent basis. Moreover, at this writing it
had lost funding (and was seeking funds) for one of its more active
information gathering mechanisms, a computer technology-watcher based
in Europe. To monitor global technology trends well, the government
must invest in perhaps 100 or more additional, highly skilled people
who are knowledgeable not only in computer technologies and their
applications but also in international market trends and foreign
languages. The necessary monitoring would not only facilitate a
focusing of export controls, it would also benefit U.S. computer
companies and researchers. To benefit from technology watchers
scattered among agencies, coordination and planning for this effort
should be provided by a lead entity with a suitably broad mission.
Recommendation #11: The Department of State should undertake periodic
__________________
reviews of technology trends along the lines this report. The rapid
change in computer technology makes trend presentations perishable and
this type of review should be conducted at least every three years.
Recommendation #12: The Departments of Commerce, Defense, and State
__________________
should review definitions or categorizations of controlled technologies
in a manner that is more timely and rapid, as well as more expert. The
committee recognizes that advisory committees are already in use; but
we are concerned that review of control decisions is neither timely nor
rapid, and evidence suggests that more expertise may be needed to reach
the best decisions and put them into effect.
-------
∂01-Nov-88 1809 G.GSB@MACBETH.STANFORD.EDU Re: anti-racism
Received: from MACBETH.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 18:09:29 PST
Date: Tue 1 Nov 88 18:04:06-PST
From: Warren Redlich <G.GSB@MACBETH.STANFORD.EDU>
Subject: Re: anti-racism
To: JMC@SAIL.STANFORD.EDU, su-etc@MACBETH.STANFORD.EDU
In-Reply-To: <1Hqq#u@SAIL.Stanford.EDU>
Message-ID: <12443211841.66.G.GSB@MACBETH.STANFORD.EDU>
< I fear that we
< will have required courses on ``other cultures'' which will
< amount to indoctrination with the students pressured into
< parrotting the views of the professors in order to get good
< grades.
Will the professors of these "other cultures" courses
be any more indoctrinating than those who teach the regular
culture courses we have now?
WAR
-------
∂01-Nov-88 1824 G.GSB@MACBETH.STANFORD.EDU re: anti-racism
Received: from MACBETH.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 18:24:31 PST
Date: Tue 1 Nov 88 18:18:59-PST
From: Warren Redlich <G.GSB@MACBETH.STANFORD.EDU>
Subject: re: anti-racism
To: JMC@SAIL.STANFORD.EDU
In-Reply-To: <lqsFS@SAIL.Stanford.EDU>
Message-ID: <12443214548.66.G.GSB@MACBETH.STANFORD.EDU>
I fear indoctrination for several reasons.
1. The motivation of the proposals for required courses is that people
don't have the beliefs they should.
2. Many of the people demanding required courses will regard them
as tokens unless they indoctrinate.
3. If opinion polls taken after the courses, show that students don't
have the correct views, there will be demand to indoctrinate even more.
4. The people who will want jobs teaching these courses have shown
little respect for freedom of opinion. Many of them regard all education
as indoctrination and only want it to be their kind of indoctrination.
--------------------
Perhaps you should try to understand that which you fear.
In point 1, you make a sweeping generalization about course proposals,
in areas where I might guess, you have studied little, if at all. I have
studied little of these things myself. The motivation for these course
proposals is quite often to teach people about different cultures, not
to indoctrinate them into a set of beliefs. Even if you disagree, how
is it different from indoctrination in Western Culture?
The problems with your conclusions in 2 and 3 stem in part from
this assumption of point 1.
In 4, you again make a sweeping generalization about people who
you (again, I guess) don't know, at least not very well. One flaw is that
the people who are clamoring to teach these courses already have jobs.
They are already Humanities (and other fields) faculty. They will not
teach any greater quantity of courses, and will thus note be "indoctrinating"
any more than they could already in their current courses.
Go easy on the sweeping generalizations.
WAR
P.S.: Feel free to post both your message and mine on su-etc. I made a
mistake in MM earlier today.
-------
∂01-Nov-88 1948 SHOHAM@Score.Stanford.EDU russians
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 19:48:36 PST
Date: Tue 1 Nov 88 19:47:47-PST
From: Yoav Shoham <SHOHAM@Score.Stanford.EDU>
Subject: russians
To: jmc@Sail.Stanford.EDU
Message-ID: <12443230716.14.SHOHAM@Score.Stanford.EDU>
John,
The IJCAI program committee is interested in names of good AI people
in the USSR, who could possibly serve as referees. It seems like they
are bneing allowed to send in many papers now, and should be included
in the process. Can you suggest names, specialities, and means of
contact?
Yoav
-------
∂01-Nov-88 2146 SHOHAM@Score.Stanford.EDU re: russians
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Nov 88 21:46:21 PST
Date: Tue 1 Nov 88 21:42:38-PST
From: Yoav Shoham <SHOHAM@Score.Stanford.EDU>
Subject: re: russians
To: JMC@Sail.Stanford.EDU
In-Reply-To: <10qBQ8@SAIL.Stanford.EDU>
Message-ID: <12443251622.14.SHOHAM@Score.Stanford.EDU>
He referred me to you. There was one guy he had in mind (academy member,
I think) whose name he couldn't recall.
-------
∂01-Nov-88 2351 qphysics-owner@neat.ai.toronto.edu lethargy?
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 1 Nov 88 23:51:16 PST
Received: from [128.100.1.65] by RELAY.CS.NET id aa07587; 1 Nov 88 21:24 EST
Received: by neat.ai.toronto.edu id 8346; Tue, 1 Nov 88 21:11:25 EST
Resent-From: qphysics-owner@ai.toronto.edu
Resent-To: qphysics-distribution@ai.toronto.edu
Resent-Sender: qphysics-owner@ai.toronto.edu
Received: from localhost (stdin) by neat.ai.toronto.edu with SMTP id 8342; Tue, 1 Nov 88 21:11:10 EST
To: qphysics@ai.toronto.edu
Subject: lethargy?
Date: Wed, 2 Nov 88 00:30:06 EST
From: Jean-Francois Lamy <lamy@ai.utoronto.ca>
Message-Id: <88Nov1.211110est.8342@neat.ai.toronto.edu>
Resent-Message-Id: <88Nov1.211125est.8346@neat.ai.toronto.edu>
Resent-Date: Tue, 1 Nov 88 21:11:12 EST
Just a reminder that qphysics@ai.utoronto.ca is patiently waiting for your
input, Surely *someone* has published a tech report that they could
announce... Machine-readable bibliographies could be made available for
retrieval, etc. (in addition to the mail server we can now support FTP from
the whole Internet on neat.ai.toronto.edu, 128.100.1.65). Abstract from
relevant talks are often a way to make contact with researchers with similar
interests.
Who knows, maybe someone could ask a question or emit a point of view that
would get (gak!) a discussion going...
Jean-Francois Lamy lamy@ai.utoronto.ca, uunet!ai.utoronto.ca!lamy
AI Group, Department of Computer Science, University of Toronto, Canada M5S 1A4
-------------------------------8< couper ici 8<--------------------------------
Submissions to
qphysics@ai.utoronto.ca or qphysics@ai.toronto.edu (all networks)
failing that, try
qphysics@utai.uucp or uunet!utai!qphysics (UUCP)
qphysics%ai.utoronto.ca@relay.cs.net (broken ARPA)
qphysics@ai.toronto.cdn (EAN X.400)
lamy@utorgpu (Bitnet/EARN)
Replace qphysics with qphysics-request to get at the moderator.
∂02-Nov-88 0633 BLUMENTHAL@A.ISI.EDU re: Final Touches
Received: from A.ISI.EDU by SAIL.Stanford.EDU with TCP; 2 Nov 88 06:33:35 PST
Date: Wed 2 Nov 88 08:06:44-EST
From: Marjory Blumenthal <BLUMENTHAL@A.ISI.EDU>
Subject: re: Final Touches
To: JMC@SAIL.STANFORD.EDU
cc: BLUMENTHAL@A.ISI.EDU
In-Reply-To: <$qtO#@SAIL.Stanford.EDU>
Message-ID: <12443332468.27.BLUMENTHAL@A.ISI.EDU>
John, we needed you at our July meeting... --Marjory
-------
∂02-Nov-88 0850 SHOHAM@Score.Stanford.EDU re: russians
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Nov 88 08:49:57 PST
Date: Wed 2 Nov 88 08:46:11-PST
From: Yoav Shoham <SHOHAM@Score.Stanford.EDU>
Subject: re: russians
To: JMC@Sail.Stanford.EDU
In-Reply-To: <lqCC0@SAIL.Stanford.EDU>
Message-ID: <12443372417.38.SHOHAM@Score.Stanford.EDU>
Thanks
-------
∂02-Nov-88 0959 MPS paper on soviet access....
I forgot what you told me to do with this.
Was it for xeroxing?
Pat
∂02-Nov-88 1204 CLT calendar item
fri 4 nov 8:00 Matt - mortgage forms
∂02-Nov-88 1738 Mailer re: anti-racism
To: JMC@SAIL.Stanford.EDU, RTC@SAIL.Stanford.EDU,
su-etc@SAIL.Stanford.EDU
From: Robert W Floyd <RWF@SAIL.Stanford.EDU>
[In reply to message from JMC rcvd 02-Nov-88 16:59-PT.]
Someone should check `wowser' in Mencken's own book,
The American Language.
∂02-Nov-88 1740 kar@polya.Stanford.EDU Reminder: Applications AI questions needed by November 7th
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Nov 88 17:40:32 PST
Received: from LOCALHOST by polya.Stanford.EDU (5.59/25-eef) id AA03094; Wed, 2 Nov 88 17:40:27 PDT
Message-Id: <8811030140.AA03094@polya.Stanford.EDU>
To: shoham@score, feigenbaum@sumex, genesereth@score, nilsson@score, jmc@sail,
val@sail, subramanian@score, myers@polya, grove@polya
Cc: kar@polya.Stanford.EDU
Subject: Reminder: Applications AI questions needed by November 7th
Date: Wed, 02 Nov 88 17:40:25 -0800
From: kar@polya.Stanford.EDU
On behalf of the Applications Comp. Committee I am writing to you
to solicit questions for the AI section of the Applications Comp.
which is to be held next January.
The text for this section is the whole of Elaine Rich's book
"Artificial Intelligence". (Note that there is a separate AI
section on the Theory comp. based on Genesereth and Nilsson's book.)
You are invited to submit *one question* (or more if you feel
enthusiastic) by November 7th. Please give hard copies to either
myself or Prof. Wiederhold. (I can sometimes be found in room 450
of MJH). An accompanying sketch solution would also
be appreciated, as the solutions are being prepared alongside the
exam for this comp. Please ensure that the questions remain
confidential.
Thank you for your cooperation.
Ken Ross.
∂02-Nov-88 1751 RFC Prancing Pony Bill
Prancing Pony bill of JMC John McCarthy 2 November 1988
Previous Balance 12.12
Monthly Interest at 1.0% 0.12
Current Charges 4.00 (bicycle lockers)
-------
TOTAL AMOUNT DUE 16.24
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 7/12/88.
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-Nov-88 0700 pehoushe@Gang-of-Four.Stanford.EDU [elloyd@NOTE.NSF.GOV: NSF support for algorithms and parallel computing systems]
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 07:00:31 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04927; Thu, 3 Nov 88 06:59:52 PST
Date: Thu, 3 Nov 88 06:59:52 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8811031459.AA04927@Gang-of-Four.Stanford.EDU>
To: jmc@sail, clt@sail, rpg@sail
Subject: [elloyd@NOTE.NSF.GOV: NSF support for algorithms and parallel computing systems]
Return-Path: <@polya.Stanford.EDU:THEORYNT%NDSUVM1.BITNET@forsythe.stanford.edu>
Date: Wed, 2 Nov 88 14:24:35 CST
Reply-To: TheoryNet List <THEORYNT@NDSUVM1.BITNET>,
Errol Lloyd <elloyd@NOTE.NSF.GOV>
Sender: TheoryNet List <THEORYNT@NDSUVM1.BITNET>
Comments: Warning -- original Sender: tag was THEORYNT@YKTVMX
From: Errol Lloyd <elloyd@NOTE.NSF.GOV>
Subject: NSF support for algorithms and parallel computing systems
To: Local Distribution <aflb-tn@POLYA.STANFORD.EDU>
-------------------------------------------------------------------------
PRELIMINARY ANNOUNCEMENT
Beginning in FY89, NSF and DARPA will jointly support research projects
in the area of algorithm design, analysis and instrumentation for parallel
computing systems. The purpose of the program is to promote closer ties
between theory and practice. By enabling the highest quality efforts to
operate at an increased scale, the program aims to encourage collaborative
efforts that link theoretical computer science with experimentation and
engineering. Specific focal points are:
Parallel algorithms and computational models,
Analysis and instrumentation techniques for complex parallel models,
Parallel algorithm design paradigms, and
Design paradigms for parallel artificial intelligence algorithms.
The research proposed under this joint program should investigate
innovative approaches and techniques that may lead to revolutionary advances
in the state of the art. Specifically excluded are approaches that are
primarily incremental improvements to the existing state of practice.
Specific areas of interest include, but are not limited to, the following:
Parallel algorithms and data structures,
Probabilistic and randomized algorithms,
Parallel models of computation including cellular automata,
Computational geometry,
Algorithm design paradigms,
Mechanizable algorithm analysis techniques,
Instrumentation techniques,
Very high level languages for expressing parallel algorithms,
Quantitative analysis methods for heuristics, and
Design paradigms for algorithms in vision, speech, planning and learning.
Excluded from this program is work of a primarily foundational nature in
structural complexity, combinatorial mathematics and algorithm analysis, as
well as primarily application-specific algorithm design and implementation.
Approximately $1.5 Million will be awarded in FY89. Proposals for efforts
of any size will be considered. In general, support will be provided for
principal investigators, graduate students, postdoctoral research support, and
specialized equipement necessary for the conduct of the research, as well as
other funds normally allowed in an award. Proposals under this program will
be subject to the normal NSF peer review process. Special emphasis in the
review process will be given to the value gained from team research and the
capability of the plan for achieving transition of results into the research
and/or industrial communities. NSF and DARPA staffs will make final selections
from meritorious proposals. Proposals should be submitted to NSF as if they
were regular NSF submissions (for consideration by: NSF-DARPA Parallel
Computing Initiative, CCR-CISE). For technical information, prospective PIs
may contact either NSF or DARPA program offices:
DARPA - Dr. William L. Scherlis, Program Manager, Defense Advanced
Research Projects Agency, (202) 694-5800, scherlis@vax.darpa.mil
NSF - Dr. Errol L. Lloyd, Division of Computer and Computation Research,
(202) 357-7375, elloyd@note.nsf.gov
Target date for submitting proposals is December 19, 1988.
∂03-Nov-88 0929 MPS Pigott Account
Your balance on this account is 2329.49.
∂03-Nov-88 0957 SHOHAM@Score.Stanford.EDU CSLI evening seminars
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 09:57:34 PST
Date: Thu 3 Nov 88 09:52:15-PST
From: Yoav Shoham <SHOHAM@Score.Stanford.EDU>
Subject: CSLI evening seminars
To: jmc@Sail.Stanford.EDU
Message-ID: <12443646589.10.SHOHAM@Score.Stanford.EDU>
John,
CSLI is initiating a series of small, informal seminars. Participants
will be CSLI and SSP faculty, and a few invited others. The organizing
committee consists of Barwise and Etchemendy (phil), Rumelhart (psych),
Sag (ling), myself and Betsy Macken. We expect a total of about 20
people. The intention is to form an intellectual forum, characetrized by
informal discussion, intersciplinarity, and high quality of participants.
We will meet the first and third Wednesday of each month, possibly
starting on Novermber 16. It will be a 7-9 pm affair, prefixed by
wine and cheese. We seek a serious commitment from the participants,
which means primarily regular attendance. I have been asked to invite
you to join, and hope you will.
Yoav
-------
∂03-Nov-88 1041 SHOHAM@Score.Stanford.EDU re: CSLI evening seminars
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 10:41:19 PST
Date: Thu 3 Nov 88 10:38:25-PST
From: Yoav Shoham <SHOHAM@Score.Stanford.EDU>
Subject: re: CSLI evening seminars
To: JMC@Sail.Stanford.EDU
In-Reply-To: <1dr9qv@SAIL.Stanford.EDU>
Message-ID: <12443654995.23.SHOHAM@Score.Stanford.EDU>
That's great. I'll relay the news.
-------
∂03-Nov-88 1056 scales@polya.Stanford.EDU research mentor
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 10:56:36 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA03632; Thu, 3 Nov 88 10:56:32 PDT
Date: Thu, 3 Nov 88 10:56:32 PDT
From: Daniel J. Scales <scales@polya.Stanford.EDU>
Message-Id: <8811031856.AA03632@polya.Stanford.EDU>
To: jmc@sail
Subject: research mentor
Professor McCarthy:
I am a first-year Ph.D. student and am currently looking around for a
"Research Mentor" group. I would like to find more about the work
going on in your QLISP project. Is there still a slot open that might
be appropriate for a first-year student to fill? Could I meet with
you sometime to talk about the work? Or is there a student I should
talk to first? My best times are Monday, Wednesday, and Friday
afternoons.
My own interests are somewhat ill-defined. I got an MSAI degree from
Stanford in 1986 working with Paul Rosenbloom. My thesis was on
improving the performance of the Rete matching algorithm for the kinds
of production rule sets that are built in SOAR. However, after
graduation, I worked for two years at Sun Microsystems on a project
developing an environment intended to help solve programming-in-the-large
problems. Now back at Stanford as a Ph.D. student, I am unsure
whether I will pursue AI or more systems-oriented research.
I currently have a departmental RAship, so I assume that would mean
that I would still be supported by the department for work I did for
you.
Dan Scales
∂03-Nov-88 1058 VAL
I remembered that you had never had a chance to chat with my father. Would you
like to join us for lunch today? He's on campus now.
∂03-Nov-88 1159 @b.NSF.GOV:mzemanko@note.nsf.gov Proposal review
Received: from note.nsf.gov by SAIL.Stanford.EDU with TCP; 3 Nov 88 11:59:21 PST
Received: from b.nsf.gov by note.nsf.gov id aa26257; 3 Nov 88 14:36 EST
To: jmc@SAIL.STANFORD.EDU
bcc:
Subject: Proposal review
Date: Thu, 03 Nov 88 14:33:34 -0500
From: Maria Zemankova <mzemanko@note.nsf.gov>
Message-ID: <8811031433.aa02665@b.nsf.gov>
Dear Professor McCarthy:
I am taking the liberty of sending you a proposal by Ron Yager
entitled "Possibility Theory in Eexpert Systems" that has been
submitted to the NSF Database and Expert Systems Program.
Since the author is using your work as one of the starting points
in his approach, I would very much appreciate your reviewing this
proposal.
I know you are a very busy scientist, but since you are THE
leading authority in the area of non-monotonic reasoning, your
review would be of great value. If you cannot review the proposal,
I wouldappreciate your looking at it and suggesting alternative
reviewers. Should you not be able to do this, I would appreciate
your letting me know, so that I can try to engage somebody else.
Thank you for your assistance.
Sincerely,
Maria Zemankova, Director
Database and Expert Systems Program
mzemanko@note.nsf.gov
202-357-9570
∂03-Nov-88 1206 weening@Gang-of-Four.Stanford.EDU Information on the "virus"
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 12:06:05 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA05666; Thu, 3 Nov 88 12:05:42 PST
Date: Thu, 3 Nov 88 12:05:42 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811032005.AA05666@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Information on the "virus"
A number of reporters have been calling you about a virus that spread
last night over the Internet. (They all say "Arpanet", but it has
actually spread throughout many of the networks connected to the
Arpanet as well.) This bug has definitely arrived at Stanford and
affected Unix hosts including Polya, Gang-of-Four, and many others.
It spreads through the mail system. Someone discovered a loophole
whereby a Unix system that receives mail of a particular type can be
made to execute a shell script, thus causing the virus to store itself
in the filesystem and propagate by sending out further messages. I
haven't yet heard of any damage done by the virus other than raising
the load on the machines that it hits.
Below is a message describing it in detail.
Newsgroups: news.announce,news.sysadmin
Date: Thu, 3 Nov 88 02:58:55 PST
From: bostic@okeeffe.Berkeley.EDU (Keith Bostic)
Subject: Virus (READ THIS IMMEDIATELYππ)
Approved: spaf@cs.purdue.edu
Distribution: world
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Subject: Fixes for the virus
Index: usr.lib/sendmail/src/srvrsmtp.c 4BSD
Description:
There's a virus running around; the salient facts. A bug in
sendmail has been used to introduce a virus into a lot of
Internet UNIX systems. It has not been observed to damage the
host system, however, it's incredibly virulent, attempting to
introduce itself to every system it can find. It appears to
use rsh, broken passwords, and sendmail to introduce itself
into the target systems. It affects only VAXen and Suns, as
far as we know.
There are three changes that we believe will immunize your
system. They are attached.
Thanks to the Experimental Computing Facility, Center for
Disease Control for their assistance. (It's pretty late,
and they certainly deserved some thanks, somewhere!)
Fix:
First, either recompile or patch sendmail to disallow the `debug'
option. If you have source, recompile sendmail after first
applying the following patch to the module svrsmtp.c:
*** /tmp/d22039 Thu Nov 3 02:26:20 1988
--- srvrsmtp.c Thu Nov 3 01:21:04 1988
***************
*** 85,92 ****
"onex", CMDONEX,
# ifdef DEBUG
"showq", CMDDBGQSHOW,
- "debug", CMDDBGDEBUG,
# endif DEBUG
# ifdef WIZ
"kill", CMDDBGKILL,
# endif WIZ
--- 85,94 ----
"onex", CMDONEX,
# ifdef DEBUG
"showq", CMDDBGQSHOW,
# endif DEBUG
+ # ifdef notdef
+ "debug", CMDDBGDEBUG,
+ # endif notdef
# ifdef WIZ
"kill", CMDDBGKILL,
# endif WIZ
Then, reinstall sendmail, refreeze the configuration file,
using the command "/usr/lib/sendmail -bz", kill any running
sendmail's, using the ps(1) command and the kill(1) command,
and restart your sendmail. To find out how sendmail is
execed on your system, use grep(1) to find the sendmail start
line in either the files /etc/rc or /etc/rc.local
If you don't have source, apply the following patch to your
sendmail binary. SAVE A COPY OF IT FIRST, IN CASE YOU MESS
UP! This is mildly tricky -- note, some versions of strings(1),
which we're going to use to find the offset of the string
"debug" in the binary print out the offsets in octal, not
decimal. Run the following shell line to decide how your
version of strings(1) works:
/bin/echo 'ππππππππabcd' | /usr/ucb/strings -o
Note, make sure the eight control 'G's are preserved in this
line. If this command results in something like:
0000008 abcd
your strings(1) command prints out locations in decimal, else
it's octal.
The patch script for sendmail. NOTE, YOUR OFFSETS MAY VARY!!
This script assumes that your strings(1) command prints out
the offsets in decimal.
Script started on Thu Nov 3 02:08:14 1988
okeeffe:tmp {2} strings -o -a /usr/lib/sendmail | egrep debug
0096972 debug
okeeffe:tmp {3} adb -w /usr/lib/sendmail
?m 0 0xffffffff 0
0t10$d
radix=10 base ten
96972?s
96972: debug
96972?w 0
96972: 25701 = 0
okeeffe:tmp {4} ↑D
script done on Thu Nov 3 02:09:31 1988
If your strings(1) command prints out the offsets in octal,
change the line "0t10$d" to "0t8$d".
After you've fixed sendmail, move both /bin/cc and /bin/ld to
something else. (The virus uses the cc and the ld commands
to rebuild itself to run on your system.)
Finally, kill any processes on your system that don't belong there.
Suspicious ones have "(sh)" or "xNNNNNNN" where the N's are random
digits, as the command name on the ps(1) output line.
One more thing, if you find files in /tmp or /usr/tmp that
have names like "xNNNNNN,l1.c", or "xNNNNNN,sun3.o", or
"xNNNNNNN,vax.o" where the N's are random digits, you've been
infected.
∂03-Nov-88 1253 weening@Gang-of-Four.Stanford.EDU Virus update
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 12:52:55 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA05829; Thu, 3 Nov 88 12:52:37 PST
Date: Thu, 3 Nov 88 12:52:37 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811032052.AA05829@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Virus update
While Gang-of-Four received the messages, it doesn't seem to have had
any effect. Several explanations that I've seen say it only affects
Vaxes and Suns, though I don't yet understand why.
My own speculation at this point is that it was done by someone who
wanted to demonstrate a weakness of BSD Unix systems in a dramatic
way, without causing actual damage. News of the virus and how to
avoid it, and fix infected systems, is quickly being spread to Unix
system administrators.
The chief danger that I see at this point is that someone can use the
same idea to do actual damage to systems before they are repaired.
Also, people may think of similar ways to attack Unix systems and
cause damage without advance warning.
∂03-Nov-88 1412 clark@sm.unisys.com
Received: from rdcf.sm.unisys.com (SM.UNISYS.COM) by SAIL.Stanford.EDU with TCP; 3 Nov 88 14:12:33 PST
Received: by rdcf.sm.unisys.com (sdcrdcf) (5.54/Domain/jpb/2.9)
id AA02168; Thu, 3 Nov 88 14:14:20 PST
Date: Thu, 3 Nov 88 14:14:20 PST
From: clark@sm.unisys.com (Clark Weissman)
Message-Id: <8811032214.AA02168@rdcf.sm.unisys.com>
To: Blumenthal@a.isi.edu, JMC@sail.stanford.edu, clark
John:
Enjoyed your essay. Here are some further thoughts.
At one of the Export meetings you missed, we got into some
discussions related to your essay. The point I particularly found
challenging had to do with USSR not having the foreign exchange
to buy much from the West even if export controls were relaxed.
Therefore, finding ways to deal with establishing favorable rates
of exchange:Rubbles to dollars, is an important dimension for
West to explore. Also, Western industries must be allowed to
remove earnings from USSR at these favorable exchange rates, else
the improved trade will always be favorable to the USSR.
These thoughts are relevant to your items 12, 15, and Summary. You decide.
A 3rd summary item might be:
3. The West needs to consider what USSR trade barriers to US
export need to be relaxed; e.g., food, travel, textiles,
literature, etc., and what items the WEST wants from USSR.
Clark
∂03-Nov-88 1720 VAL Nonmonotonic Reasoning Seminar: Reminder and correction
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
This week we meet in Room 252, not 301 as was announced earlier.
ARGUMENT SYSTEMS:
A UNIFORM BASIS FOR NONMONOTONIC REASONING
Fangzhen Lin (lin@polya)
Yoav Shoham (shoham@score)
Stanford University
Friday, November 4, 3:15pm
MJH 252
We introduce argument systems. The key notions are inference rules, arguments,
argument structures, and completeness conditions. Inference rules are
primitive relations between premises and conclusions. Arguments are trees of
inference rules and are used to establish propositions. Argument structures
are aggregations of arguments and must satisfy some conditions in order to be
viewed as a candidate of supporting a set of beliefs held by a rational agent.
Completeness conditions capture when our knowledge is complete.
All of the notions are simple and natural. It turns out that all of the major
existing nonmonotonic logics can be viewed as special argument systems. By
showing this we are able to unveil some common features of existing
nonmonotonic logics. For example, all of them are presumptive. It also
suggests that a generalized "negation as failure" rule may be useful in
implementing these logics.
∂03-Nov-88 1832 Mailer re: Article on meat eating
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 18:32:41 PST
Received: by sierra.STANFORD.EDU (3.2/4.7); Thu, 3 Nov 88 18:30:45 PST
From: singh@sierra.STANFORD.EDU (Harinder J. Singh)
Date: Thu, 3 Nov 1988 18:30:44 PST
To: John McCarthy <JMC@SAIL.Stanford.EDU>
Cc: singh@SIERRA.Stanford.EDU, su-etc@SAIL.Stanford.EDU
Subject: re: Article on meat eating
In-Reply-To: Your message of 03 Nov 88 1756 PST
Prof. McCarthy says:
``
My remarks on chimpanzees eating meat were based on Jane Goodall's
noticing, some years after she began observing chimpanzees, that they
occasionally caught and ate small monkeys. I don't believe she claimed
that eating meat was important for their health. My guess is that
animals sometimes subject to shortage of food or shortage of certain
components of food will survive better the wider their choice of
diet.''
Jane Goodall's is only one set of observations. Even she
took many years to observe the `occasional' non-veg episodes. What
does that tell us about the predominant aspect of chimp cuisine?
What about the points I made about the 95%-5% composition of the
diet of chimpanzees? Along the lines of your argument, the folks in
the Donner Party are known to have eaten human flesh due to
``shortage of food.'' Those and other observations of cannibalism,
by your reasoning, should lead us to believe that we're designed to
eat humans too. [Assuming that the eatee is already dead, why
not? Why the hue and cry after the survivors return to civilization?]
``
The conclusion that Ramapithecus didn't eat meat when it could
get it still seems unsupported...''
It may have been no more than a plausibility argument,
as opposed to rigorous proof, and meant to be factored in with
the rest of the evidence. I have no problem with conceding that
humans and their ancestors can frequently get by on flesh-foods.
More importantly, what do you have to say about the length of
the human intestinal tract, as compared to that of carnivores?
While we're about it, how about addressing the rest of
the facts brought to bear on the comparisons, eg the dental
structure?
``.... Others believe, perhaps on somewhat better evidence, that
substantial human meat use depended on fire and on tools for
butchering carcasses.''
What others? What `perhaps on somewhat better evidence'?
If that ain't pure and unadulterated conjecture I don't know what
is :-) Mebbe `Just So' stories are fine so long as they support
the good Professor's position!
``... By the way, I was taken to a vegetarian restaurant in Peking.
The imitations of meat were realistic in appearance, but didn't taste
very good to me.''
It is a major mistake, often made by carnivores, to assess
vegetarian food through imitations of meat. An imitation can only
try to approximate whatever it is imitating - if that is your sole
exposure to vegetarian cuisine then one has to chalk up the subsequent
evaluation to ignorance! Who knows, maybe you were reacting to bad
and unfamiliar Chinese food. You're entitled to your preferences of
taste, for sure, but if you make claims to fairness then you need to
go to some of the better vegetarian restaurants in the area. Sampling
imitations of meat isn't a substantive basis - it's a joke.
``Finally, I don't see that facts about Ramapithecus are relevant about
whether a person should become a vegetarian. If primitive humans
lived entirely on meat, you could still decide that humanity has
advanced morally to the point where eating meat should be stopped.
If they never ate meat, you could still regard eating meat as an
advance that should not be given up.''
Finally some agreement. I agree that this Rama-whoever-dude
ought not to be a _major_ issue in the discussion. This week is the
first I ever heard of him and his antics. The Rama-something thing
can be a useful _adjunct_ to an examination of the issues but is only
one small fragment of the overall picture.
``I decline a formal confrontation on the subject...''
It is everyone's prerogative to choose to engage in one
or another discussion or attend to any further information that
is relevant. [I, for instance, couldn't care less what's been goin'
on with the AI boondoggle :-)]
But Professor, by ignoring the available facts that
contradict your favorite position on any subject other than
the couple you listed as open for discussion (AI etc), you
do greatly undermine your credibility for future discourse.
Not that scholarship or careful thought are any prerequisites
for holding forth on su-etc on any given subject :-)
Happy times,
I.
∂03-Nov-88 1937 RLM@Score.Stanford.EDU Red meat fun with McDonald's
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 3 Nov 88 19:37:16 PST
Date: Thu 3 Nov 88 19:36:18-PST
From: Robert L. Miller <RLM@Score.Stanford.EDU>
Subject: Red meat fun with McDonald's
To: su-etc@Score.Stanford.EDU
cc: jmc@Sail.Stanford.EDU, singh@Sierra.Stanford.EDU, rlm@Score.Stanford.EDU,
p.redlich@GSB-WHY.Stanford.EDU, phil@Score.Stanford.EDU
Message-ID: <12443752913.12.RLM@Score.Stanford.EDU>
I called up McDonald's outpost at the Stanford Shopping Center and
asked nicely to speak to the manager. Granted, the this is not a
piercing subject, but the dialog was fun. It went something like
this:
"Do you serve red meat?"
"Red meat?"
"Yes, red meat. Beef."
"On our hamburgers?"
(pause)
"Just a minute..." The manager put the phone down. A minute or two
passed. I could hear rumblings of fast food action...
"Hello?"
"Yes."
"It says 100% U.S. Beef on the box."
"I see. Can you tell me how McDonald's feels about the vegetarian
movement? Do you feel threatened by it?"
"Just a minute..."
Phone down.
"Hello," a new, stronger voice came on. "This is the manager."
"I thought I was just speaking to the manager."
"You were... I'm the other manager."
"Oh. Can you tell me if McDonald's feels threatened by the growing
vegetarian movement? I've been told that..."
"We can't really comment on that, but you're welcome to call our
corporate office at... 408 922-0990."
I thought about asking them if they wanted to cater half a debate. I'd
try to rope in Hobee's for the other half.
I put'eth down the ball.
RLM
-------
∂04-Nov-88 0718 @b.NSF.GOV:mzemanko@note.nsf.gov Re: Proposal review
Received: from note.nsf.gov by SAIL.Stanford.EDU with TCP; 4 Nov 88 07:18:34 PST
Received: from b.nsf.gov by note.nsf.gov id ab01815; 4 Nov 88 10:01 EST
To: John McCarthy <JMC@SAIL.STANFORD.EDU>
Subject: Re: Proposal review
In-reply-to: Your message of 03 Nov 88 16:16:00 -0800.
<hrtCB@SAIL.Stanford.EDU>
Date: Fri, 04 Nov 88 09:55:05 -0500
From: Maria Zemankova <mzemanko@note.nsf.gov>
Message-ID: <8811040955.aa06199@b.nsf.gov>
Thank you very much for your valuable assistance.
Maria Zemankova
∂04-Nov-88 1655 betsy@russell.Stanford.EDU CSLI Stanford Faculty Meeting
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 4 Nov 88 16:55:47 PST
Received: by russell.Stanford.EDU (4.0/4.7); Fri, 4 Nov 88 16:58:07 PST
Date: Fri 4 Nov 88 16:58:06-PST
From: Betsy Macken <BETSY@CSLI.Stanford.EDU>
Subject: CSLI Stanford Faculty Meeting
To: jmc@SAIL.STANFORD.EDU
Message-Id: <594694686.0.BETSY@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Would you please plan to attend a CSLI Stanford Faculty meeting on
Thursday, 17 November, at 4:00 in the Cordura Conference Room?
Stanley will discuss at least the following: a report from the
Research Committee about possibilities for government funding, CSLI's
Industrial Affiliates Program, and CSLI's Advisory Board (not the same
as our Advisory Panel). All of these topics affect the future life of
CSLI, and we need your ideas and your support. Please put it on your
calendar.
Thanks,
Betsy
PS
I'll order a special tea for that day.
-------
∂05-Nov-88 0910 CLT You haven't sent me the number of the
Inamori foundation -- or some such place for
Hazel to call if need be.
∂05-Nov-88 1012 CLT thanks
yes, you sent the hotel name
∂05-Nov-88 1117 singh@sierra.STANFORD.EDU re: Mud-slinging and JMC's heroes [was Re: more Coverup]
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 5 Nov 88 11:16:57 PST
Received: by sierra.STANFORD.EDU (3.2/4.7); Sat, 5 Nov 88 11:14:56 PST
From: singh@sierra.STANFORD.EDU (Harinder J. Singh)
Date: Sat, 5 Nov 1988 11:14:55 PST
To: John McCarthy <JMC@sail.stanford.edu>
Subject: re: Mud-slinging and JMC's heroes [was Re: more Coverup]
In-Reply-To: Your message of 04 Nov 88 1835 PST
Which one(s) do you think?
∂07-Nov-88 0429 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Bessel, yo is my worry now
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 7 Nov 88 04:29:27 PST
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 332446; Mon 7-Nov-88 07:27:53 EST
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 74697; Mon 7-Nov-88 04:23:35 PST
Date: Mon, 7 Nov 88 04:20 PST
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Bessel, yo is my worry now
To: math-fun@RUSSIAN.SPA.Symbolics.COM, "ilan@score.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19881107122047.8.RWG@TSUNAMI.SPA.Symbolics.COM>
Can anybody give >1 term of an expansion at x = 0 of
J[1+1/x](1/x)/J[1/x](1/x) ?
A physicist at BU had a continued fraction which he thought was
1 - k x↑(1/3) + . . ., but it came out to this Bessel ratio.
∂07-Nov-88 0630 ghh@confidence.Princeton.EDU AI as CS and the scientific epistemology of the common sense world
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 7 Nov 88 06:30:45 PST
Received: from clarity.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA15678; Mon, 7 Nov 88 09:30:30 EST
Received: by clarity.Princeton.EDU (3.2/1.63)
id AA27699; Mon, 7 Nov 88 09:32:32 EST
Date: Mon, 7 Nov 88 09:32:32 EST
From: ghh@confidence.Princeton.EDU (Gilbert Harman)
Message-Id: <8811071432.AA27699@clarity.Princeton.EDU>
To: JMC@SAIL.STANFORD.EDU (John McCarthy)
In-Reply-To: JMC@SAIL.STANFORD.EDU's message of 1 Nov 88 05:54:00 GMT
Subject: AI as CS and the scientific epistemology of the common sense world
Where can I read your paper, "Ascribing Mental Qualities to Machines"?
Gil
Gilbert Harman
Princeton University Cognitive Science Laboratory
221 Nassau Street, Princeton, NJ 08542
ghh@princeton.edu
HARMAN@PUCC.BITNET
∂07-Nov-88 0954 betsy@russell.Stanford.EDU re: CSLI Stanford Faculty Meeting
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Nov 88 09:53:59 PST
Received: by russell.Stanford.EDU (4.0/4.7); Mon, 7 Nov 88 09:56:19 PST
Date: Mon 7 Nov 88 09:56:18-PST
From: Betsy Macken <BETSY@CSLI.Stanford.EDU>
Subject: re: CSLI Stanford Faculty Meeting
To: JMC@SAIL.STANFORD.EDU
Message-Id: <594928578.0.BETSY@CSLI.Stanford.EDU>
In-Reply-To: <$r$Bd@SAIL.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
I'm sorry you won't be there. Thanks for letting me know though.
Have you used your office? How do you like it?
Betsy
-------
∂07-Nov-88 1000 eaf@sumex-aim.stanford.edu [davies@cascade.Stanford.EDU (Byron Davies) : Knowledge quantum ]
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 7 Nov 88 10:00:11 PST
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
id AA01609; Mon, 7 Nov 88 10:00:27 PST
Date: Mon, 7 Nov 1988 10:00:26 PST
From: Edward A. Feigenbaum <eaf@sumex-aim.stanford.edu>
To: ai.lenat@mcc.com, jmc@sail.stanford.edu
Subject: [davies@cascade.Stanford.EDU (Byron Davies) : Knowledge quantum ]
Message-Id: <CMM.0.88.594928826.eaf@sumex-aim.stanford.edu>
Return-Path: <davies@cascade.Stanford.EDU>
Received: from cascade.Stanford.EDU by sumex-aim.stanford.edu (4.0/inc-1.0)
id AA23400; Sun, 6 Nov 88 17:39:55 PST
Received: by cascade.Stanford.EDU (5.59/inc-1.0)
id AA12048; Sun, 6 Nov 88 17:40:23 PDT
Date: Sun, 6 Nov 88 17:40:23 PDT
>From: davies@cascade.Stanford.EDU (Byron Davies)
Message-Id: <8811070140.AA12048@cascade.Stanford.EDU>
To: Feigenbaum@SUMEX.stanford.edu
Subject: Knowledge quantum
Is there a standard term for an elementary particle of knowledge?
"Rule", "fact", "assertion", "concept", "symbol", or anything else I
can think of is too wedded to a particular technology for representing
knowledge.
-- Byron
∂07-Nov-88 1146 TAJNAI@Score.Stanford.EDU Seeking sponsor for Visiting Scholar
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 7 Nov 88 11:46:15 PST
Date: Mon 7 Nov 88 11:37:12-PST
From: Carolyn Tajnai <TAJNAI@Score.Stanford.EDU>
Subject: Seeking sponsor for Visiting Scholar
To: Feigenbaum@SUMEX-AIM.Stanford.EDU, genesereth@Score.Stanford.EDU,
shoham@Score.Stanford.EDU, jmc@Sail.Stanford.EDU,
engelmore@SUMEX-AIM.Stanford.EDU
cc: hiller@Score.Stanford.EDU
Message-ID: <12444714270.36.TAJNAI@Score.Stanford.EDU>
Prof. Nilsson received a letter from Dr. Fumihiko Mori, Senior Researcher,
Hitachi inquiring about a visiting scholar position for Mr. Hideo Ohata.
''Mr. Ohata received a B.Eng. in 1980 and an M.Eng. in 1982 both from
Kyoto University. He has been engaged in R&D of AI software.
Major professional experience so far concerns application of knowledge
engineering to real world problems such as those in banking business.
In his view, accentuation on fundamental research themes like
advanced knowledge representation and reasoning to deal with common sense,
analogy, induction, etc., and knowledge acquisition aid combined with
learning facilities is getting more and more needed to put AI technology
further ahead into real world application successfully."
If you will consider sponsoring this person, we will get more information
and him. We also need to find space (a desk).
The CSD fee is $40,000/per year. The faculty sponsor will receive
$35,500 ($500/Forum; $4K/CSD). KSL has a higher rate, and if KSL is
interested, then I will negotiate that figure.
Please let me know.
Carolyn
-------
∂07-Nov-88 1420 scherlis@vax.darpa.mil PI Meeting
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 7 Nov 88 14:19:54 PST
Received: from sun45.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA11028; Mon, 7 Nov 88 16:28:22 EST
Posted-Date: Mon 7 Nov 88 16:29:20-EST
Received: by sun45.darpa.mil (5.54/5.51)
id AA00653; Mon, 7 Nov 88 16:29:21 EST
Date: Mon 7 Nov 88 16:29:20-EST
From: William L. Scherlis <SCHERLIS@vax.darpa.mil>
Subject: PI Meeting
To: SW-PI@vax.darpa.mil
Message-Id: <594941360.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
There is some confusion: Unless you have had a discussion
with Mark Pullen, you have no obligation to prepare a
presentation for the PI Meeting. (Consider the alternative:
350 presentations of 7 minutes each....)
I'll be sending a message soon with a proposal for an
agenda for the software session Thursday afternoon.
If you are going to respond to the earlier message,
please do so by tomorrow (Tuesday) evening, so we
can act on the responses. Thanks,
Bill
-------
∂07-Nov-88 1429 honavar@cs.wisc.edu Re: AI as CS and the scientific epistemology of the common sense world
Received: from goat.cs.wisc.edu by SAIL.Stanford.EDU with TCP; 7 Nov 88 14:29:11 PST
Date: Mon, 7 Nov 88 16:27:59 CST
From: honavar@cs.wisc.edu (A Buggy AI Program)
Message-Id: <8811072227.AA02652@goat.cs.wisc.edu>
Received: by goat.cs.wisc.edu; Mon, 7 Nov 88 16:27:59 CST
To: JMC@SAIL.STANFORD.EDU
Subject: Re: AI as CS and the scientific epistemology of the common sense world
Newsgroups: comp.ai.digest
In-Reply-To: <4pcvX@SAIL.Stanford.EDU>
Organization: U of Wisconsin CS Dept
Cc:
Dear Prof. McCarthy:
Could you please mail me a reprint of your paper titled
``Ascribing Mental Qualities to Machines'' that was
cited in your recent ai-digest article? Thanks in advance.
Vasant Honavar
Computer Sciences Department
University of Wisconsin-Madison
1210 W. Dayton St.
Madison, WI 53706.
honavar@ai.cs.wisc.edu
∂08-Nov-88 1154 VAL Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
INHERITANCE HIERARCHIES AND AUTOEPISTEMIC LOGIC
Michael Gelfond
University of Texas at El Paso
Friday, November 11, 3:15pm
MJH 301
We propose a new semantics for inheritance hierarchies with
exceptions. Our approach is based on a translation of an inheritance
hierarchy into a logical theory stated in autoepistemic logic. This
translation interprets normative statements of the form "Typically
property F(x) holds" as statements about the beliefs of an agent whose
premises are determined by the inheritance hierarchy. We hope that
this interpretation will provide some insights into the nature of
inheritance-based reasoning.
This is joint work with Halina Przymusinska.
∂08-Nov-88 1225 jwalton@vax.darpa.mil [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: Deferred: Connecti]
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 8 Nov 88 12:25:12 PST
Resent-Message-Id: <8811081813.AA13808@vax.darpa.mil>
Received: by vax.darpa.mil (5.59/5.51)
id AA13808; Tue, 8 Nov 88 13:13:10 EST
Received-Date: Tue, 8 Nov 88 12:09:54 EST
Received: from sun30.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA13525; Tue, 8 Nov 88 12:09:54 EST
Posted-Date: Tue 8 Nov 88 12:10:34-EST
Received: by sun30.darpa.mil (5.54/5.51)
id AA04681; Tue, 8 Nov 88 12:10:35 EST
Date: Tue 8 Nov 88 12:10:34-EST
From: Mark Pullen <PULLEN@vax.darpa.mil>
Subject: [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: Deferred: Connecti]
To: jwalton@vax.darpa.mil
Message-Id: <595012234.0.PULLEN@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Resent-Date: Tue 8 Nov 88 13:13:08-EST
Resent-From: Juanita Walton <JWALTON@VAX.DARPA.MIL>
Resent-To: ISTO-PI-LIST@vax.darpa.mil
Juanita,
Please retry on the messages that didn't get thru.
Mark
---------------
Date: Mon, 7 Nov 88 17:25:56 EST
From: MAILER-DAEMON (Mail Delivery Subsystem)
Subject: Returned mail: Deferred: Connection timed out during user open with vlsi.caltech.edu
Posted-Date: Mon, 7 Nov 88 17:25:56 EST
Received-Date: Mon, 7 Nov 88 17:25:56 EST
Message-Id: <8811072225.AA11176@vax.darpa.mil>
Received: from sun30.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA11176; Mon, 7 Nov 88 17:25:56 EST
To: <pullen>
----- Transcript of session follows -----
421 note.nsf.gov.tcp... Deferred: Connection timed out during user open with note.nsf.gov
550 /darpa/isto/list/arch-pi.list: line 132: hlum%nasamail@ames.arc.nasa.gov(Henry Lum-NASA)... Host unknown
421 nswc-wo.arpa.tcp... Deferred: Connection refused by NSWC-WO.ARPA
>>> RCPT To:<mtl.lab@WSMR-SIMTEL20.ARMY.MIL>
<<< 550 No such local mailbox as "mtl.lab", recipient rejected
550 /darpa/isto/list/arch-pi.list: line 180: mtl.lab@simtel20.arpa(Dan Koff-AMTL)... User unknown
421 brl.arpa.tcp... Deferred: Connection timed out during user open with BRL.ARPA
>>> QUIT
<<< 421 research.att.com too busy, please try later
>>> QUIT
<<< 421 research.att.com too busy, please try later
451 /darpa/isto/list/arch-pi.list: line 285: putbody: write error: Network is unreachable
421 vlsi.caltech.edu.tcp... Deferred: Connection timed out during user open with vlsi.caltech.edu
----- Unsent message follows -----
Received-Date: Mon, 7 Nov 88 17:25:56 EST
Received: from sun30.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA11154; Mon, 7 Nov 88 17:25:56 EST
Posted-Date: Mon 7 Nov 88 17:26:41-EST
Received: by sun30.darpa.mil (5.54/5.51)
id AA03935; Mon, 7 Nov 88 17:26:42 EST
Date: Mon 7 Nov 88 17:26:41-EST
From: Mark Pullen <PULLEN@vax.darpa.mil>
Subject: PI Meeting Agenda
To: ISTO-PI-LIST
Cc: ISTO-AGENTS, pimeet
Message-Id: <594944801.0.PULLEN@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Dear ISTO PI,
This message provides an update on the schedule for the upcoming
DARPA/ISTO PIU Meeting in Dallas, Texas, 14-18 November 1988.
With the help of the Program Committee, we were able to include
many of the good "New Ideas" white papers in the afternoon sessions.
These sessions will be shaped by their coordinators to include
short presentation, questions, and lots of discussion. If your
submission is listed below, please contact the session coordinator.
The planned agenda is as follows:
Monday 14 November
Registration until 1400
1400 Jack Schwartz: Keynote
1500 Chuck Seitz: Parallel Computing Overview
1600 Raj Reddy: Artificial Intelligence Overview
1700 Alfred Spector,Dave Clark,Mike Frankel:
Distributed Systems/Networking/C3 Overview
1830 Reception
Tuesday 15 November
0830 Steve Jacobsen: Robotics Overview
0915 Marty Tenenbaum: Concurrent Product and Process Design
1000 Break
1030 Salim Roukos: Toward Spoken Language Systems
1100 Bob Wilensky: The UNIX Consultant
1130 Dick Shively: Aspen Multicomuter
1200 Rick Rashid: Mach project Status
1230 Lunch, talk by Bob Kahn: Toward a National Network
1400 Parallel Sessions on New Ideas
New Technology in VLSI (Takeo Kanade coord)
pres by Kanade,Plummer/Horowitz/Losleben,Koch/Olin
Computation in Design (Saul Amarel coord)
pres by Amarel,Chandrasekran,Walker/Blumenthal,Losleben
Knowledge-Based Management (Ullman coord)
pres by Ullman/Weiderhold,Grishman,Gurfield/Katz/Postel
Application-Driven Architectures (Bert Halstead coord)
pres by Halstead,Weems,Albus
Connectionist Systems (David Waltz coord)
Free-Form Fabrication: Stereolithography and
Selected Laser Sintering (Isler coord)
1900 Dinner (Texas barbeque)
Wednesday 16 November
0830 Takeo Kanade: Computer Vision
0900 Chip Weems: Image Understanding Benchmarks
0930 H.T. Kung: Warp,Iwarp and Nectar
1000 Duncan Miller: SIMNET
1030 Break
1100 Arvind: Dataflow and Implicit Parallelism
1130 Harry Forsdick: Collaboration Technology
1200 Bob Balzer: Comprehensive Prototyping System
1230 Lunch, talk by John McCarthy: Lisp, Mathematic Logic, and AI
1400 Parallel Sessions on New Ideas
Architecture Compilation (Phil Kuekes coord)
pres by Kuekes/Shen,Despain,Lipton,Shen
Battle Management/C3I (Mike Frankel coord)
pres by Frankel,Hayes-Roth,Birman/Marzullo,Hobbs
Parallel/Reliable/Distributed Systems (Dan Siewiorek coord)
pres by Siewiorek/Segall/Strosnider,Cheriton,Spector
CPS(Balzer)/Software CAD Databases (Larry Rowe coord)
pres by Rowe,Dewitt,Osterweil,Zdonik
DICE Meets First Cut and other related ISTO Projects
(Isler coord)
Intelligent Training Systems (Keith Uncapher coord)
1900 Dinner, talk by Bob Cooper: AI Retrospective
Thursday 17 November
0830 Rich Sincovec: Center for Advanced Architectures
0900 Jon Postel: Regional Nets Replace ARPANET
0930 Mike Stonebraker: Database Management
1000 Danny Cohen: Computerized Commerce
1030 Break
1100 Bob Broderson: Microsystem Design
1130 Bill Wedlake: Airland Battle Management
1200 Paul Rosenbloom: Machine Learning
1230 Lunch, talk by Craig Fields: National Defense and
Computer/Infomation Science
1400 Program Area Meetings
Boesch/Sowa/Thorpe/Shiflett: Distributed Systems and C3
Pullen/Richer: Computing/Networking Infrastructure
Rosenfeld: Robotics
Scherlis: Software and Algorithms
Simpson/Wayne/Frew: Artificial Intelligence/AI Applications
Squires: Parallel Architectures and Teraops
Toole: Microsystems Design and CAD/CAM
1930 Dinner, for DARPA/ex-DARPA personnel only
(the rest of you are on your own!)
Friday 18 November
0830 Panel: (Bob Simpson,Lou Kallis,Mark Pullen,Ron Register,
Stephen Squires,Jack Verkoski):Program Management
1000 Break
1030 Michael Dertouzos: Computers for Productivity
1100 Panel(DARPA types): New Program Plans
1230 Adjourn
As you can see, this should be a very interesting meeting!
See you there.
Mark Pullen
-------
-------
-------
-------
∂08-Nov-88 1507 pullen@vax.darpa.mil ISTO PI Meeting Agenda
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 8 Nov 88 15:07:38 PST
Received: from sun30.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA14707; Tue, 8 Nov 88 16:42:47 EST
Posted-Date: Tue 8 Nov 88 16:43:25-EST
Received: by sun30.darpa.mil (5.54/5.51)
id AA04981; Tue, 8 Nov 88 16:43:27 EST
Date: Tue 8 Nov 88 16:43:25-EST
From: Mark Pullen <PULLEN@vax.darpa.mil>
Subject: ISTO PI Meeting Agenda
To: ISTO-PI-LIST@vax.darpa.mil
Cc: ISTO-AGENTS@vax.darpa.mil, pimeet@vax.darpa.mil, PGM-CMTE@vax.darpa.mil
Message-Id: <595028605.0.PULLEN@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
******* CORRECTED RETRANSMISSION ********************************
Dear ISTO PI,
This message provides an update on the schedule for the upcoming
DARPA/ISTO PI Meeting in Dallas, Texas, 14-18 November 1988.
With the help of the Program Committee, we were able to include
many of the good "New Ideas" white papers in the afternoon sessions.
These sessions will be shaped by their coordinators to include
short presentation, questions, and lots of discussion. If your
submission is listed below, please contact the session coordinator.
The planned agenda is as follows:
Monday 14 November
Registration until 1400
1400 Jack Schwartz: Keynote
1500 Chuck Seitz: Parallel Computing Overview
1600 Raj Reddy: Artificial Intelligence Overview
1700 Alfred Spector,Dave Clark,Mike Frankel:
Distributed Systems/Networking/C3 Overview
1830 Reception
Tuesday 15 November
0830 Steve Jacobsen: Robotics Overview
0915 Marty Tenenbaum: Concurrent Product and Process Design
1000 Break
1030 Salim Roukos: Toward Spoken Language Systems
1100 Bob Wilensky: The UNIX Consultant
1130 Dick Shively: Aspen Multicomuter
1200 Rick Rashid: Mach project Status
1230 Lunch, talk by Bob Kahn: Toward a National Network
1400 Parallel Sessions on New Ideas
Parallel/Reliable/Distributed Systems(Dan Siewiorek coord)
pres by Siewiorek/Segall/Strosnider,Cheriton,Spector
Computation in Design (Saul Amarel coord)
pres by Amarel,Chandrasekran,Walker/Blumenthal,Losleben
Knowledge-Based Management (Ullman coord)
pres by Ullman/Weiderhold,Grishman,Gurfield/Katz/Postel
Application-Driven Architectures (Bert Halstead coord)
pres by Halstead,Weems,Albus
Intelligent Training Systems (Elliot Soloway coord)
Free-Form Fabrication: Stereolithography and
Selected Laser Sintering (Isler coord)
1900 Dinner (Texas barbeque)
Wednesday 16 November
0830 Takeo Kanade: Computer Vision
0900 Chip Weems: Image Understanding Benchmarks
0930 H.T. Kung: Warp,Iwarp and Nectar
1000 Duncan Miller: SIMNET
1030 Break
1100 Arvind: Dataflow and Implicit Parallelism
1130 Harry Forsdick: Collaboration Technology
1200 Bob Balzer: Comprehensive Prototyping System
1230 Lunch, talk by John McCarthy: Lisp, Mathematic Logic, and AI
1400 Parallel Sessions on New Ideas
Architecture Compilation (Phil Kuekes coord)
pres by Kuekes/Shen,Despain,Lipton,Shen
Battle Management/C3I (Mike Frankel coord)
pres by Frankel,Hayes-Roth,Birman/Marzullo,Hobbs
New Technology in VLSI(Takeo Kanade coord)
pres by Kanade,Plummer/Horowitz/Losleben,Koch/Olin
CPS(Balzer)/Software CAD Databases (Larry Rowe coord)
pres by Rowe,Dewitt,Osterweil,Zdonik
DICE Meets First Cut and other related ISTO Projects
(Isler coord)
Connectionist Systems (David Waltz coord)
1900 Dinner, talk by Bob Cooper: AI Retrospective
Thursday 17 November
0830 Rich Sincovec: Center for Advanced Architectures
0900 Jon Postel: Regional Nets Replace ARPANET
0930 Mike Stonebraker: Database Management
1000 Danny Cohen: Computerized Commerce
1030 Break
1100 Bob Broderson: Microsystem Design
1130 Bill Wedlake: Airland Battle Management
1200 Paul Rosenbloom: Machine Learning
1230 Lunch, talk by Craig Fields: National Defense and
Computer/Infomation Science
1400 Program Area Meetings
Boesch/Sowa/Thorpe/Shiflett: Distributed Systems and C3
Pullen/Richer: Computing/Networking Infrastructure
Rosenfeld: Robotics
Scherlis: Software and Algorithms
Simpson/Wayne/Frew: Artificial Intelligence/AI Applications
Squires: Parallel Architectures and Teraops
Toole: Microsystems Design and CAD/CAM
1930 Dinner, for DARPA/ex-DARPA personnel only
(the rest of you are on your own!)
Friday 18 November
0830 Panel: (Bob Simpson,Lou Kallis,Mark Pullen,Ron Register,
Stephen Squires,Jack Verkoski):Program Management
1000 Break
1030 Michael Dertouzos: Computers for Productivity
1100 Panel(DARPA types): New Program Plans
1230 Adjourn
As you can see, this should be a very interesting meeting!
See you there.
Mark Pullen
-------
∂08-Nov-88 1938 qphysics-owner@neat.ai.toronto.edu lethargy?
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 8 Nov 88 19:38:43 PST
Received: from [128.100.1.65] by RELAY.CS.NET id aa09188; 8 Nov 88 3:39 EST
Received: by neat.ai.toronto.edu id 8197; Tue, 8 Nov 88 03:04:05 EST
Resent-From: qphysics-owner@ai.toronto.edu
Resent-To: qphysics-distribution@ai.toronto.edu
Resent-Sender: qphysics-owner@ai.toronto.edu
Received: from RELAY.CS.NET (relay1.cs.net) by neat.ai.toronto.edu with SMTP id 8473; Tue, 8 Nov 88 03:03:49 EST
Received: from nooksack.cs.washington.edu by RELAY.CS.NET id aa07623;
8 Nov 88 1:44 EST
Received: by nooksack.cs.washington.edu (5.52.1/6.13)
id AA02601; Mon, 7 Nov 88 11:04:06 PST
Date: Mon, 7 Nov 88 14:04:06 EST
From: Dan Weld <weld@nooksack.cs.washington.edu>
Return-Path: <weld@ai.toronto.edu>
Message-Id: <8811071904.AA02601@nooksack.cs.washington.edu>
To: qphysics@ai.toronto.edu
Subject: lethargy?
Resent-Message-Id: <88Nov8.030405est.8197@neat.ai.toronto.edu>
Resent-Date: Tue, 8 Nov 88 03:03:53 EST
I thought long and hard trying to come up with an inspirational (or at least
inflamitory) message for this group, but dozed off with dreams of a
fricitionless spring moving to the right, slowing down, now back to the left...
Since I can't be truly exciting, at least I'll announce the availability of my
thesis (MIT AI tech report 1035). Unfortunately, I gave all my copies away so
you have to get one from PUBLICATIONS@WHEATIES.AI.MIT.EDU Abstract follows.
THEORIES OF COMPARATIVE ANALYSIS
Dan Weld
Comparative analysis is the problem of predicting how a system will react to
perturbations in its parameters, and why. For example, comparative analysis
could be asked to explain why the period of an oscillating spring/block system
would increase if the mass of the block were larger. This thesis formalizes
the task of comparative analysis and presents two solution techniques:
differential qualitative (DQ) analysis and exaggeration. Both techniques
solve many comparative analysis problems, providing explanations suitable for
use by design systems, automated diagnosis, intelligent tutoring systems, and
explanation based generalization.
This thesis explains the theoretical basis for each technique, describes how
they are implemented, and discusses the difference between the two.
DQ analysis is sound; it never generates an incorrect answer to a comparative
analysis question. Although exaggeration does occasionally produce misleading
answers, it solves a larger class of problems than DQ analysis and frequently
results in simpler explanations.
∂08-Nov-88 2106 qphysics-owner@neat.ai.toronto.edu Lethargy?
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 8 Nov 88 21:06:09 PST
Received: from [128.100.1.65] by RELAY.CS.NET id ai15133; 8 Nov 88 11:38 EST
Received: by neat.ai.toronto.edu id 8652; Tue, 8 Nov 88 11:09:37 EST
Resent-From: qphysics-owner@ai.toronto.edu
Resent-To: qphysics-distribution@ai.toronto.edu
Resent-Sender: qphysics-owner@ai.toronto.edu
Received: from RELAY.CS.NET (relay1.cs.net) by neat.ai.toronto.edu with SMTP id 8677; Tue, 8 Nov 88 11:09:19 EST
Received: from a.cs.uiuc.edu by RELAY.CS.NET id aa12392; 8 Nov 88 7:16 EST
Received: from p.cs.uiuc.edu by a.cs.uiuc.edu with SMTP (UIUC-5.52/9.7)
id AA22123; Mon, 7 Nov 88 16:17:03 CST
Received: by p.cs.uiuc.edu (UIUC-5.52/9.7)
id AA27283; Mon, 7 Nov 88 16:17:12 CST
Date: Mon, 7 Nov 88 17:17:12 EST
From: Kenneth Forbus <forbus@P.CS.UIUC.EDU>
Message-Id: <8811072217.AA27283@p.cs.uiuc.edu>
To: qphysics@ai.toronto.edu
Subject: Lethargy?
Cc: forbus@P.CS.UIUC.EDU
Resent-Message-Id: <88Nov8.110937est.8652@neat.ai.toronto.edu>
Resent-Date: Tue, 8 Nov 88 11:09:34 EST
Given the close proximity of the IJCAI deadline, I'm not too surprised
people have their blinders on...
∂09-Nov-88 1204 VAL Special Seminar: Keith Clark's FGCS invited talk
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
LOGIC PROGRAMMING SCHEMES
Keith L. Clark
Imperial College of Science and Technology
London, England
Monday, November 14, 4:15pm
MJH252
ABSTRACT. The essential logical and operation properties of a
clause based logic programming language are discussed by
presenting a very general framework that covers Kowalski's
original SLD scheme and the more recent constraint programming
schemes.
(This is an invited talk to be given at the International
Conference on Fifth Generation Computer Systems in Japan.)
∂09-Nov-88 1305 GOODMAN%uamis@rvax.ccit.arizona.edu finished
Received: from rvax.ccit.arizona.edu by SAIL.Stanford.EDU with TCP; 9 Nov 88 13:04:46 PST
Date: Wed, 9 Nov 88 12:31 MST
From: GOODMAN%uamis@rvax.ccit.arizona.edu
Subject: finished
To: DUANE.ADAMS@c.cs.cmu.edu, BLUMENTHAL@a.isi.edu, 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, KNEMEYER@a.isi.edu,
MCHENRY@GUVAX.BITNET, OUSTER@ginger.berkeley.edu, Ralston@mcc.com,
CWEISSMAN@dockmaster.arpa
X-VMS-To: @NAS, THORNTON, BILLM
Marjory and I just got off the phone after a few final frantic days of
editing etc. We are putting it forward to the printer!
Once again, I'd like to thank all of you for your efforts. Virtually
everyone in the committee (17 members) ended up doing something
substantial, giving us the breadth of participation we wanted and
needed. Given the scope of our assignment, and the severity of our
resource constraints, we have every reason to feel good about the
end result.
We've been a good group. Thanks again.
∂09-Nov-88 1333 jwalton@vax.darpa.mil [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: Deferred: Connecti]
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 9 Nov 88 13:32:55 PST
Resent-Message-Id: <8811081813.AA13808@vax.darpa.mil>
Received: by vax.darpa.mil (5.59/5.51)
id AA13808; Tue, 8 Nov 88 13:13:10 EST
Received-Date: Tue, 8 Nov 88 12:09:54 EST
Received: from sun30.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA13525; Tue, 8 Nov 88 12:09:54 EST
Posted-Date: Tue 8 Nov 88 12:10:34-EST
Received: by sun30.darpa.mil (5.54/5.51)
id AA04681; Tue, 8 Nov 88 12:10:35 EST
Date: Tue 8 Nov 88 12:10:34-EST
From: Mark Pullen <PULLEN@vax.darpa.mil>
Subject: [MAILER-DAEMON (Mail Delivery Subsystem): Returned mail: Deferred: Connecti]
To: jwalton@vax.darpa.mil
Message-Id: <595012234.0.PULLEN@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Resent-Date: Tue 8 Nov 88 13:13:08-EST
Resent-From: Juanita Walton <JWALTON@vax.darpa.mil>
Resent-To: ISTO-PI-LIST@vax.darpa.mil
Juanita,
Please retry on the messages that didn't get thru.
Mark
---------------
Date: Mon, 7 Nov 88 17:25:56 EST
From: MAILER-DAEMON (Mail Delivery Subsystem)
Subject: Returned mail: Deferred: Connection timed out during user open with vlsi.caltech.edu
Posted-Date: Mon, 7 Nov 88 17:25:56 EST
Received-Date: Mon, 7 Nov 88 17:25:56 EST
Message-Id: <8811072225.AA11176@vax.darpa.mil>
Received: from sun30.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA11176; Mon, 7 Nov 88 17:25:56 EST
To: <pullen>
----- Transcript of session follows -----
421 note.nsf.gov.tcp... Deferred: Connection timed out during user open with note.nsf.gov
550 /darpa/isto/list/arch-pi.list: line 132: hlum%nasamail@ames.arc.nasa.gov(Henry Lum-NASA)... Host unknown
421 nswc-wo.arpa.tcp... Deferred: Connection refused by NSWC-WO.ARPA
>>> RCPT To:<mtl.lab@WSMR-SIMTEL20.ARMY.MIL>
<<< 550 No such local mailbox as "mtl.lab", recipient rejected
550 /darpa/isto/list/arch-pi.list: line 180: mtl.lab@simtel20.arpa(Dan Koff-AMTL)... User unknown
421 brl.arpa.tcp... Deferred: Connection timed out during user open with BRL.ARPA
>>> QUIT
<<< 421 research.att.com too busy, please try later
>>> QUIT
<<< 421 research.att.com too busy, please try later
451 /darpa/isto/list/arch-pi.list: line 285: putbody: write error: Network is unreachable
421 vlsi.caltech.edu.tcp... Deferred: Connection timed out during user open with vlsi.caltech.edu
----- Unsent message follows -----
Received-Date: Mon, 7 Nov 88 17:25:56 EST
Received: from sun30.darpa.mil by vax.darpa.mil (5.59/5.51)
id AA11154; Mon, 7 Nov 88 17:25:56 EST
Posted-Date: Mon 7 Nov 88 17:26:41-EST
Received: by sun30.darpa.mil (5.54/5.51)
id AA03935; Mon, 7 Nov 88 17:26:42 EST
Date: Mon 7 Nov 88 17:26:41-EST
From: Mark Pullen <PULLEN@vax.darpa.mil>
Subject: PI Meeting Agenda
To: ISTO-PI-LIST
Cc: ISTO-AGENTS, pimeet
Message-Id: <594944801.0.PULLEN@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Dear ISTO PI,
This message provides an update on the schedule for the upcoming
DARPA/ISTO PIU Meeting in Dallas, Texas, 14-18 November 1988.
With the help of the Program Committee, we were able to include
many of the good "New Ideas" white papers in the afternoon sessions.
These sessions will be shaped by their coordinators to include
short presentation, questions, and lots of discussion. If your
submission is listed below, please contact the session coordinator.
The planned agenda is as follows:
Monday 14 November
Registration until 1400
1400 Jack Schwartz: Keynote
1500 Chuck Seitz: Parallel Computing Overview
1600 Raj Reddy: Artificial Intelligence Overview
1700 Alfred Spector,Dave Clark,Mike Frankel:
Distributed Systems/Networking/C3 Overview
1830 Reception
Tuesday 15 November
0830 Steve Jacobsen: Robotics Overview
0915 Marty Tenenbaum: Concurrent Product and Process Design
1000 Break
1030 Salim Roukos: Toward Spoken Language Systems
1100 Bob Wilensky: The UNIX Consultant
1130 Dick Shively: Aspen Multicomuter
1200 Rick Rashid: Mach project Status
1230 Lunch, talk by Bob Kahn: Toward a National Network
1400 Parallel Sessions on New Ideas
New Technology in VLSI (Takeo Kanade coord)
pres by Kanade,Plummer/Horowitz/Losleben,Koch/Olin
Computation in Design (Saul Amarel coord)
pres by Amarel,Chandrasekran,Walker/Blumenthal,Losleben
Knowledge-Based Management (Ullman coord)
pres by Ullman/Weiderhold,Grishman,Gurfield/Katz/Postel
Application-Driven Architectures (Bert Halstead coord)
pres by Halstead,Weems,Albus
Connectionist Systems (David Waltz coord)
Free-Form Fabrication: Stereolithography and
Selected Laser Sintering (Isler coord)
1900 Dinner (Texas barbeque)
Wednesday 16 November
0830 Takeo Kanade: Computer Vision
0900 Chip Weems: Image Understanding Benchmarks
0930 H.T. Kung: Warp,Iwarp and Nectar
1000 Duncan Miller: SIMNET
1030 Break
1100 Arvind: Dataflow and Implicit Parallelism
1130 Harry Forsdick: Collaboration Technology
1200 Bob Balzer: Comprehensive Prototyping System
1230 Lunch, talk by John McCarthy: Lisp, Mathematic Logic, and AI
1400 Parallel Sessions on New Ideas
Architecture Compilation (Phil Kuekes coord)
pres by Kuekes/Shen,Despain,Lipton,Shen
Battle Management/C3I (Mike Frankel coord)
pres by Frankel,Hayes-Roth,Birman/Marzullo,Hobbs
Parallel/Reliable/Distributed Systems (Dan Siewiorek coord)
pres by Siewiorek/Segall/Strosnider,Cheriton,Spector
CPS(Balzer)/Software CAD Databases (Larry Rowe coord)
pres by Rowe,Dewitt,Osterweil,Zdonik
DICE Meets First Cut and other related ISTO Projects
(Isler coord)
Intelligent Training Systems (Keith Uncapher coord)
1900 Dinner, talk by Bob Cooper: AI Retrospective
Thursday 17 November
0830 Rich Sincovec: Center for Advanced Architectures
0900 Jon Postel: Regional Nets Replace ARPANET
0930 Mike Stonebraker: Database Management
1000 Danny Cohen: Computerized Commerce
1030 Break
1100 Bob Broderson: Microsystem Design
1130 Bill Wedlake: Airland Battle Management
1200 Paul Rosenbloom: Machine Learning
1230 Lunch, talk by Craig Fields: National Defense and
Computer/Infomation Science
1400 Program Area Meetings
Boesch/Sowa/Thorpe/Shiflett: Distributed Systems and C3
Pullen/Richer: Computing/Networking Infrastructure
Rosenfeld: Robotics
Scherlis: Software and Algorithms
Simpson/Wayne/Frew: Artificial Intelligence/AI Applications
Squires: Parallel Architectures and Teraops
Toole: Microsystems Design and CAD/CAM
1930 Dinner, for DARPA/ex-DARPA personnel only
(the rest of you are on your own!)
Friday 18 November
0830 Panel: (Bob Simpson,Lou Kallis,Mark Pullen,Ron Register,
Stephen Squires,Jack Verkoski):Program Management
1000 Break
1030 Michael Dertouzos: Computers for Productivity
1100 Panel(DARPA types): New Program Plans
1230 Adjourn
As you can see, this should be a very interesting meeting!
See you there.
Mark Pullen
-------
-------
-------
-------
∂09-Nov-88 1345 postmaster@guvax Returned Network Mail
Received: from guvax (GUVAX.GEORGETOWN.EDU) by SAIL.Stanford.EDU with TCP; 9 Nov 88 13:44:53 PST
Date: 2 Nov 88 01:34:00 EST
From: "RSCS Postmaster" <postmaster@guvax>
Subject: Returned Network Mail
To: "jmc" <jmc@sail.stanford.edu>
Your mail is being returned to you.
Reason for return is:
%MAIL-E-OPENOUT, error opening DISK$ACC1:[MCHENRY.MAIL]MAIL.MAI; as output
-SYSTEM-F-IVDEVNAM, invalid device name
Returned mail follows:
------------------------------
Received: From CUNYVM(MAILER) by GUVAX with Jnet id 3071
for MCHENRY@GUVAX; Wed, 2 Nov 88 01:34 EST
Received: from CUNYVM by CUNYVM.BITNET (Mailer X2.00) with BSMTP id 7242; Tue,
01 Nov 88 23:09:17 EST
Received: from SAIL.Stanford.EDU by CUNYVM.CUNY.EDU (IBM VM SMTP R1.1) with TCP;
Tue, 01 Nov 88 23:08:55 EST
Message-ID: <$qtO#@SAIL.Stanford.EDU>
Date: 01 Nov 88 1629 PST
From: John McCarthy <JMC@SAIL.Stanford.EDU>
Subject: re: Final Touches
To: BLUMENTHAL@a.ISI.EDU, duane.adams@CS.CMU.EDU, dongarra@MCS.ANL.GOV,
gannon%rdvax.dec@decwrl.dec.com, gossard@CADLAB2.MIT.EDU,
hearn@RAND-UNIX.ARPA, jlh@VSOP.STANFORD.EDU,
mchenry%guvax.bitnet@CUNYVM.CUNY.EDU, ouster@GINGER.BERKELEY.EDU,
ralston@MCC.COM, thornj@max.acs.washington.edu, CWeissman@DOCKMASTER.ARPA,
troywil@IBM.COM
[In reply to message from BLUMENTHAL@A.ISI.EDU sent Tue 1 Nov 88 17:20:47-EST.]
I have been thinking about Soviet access to Western computer
technology on lines somewhat orthogonal to the committee's deliberations.
I'm sorry I didn't write it up sooner, because I think it's too late
to incorporate its ideas in the report even if the committee considered
it appropriate. I would like to refer to the committee report when it
is released. Here are my ideas, and I would welcome comments including
comments about where it might be appropriate to direct the essay.
%soviet[f88,jmc] Terms for Soviet access to Western computer technology
\input memo.tex[let,jmc]
\title{TERMS FOR SOVIET ACCESS TO WESTERN COMPUTER TECHNOLOGY}
The object of this essay is to suggest that the United
States and its allies undertake to set terms for greatly
increased Soviet access to Western computer and other electronic
technology. We will also have to sell the Soviets on changing
their ways of importing technology and on the advantages of paying
the price we ask.
Here are some considerations.
1. The Soviets are far behind in these areas. They have been
behind ever since the computer industry started and are not catching up.
2. The Western countries through the COCOM consortium restrict
technology exports to the Soviet Union for defense reasons. In the
computer area the newest technologies are restricted, but the restrictions
on any particular technology are removed after some years, averaging
five years.
3. The Soviets import some Western computer technology in
compliance with the restrictions.
4. The Soviets steal other technology, chiefly through getting
unscrupulous Western businessmen to set up dummy companies, purchase
the computers and smuggle them. Every so often Western countries catch
someone at it and arrest them. Only small numbers of computers are
illegally imported. Probably they are mainly imported to be copied
rather than just for use. The Soviets copied the IBM 360/370 line starting
in the middle 1960s with only moderate success. In the middle 1970s
they started making computers compatible with the D.E.C PDP-11 and
more recently the VAX. This hasn't been very successful either.
5. Theft as a means of getting technology has serious disadvantages
for the Soviets. Here are some.
a. The documentation obtained often doesn't agree with the
hardware.
b. The normal use of computer technology involves continued
communication between the users of the hardware or software and the
suppliers. This communication involves correcting users' mistakes, resolving
ambiguities and incompleteness in the documentation, getting bugs that
have arisen in the users' work corrected, 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 initial compatibility with foreign
technology often lost when the Soviets have to improvise a solution to
a problem that arises. Their solution is likely to be incompatibile with
changes made by the supplier.
c. The KGB or whoever steals the technology insists that
the technology be kept under wraps, and this interferes with communication
within the Soviet Union.
d. Institutions with stolen technology are restricted
in their communication with foreigners.
6. Very likely the KGB doesn't understand the difficulties
their methods make for their Soviet customers. Most likely they are
proud of their intelligence coups. Our intelligence people are chagrined
at the KGB's successes but may not be in a position to analyze how much
use Soviet industry gets from it.
7. The COCOM restrictions have important effects in
limiting Soviet computer technology. However, they are
probably less than the effects of the Soviets' own restrictions on
their ability to absorb foreign technology. Here are some details.
a. They restrict foreign travel by their own scientists
and engineers far beyond the restrictions imposed by their lack
of foreign currency. While they get all the important foreign
scientific journals in their central libraries,
distribution throughout the country is weak, and there are very
few individual subscriptions. This makes use of foreign ideas
difficult, and encourages complacency about how well they are
doing.
b. When they do buy foreign computers legally, they usually
restrict their contacts with the service organizations of the
companies from which they purchase. For example, they don't
let them set up service organizations within the Soviet Union.
This makes service calls very difficult.
8. The Gorbachev reforms are making the Soviet Union
more congenial to many people all over the world. Lots of
people didn't see the defense importance of observing COCOM
restrictions in the past, and this number will increase. The
number of suppliers has increased. The COCOM system may weaken
considerably.
9. It would increase the Soviet standard of living
considerably over the long term to induce the West to
relax or abandon restrictions on technology transfer. It
would be to their advantage to pay a considerable price
for this relaxation. Running a technology race with the
rest of the world is something they can't win.
10. The Soviet standard of living hasn't been the
dominant consideration with the Politburo in the past, and it
isn't obvious today what its priority is relative to military
advantage.
11. Some Western people favor relaxing the restrictions
unilaterally to encourage Gorbachev, to promote peace, because
they consider them wasted effort or for other reasons. They have
had some success from time to time, but there is no reason to
suppose they will get 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.
12. To the extent that the West is agreeable, the best Soviet
strategy is to rejoin the world technologically. This means
buying Western products and technology from a variety of countries
using the same commercial practices as are used among Western
countries. It means letting Western companies 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 Trade. It means letting
Soviet engineers and scientists subscribe freely to foreign
publications and travel abroad freely when it advances their work.
13. The advantages of rejoining the world technologically
will not be obvious to organizations like the KGB, proud of their
success in stealing technology. It also goes against the
tendency of the Party to control everything. However, the
Gorbachev Administration has been taking some steps in this
direction.
14. The West needs to figure out how to sell the Soviets
on the advantages of rejoining the world. Otherwise, the negotiations
will fail, because the Soviet diplomats won't find the price
worth paying.
15. It is beyond the scope of this paper to treat comprehensively
the price we should ask. Here are a few considerations.
a. The Soviets might like agreements purely in the
technological area - we exchange our technology for theirs. However,
we have so much more that they need than vice-versa that purely
technological exchanges won't go very far.
b. We should imagine a sequence of successively more
comprehensive agreements.
c. 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 anyway, and
we are getting some concessions for nothing. For example, they now
have promised the Soviet public to publish correct maps, including
a road atlas of the Soviet Union.
d. Probably the main concessions have to be in
the military area. They need to give up some of their conventional
armaments advantage in Europe, maybe even their draft.
16. 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.
17. According to Arkady Shevchenko, the Soviets have never
feared an unprovoked Western attack. This suggests that we make
no presumptions about their state of mind based on ideas of symmetry
between their situation and ours.
18. The single most effective way for them to improve their
standard of living is to reduce military and police expenditures.
Howver, this offers institutional difficulties, the extent of which
we cannot predict.
\noindent Summary.
1. The West should decide on terms for reducing technological
restrictions. Otherwise, they may just evaporate with no corresponding
gain in Western security and with reinforcement of the KGB doctrine
that the West consists of villains and suckers.
2. We need to sell the Soviets on the advantages to their
standard of living of rejoining the world technologically.
∂09-Nov-88 1743 qphysics-owner@neat.ai.toronto.edu another TR announcement
Received: from RELAY.CS.NET by SAIL.Stanford.EDU with TCP; 9 Nov 88 17:43:28 PST
Received: from [128.100.1.65] by RELAY.CS.NET id ac02664; 9 Nov 88 18:20 EST
Received: by neat.ai.toronto.edu id 11304; Wed, 9 Nov 88 16:29:20 EST
Resent-From: qphysics-owner@ai.toronto.edu
Resent-To: qphysics-distribution@ai.toronto.edu
Resent-Sender: qphysics-owner@ai.toronto.edu
Received: from RELAY.CS.NET ([10.4.0.5]) by neat.ai.toronto.edu with SMTP id 11267; Wed, 9 Nov 88 16:29:04 EST
Received: from wpafb-avlab.arpa by RELAY.CS.NET id aa13320; 9 Nov 88 10:05 EST
Received: from aruba.local (aruba.ARPA) by wpafb-avlab.arpa (1.2/Ultrix2.2-1)
id AA03799; Wed, 9 Nov 88 09:21:24 est
Received: from localhost.ARPA by aruba.local (1.2/Ultrix2.0-B)
id AA04057; Wed, 9 Nov 88 09:26:35 est
Message-Id: <8811091426.AA04057@aruba.local>
To: qphysics%ai.utoronto.ca@RELAY.CS.NET
Subject: another TR announcement
Date: Wed, 9 Nov 88 09:26:34 EST
From: wellman@ARUBA.BBN.COM
Resent-Message-Id: <88Nov9.162920est.11304@neat.ai.toronto.edu>
Resent-Date: Wed, 9 Nov 88 16:29:08 EST
Following Dan Weld's lead, I will use this list to advertise my own
dissertation tech report. (By the way, you should all look at Dan's
work if you haven't already.) It's number MIT/LCS/TR-427, available from:
Publications Office
MIT Laboratory for Computer Science
545 Technology Sq
Cambridge, MA 02139
-----------------------------
Title: Formulation of Tradeoffs in Planning Under Uncertainty
Author: Michael P. Wellman
Abstract:
The {\em tradeoff formulation} task is to identify the central
issues in a decision problem by recognizing strategies that are
qualitatively inadmissible. SUDO-Planner formulates tradeoffs for an
example medical decision problem by proving decision-theoretically
that certain classes of plans are dominated based on qualitative
relations in the domain.
The inadequacy of the traditional predicate representation of goals
for choice among plans that may achieve objectives in part or with
uncertainty motivates SUDO-Planner's {\em dominance-proving architecture}, a
general framework for planning for partially satisfiable goals.
Dominance-proving planners delimit the space of admissible plans by
maintaining a specialization graph of plan classes annotated with
dominance conditions derived from a domain model.
{\em Qualitative Probabilistic Networks} (QPNs) are decision models
expressing constraints on the joint probability distribution over a
set of variables. {\em Qualitative influences} describe the direction
of the relationship between two variables. {\em Qualitative
synergies} describe interactions among influences. The probabilistic
definitions of these constraints justify sound and efficient inference
procedures based on graphical manipulations of the network. SUDO-Planner's
dominance prover uses these procedures to establish dominance
relations among plan classes. SUDO-Planner constructs decision models
(QPNs) from a knowledge base describing the effects of actions and
relations among events at multiple levels of abstraction. The
planning process alternates between model construction and dominance
proving, producing a plan graph with dominance conditions ruling out
the inadmissible therapy strategies for its medical decision example.
------------------------------------
--Mike.
(wellman%aruba.dnet@wpafb-avlab.arpa)
∂10-Nov-88 0225 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Bessel, yo is my worry now
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 10 Nov 88 02:25:46 PST
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 333374; Thu 10-Nov-88 05:23:24 EST
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 75046; Thu 10-Nov-88 02:19:06 PST
Date: Thu, 10 Nov 88 02:16 PST
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Bessel, yo is my worry now
To: math-fun@RUSSIAN.SPA.Symbolics.COM, "ilan@score.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: rwg@RUSSIAN.SPA.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
Message-ID: <19881110101606.7.RWG@TSUNAMI.SPA.Symbolics.COM>
From: rwg
Date: a couple of days ago
Can anybody give >1 term of an expansion at x = 0 of
J[1+1/x](1/x)/J[1/x](1/x) ?
A physicist at BU had a continued fraction which he thought was
1 - k x↑(1/3) + . . ., but it came out to this Bessel ratio.
Turns out we were both right. With the help of the formulas on p 232
of Watson's (804 page) Treatise on the Bessel Function,
1 1 1/3
J (-) (- -)! (6 X)
1 1 + 1/X X 3
------------------------------------ = ----------- = 1 - --------------- - . . .
1 1 2
2 X + 2 - -------------------------- J (-) (- -)!
1 1/X X 3
4 X + 2 - ----------------
1
6 X + 2 - ------
. . .
(I was (stupidly) unprepared for the branchpoint at 0.)
This was the small X case of what he really wanted:
1
-------------------------------------------------------
X 1 1
C e - Y - - - ----------------------------------------
Y 2 X 1 1
C e - Y - - - -----------------------
Y 3 X 1 1
C e - Y - - - ------
Y . . .
(N + 1) X
N - ---------
==== /===\ 2
\ ! ! e
> ! ! --------------------------------
/ ! ! - K X - K X X
==== K = 1 (1 - e ) Y (e Y - C e )
N>=0
= ---------------------------------------------------- .
(N + 1) X
N - ---------
==== /===\ 2
X \ ! ! e
(C e - Y) > ! ! -----------------------------
/ ! ! - K X - K X
==== K = 1 (1 - e ) Y (e Y - C)
N>=0
Notice the nonobvious insensitivity of the rhs to reciprocating Y.
E.g., for Y = i, the lhs is neat, and the rhs looks imaginary.
∂10-Nov-88 1000 STAGER@Score.Stanford.EDU Text orders???
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Nov 88 10:00:25 PST
Date: Thu 10 Nov 88 10:00:17-PST
From: Claire Stager <STAGER@Score.Stanford.EDU>
Subject: Text orders???
To: guibas@Score.Stanford.EDU, shortliffe@SUMEX-AIM.Stanford.EDU,
genesereth@Score.Stanford.EDU, jmc@Sail.Stanford.EDU,
binford@Coyote.Stanford.EDU
Office: CS-TAC 29, 723-6094
Message-ID: <12445483061.17.STAGER@Score.Stanford.EDU>
This will be my last attempt to force a textbook order for Winter Qtr
out of you. Please let me know if you need any texts ordered asap.
Thanks again.
Claire
-------
∂10-Nov-88 1206 weening@Gang-of-Four.Stanford.EDU Network connection to USSR
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Nov 88 12:06:22 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02155; Thu, 10 Nov 88 12:05:49 PST
Date: Thu, 10 Nov 88 12:05:49 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811102005.AA02155@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Network connection to USSR
The following two messages recently appeared on USENET. Let me know
if you are interested in any followups.
∂10-Nov-88 1206 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Nov 88 12:06:37 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02160; Thu, 10 Nov 88 12:06:04 PST
Date: Thu, 10 Nov 88 12:06:04 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811102006.AA02160@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: crunch@well.UUCP (John Draper)
Newsgroups: news.admin
Subject: Where, oh Where does it go??
Message-ID: <7601@well.UUCP>
Date: 9 Nov 88 10:27:02 GMT
Reply-To: crunch@well.UUCP (John Draper)
Organization: Whole Earth Lectroinic Link, Sausalito, CA
Lines: 44
I just recently returned from the Soviet Union, and have documented
some VERY INTERESTING things going on over there. I think that the
UseNet sites might be able to use some of the very useful information
I was able to gather. Some of this information is:
A) Whats involved in setting up joint Software Development Projects
with the Soviets.
B) Inexpensive Electronic Mail facilities now available between
San Francisco and Moscow.
C) Do's and Don'ts on dealing with the Soviets.
D) Some very interesting opportunities for information exchange
between the Soviets and Western worlds.
I have just returned from the USSR, and for the past week, have
been recording and documenting my 3 week trip which covered Leningrad,
Tbilisi, and Moscow.
I was allowed total freedom to meet with whoever I wanted, and hung
out with the Soviet computer programmers from the Academy of Sciences
in Leningrad and Moscow. I know what kind of computers they are using,
and the languages they prefer to use.
I suspect that I'll have the paper completed by sometime over the
weekend, and are asking the UseNet gurus where this information might
reside.
If no specific newsgroup is set of for Joint Ventures with the Soviets
how does one set up something like that.
In 3 weeks, I'll be directly connected to Moscow through a Satellite
network, and I hope to be connected to a number of institutions using
computers, and special languages. I want to share these experiences
with others, and hope that someday they can ALSO visit the Soviet
Union and meet the fine people I had the chance to meet.
Please Email me at: uunet!acad!well!crunch
John Draper
∂10-Nov-88 1208 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 10 Nov 88 12:08:14 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02165; Thu, 10 Nov 88 12:07:40 PST
Date: Thu, 10 Nov 88 12:07:40 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811102007.AA02165@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: crunch@well.UUCP (John Draper)
Newsgroups: news.admin
Subject: Re Soviets
Message-ID: <7611@well.UUCP>
Date: 10 Nov 88 01:57:38 GMT
Reply-To: crunch@well.UUCP (John Draper)
Organization: Whole Earth Lectroinic Link, Sausalito, CA
Lines: 64
In response to the latest message I posted about Joint Ventures,
this is my response to all of TODAYS responses.
>I think the right place to post this info would be in comp.misc.
>Any chance of getting the Soviets on UseNet? I understand that most
>of the scientifically involved Soviets also speak English.
The Soviet programmers, at least 80 percent usually can read and
speak english because they have to be able to read all the manuals
which are in English.
Currently, I will be the gateway between here and the USSR, and am
trying to set up a UseNet link. I don't know the details yet.
If anyone out there can give me current info on usenet links into
Finland, or Checkloslovokia (Scuse spelling), please let me know.
Getting UseNet links into the USSR requires someone on the Soviet
side of an OFFICIAL capacity to authorize a connection to one of their
UNIX machines. They DON'T have many UNIX machines. Communications
between Finland and the USSR is about 6 marks a minute. Or about
$1.50 per MINUTE. There has to be a better way to do this. I have
people in the Soviet Union investigating this.
>How about comp.misc? Something like comp.society.russia would be better
>but that doesn't exist and I can't see it being created for just these
>articles.
> Jerry
Score 2 for comp.misc, any MORE requests.
> I have one question about the relations with the Soviets in this area... Wilr
> Will there be some sort of link established between say Usenet or the Internet?
>Granted current events raise questions about how secure such a link can be, bute
>to have established - make Usenenet truly world-wide. It would also help allow t
>exchange of information between people in the USA and USSR coincerninh their wo.
I currently have total freedom of information exchange with the Soviets. I
was totally surprised at their openness, and they have NO experience with
networks. I will have a DIRECT LINK to the Soviet Union, and I'm paying
for this out out of my own pockets, so don't expect megabytes of information to
be passed back and forth. The link I use is EXPENSIVE, but if used in
moderation, I can afford it.
I suspect I'll be using this link to extract the BEST stuff from Usenet
(Upon approval from the origional authors of course) then I can shoot it over
to Moscow. I'll have an account on the Moscow system. Most of the traffic
will be used to set up more direct links between the Soviet Union. BUT!!
If some agency is willing to foot the bill, I would be most happy to oblige,
and be the gateway. However, by my observations, I would like to have
some control of what gets passed, as in my experience, Americans tend to
say things to Soviets (Mostly un-intentional) that offend them. Currently,
I'm in "Fat city" as far as the Soviet Officials are concerned, and I intend
to respect their authority. Although the Soviet Union is opening up,
they STILL have different ways of viewing things, and are not experienced
in dealing directly with Americans. I think after you read the Joint
Venture paper (Soon to be published in comp.misc), you will understand
why I feel that way.
Anyway, so be looking for it in comp.misc. I have more changes to make
before putting it there. A few more days.....
Crunch uunet!acad!well!crunch
∂10-Nov-88 1703 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
INHERITANCE HIERARCHIES AND AUTOEPISTEMIC LOGIC
Michael Gelfond
Halina Przymusinska
University of Texas at El Paso
Friday, November 11, 3:15pm
MJH 301
We propose a new semantics for inheritance hierarchies with
exceptions. Our approach is based on a translation of an inheritance
hierarchy into a logical theory stated in autoepistemic logic. This
translation interprets normative statements of the form "Typically
property F(x) holds" as statements about the beliefs of an agent whose
premises are determined by the inheritance hierarchy. We hope that
this interpretation will provide some insights into the nature of
inheritance-based reasoning.
∂10-Nov-88 2352 harnad@confidence.Princeton.EDU BBS Call For Commentators: The Tag Assignment Problem
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 10 Nov 88 23:52:04 PST
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA24285; Fri, 11 Nov 88 02:50:27 EST
Received: by psycho.Princeton.EDU (3.2/1.64)
id AA00847; Fri, 11 Nov 88 02:38:39 EST
Date: Fri, 11 Nov 88 02:38:39 EST
From: harnad@confidence.Princeton.EDU (Stevan Harnad)
Message-Id: <8811110738.AA00847@psycho.Princeton.EDU>
To: neuron-request%ti-csl.CSNET@relay.cs.net
Subject: BBS Call For Commentators: The Tag Assignment Problem
Below is the abstract of a forthcoming target article to appear in
Behavioral and Brain Sciences (BBS), an international,
interdisciplinary journal providing Open Peer Commentary on important
and controversial current research in the biobehavioral and cognitive
sciences. To be considered as a commentator or to suggest other appropriate
commentators, please send email to:
harnad@confidence.princeton.edu or write to:
BBS, 20 Nassau Street, #240, Princeton NJ 08542 [tel: 609-921-7771]
____________________________________________________________________
A SOLUTION TO THE TAG-ASSIGNMENT PROBLEM FOR NEURAL NETWORKS
Gary W. Strong Bruce A. Whitehead
College of Information Studies Computer Science Program
Drexel University University of Tennessee Space Institute
Philadelphia, PA 19104 USA Tullahoma, TN 37388 USA
ABSTRACT: Purely parallel neural networks can model object recognition
in brief displays -- the same conditions under which illusory
conjunctions (the incorrect combination of features into perceived
objects in a stimulus array) have been demonstrated empirically
(Treisman & Gelade 1980; Treisman 1986). Correcting errors of illusory
conjunction is the "tag-assignment" problem for a purely parallel
processor: the problem of assigning a spatial tag to nonspatial
features, feature combinations and objects. This problem must be solved
to model human object recognition over a longer time scale. A neurally
plausible model has been constructed which simulates both the parallel
processes that may give rise to illusory conjunctions and the serial
processes that may solve the tag-assignment problem in normal
perception. One component of the model extracts pooled features and
another provides attentional tags that can correct illusory
conjunctions. Our approach addresses two questions: (i) How can objects
be identified from simultaneously attended features in a parallel,
distributed representation? (ii) How can the spatial selection
requirements of such an attentional process be met by a separation of
pathways between spatial and nonspatial processing? Analysis of these
questions yields a neurally plausible simulation model of tag
assignment, based on synchronization of neural activity for features
within a spatial focus of attention.
KEYWORDS: affordance; attention; connectionist network; eye
movements; illusory conjunction; neural network; object recognition;
retinotopic representations; saccades; spatial localization
∂11-Nov-88 0908 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Nov 88 09:08:13 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04850; Fri, 11 Nov 88 09:07:35 PST
Date: Fri, 11 Nov 88 09:07:35 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811111707.AA04850@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: geoff@sunfs3.camex.uucp (Geoffrey Knauth)
Newsgroups: news.admin
Subject: John Draper's note re: USSR & joint ventures in software
Message-ID: <257@sunfs3.camex.uucp>
Date: 10 Nov 88 19:32:18 GMT
Lines: 43
John Draper mentioned four major points:
A) Joint Software Development Projects with the Soviets.
B) Inexpensive e-mail available between San Francisco and Moscow.
C) Do's and Don'ts on dealing with the Soviets.
D) Some very interesting opportunities for information exchange
between the Soviets and Western worlds.
As one occasionally involved with US-USSR trade, I offer the following.
A) I have been approached by a programmer, Oleg Yachny of Akademgorok,
Novosibirsk, USSR, about the possibility of establishing a joint
software venture. Reaction in the US has ranged from "Look out, the
KGB is trying to recruit you" to "Check it out, but be careful." I
first became acquainted with Oleg as a pen-pal while at Harvard, and I
visited his home in Siberia in 1985.
B) The only SFO-SVU e-mail I know of was described in the 5/31/88
issue of MacWeek. It is a private venture between a somewhat bold
entrepreneur in the US and state agencies of the USSR. I call the
entrepreneur bold because he has had occasional legal difficulties in
the US.
C) What constitutes legitimate, legal trade with the USSR? No
American wants to find himself suspected of treasonous activity.
*************************************************************************
* The greatest obstacle to US-USSR trade and technical cooperation, I *
* believe, is the lack of a coordinating agency in the United States *
* that can serve to guide inexperienced, disorganized Americans in *
* dealing with the Soviet Union. *
*************************************************************************
D) The USSR is undoubtedly one of the most interesting places on
Earth. But not only are there opportunities and challenges--there are
also threats and traps, even for the seasoned trader/traveller. Until
the US develops an integrated, organized plan for US-USSR trade, I
think progress in trade relations will be, and must be slow.
--
Geoffrey S. Knauth ARPA: geoff%lloyd@husc6.harvard.edu
Camex, Inc. UUCP: geoff@lloyd.uucp or husc6!lloyd!geoff
75 Kneeland St., Boston, MA 02111
Tel: (617)426-3577 Fax: 426-9285 I do not speak for Camex.
∂11-Nov-88 0923 pehoushe@Gang-of-Four.Stanford.EDU My Vacation Time
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Nov 88 09:23:20 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04899; Fri, 11 Nov 88 09:22:42 PST
Date: Fri, 11 Nov 88 09:22:42 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8811111722.AA04899@Gang-of-Four.Stanford.EDU>
To: sloan@score, mps@sail
Cc: pehoushek@Gang-of-Four.Stanford.EDU, jmc@sail, clt@sail
Subject: My Vacation Time
Somehow my vacation time got truncated. I would like this corrected,
as per our discussions. I was hired on July 17, 1987 and deserve full
credit for accrual of vacation time from that date. If sick leave
has also been truncated, please replace that too. Thanks.
-Dan
∂11-Nov-88 0952 CN.MCS@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 11 Nov 88 09:52:12 PST
Date: Fri, 11 Nov 88 09:52:01 PST
To: jmc@sail
From: "Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>
John,
I have a question regarding the COMTEX (Scientific Datalink)
microfiche that came to Harry as part of the contract Stanford
signed for the technical reports.
Harry gave the fiche to the Interlibrary Loan department of SUL for
them to use when requests come in for CSD technical reports. Do you
think this use of these reports is an infringement of copyright?
Can SUL leagally make copies of these fiche and send them to
requestors? There is no copyright statement on the fiche. I am
asking you because you are the only one who seems to remember
anything about this Comtex deal. So any information you can supply
will be helpful.
Also, I am wondering if you have your own Dialog account? I am
beginning to talk with faculty about using Knowledge Index which is
an inexpensive way to search on Dialog after 6 pm. Are you
interested in this?
Oh, and the last thing. I did recommend your name to Ray Bacchetti
to serve on the NWC Committee. I am not sure when that committee
will meet.
Rebecca Lasher
∂11-Nov-88 1412 weening@Gang-of-Four.Stanford.EDU passwords
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 11 Nov 88 14:12:32 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA05917; Fri, 11 Nov 88 14:10:46 PST
Date: Fri, 11 Nov 88 14:10:46 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811112210.AA05917@Gang-of-Four.Stanford.EDU>
To: ag@Gang-of-Four.Stanford.EDU, air@Gang-of-Four.Stanford.EDU,
alex@Gang-of-Four.Stanford.EDU, andy@Gang-of-Four.Stanford.EDU,
arg@Gang-of-Four.Stanford.EDU, ark@Gang-of-Four.Stanford.EDU,
avva@Gang-of-Four.Stanford.EDU, barbara@Gang-of-Four.Stanford.EDU,
blee@Gang-of-Four.Stanford.EDU, carol@Gang-of-Four.Stanford.EDU,
clt@Gang-of-Four.Stanford.EDU, diana@Gang-of-Four.Stanford.EDU,
dkeisen@Gang-of-Four.Stanford.EDU, farhad@Gang-of-Four.Stanford.EDU,
ferziger@Gang-of-Four.Stanford.EDU, galbiati@Gang-of-Four.Stanford.EDU,
glb@Gang-of-Four.Stanford.EDU, grossman@Gang-of-Four.Stanford.EDU,
hbs@Gang-of-Four.Stanford.EDU, helen@Gang-of-Four.Stanford.EDU,
ilan@Gang-of-Four.Stanford.EDU, jk@Gang-of-Four.Stanford.EDU,
jmc@Gang-of-Four.Stanford.EDU, jonl@Gang-of-Four.Stanford.EDU,
kent@Gang-of-Four.Stanford.EDU, larus@Gang-of-Four.Stanford.EDU,
les@Gang-of-Four.Stanford.EDU, lincoln@Gang-of-Four.Stanford.EDU,
moreau@Gang-of-Four.Stanford.EDU, nancy@Gang-of-Four.Stanford.EDU,
okuno@Gang-of-Four.Stanford.EDU, orca@Gang-of-Four.Stanford.EDU,
pab@Gang-of-Four.Stanford.EDU, pchen@Gang-of-Four.Stanford.EDU,
pehoushe@Gang-of-Four.Stanford.EDU, pereyra@Gang-of-Four.Stanford.EDU,
rag@Gang-of-Four.Stanford.EDU, ramana@Gang-of-Four.Stanford.EDU,
ramani@Gang-of-Four.Stanford.EDU, rdz@Gang-of-Four.Stanford.EDU,
rivin@Gang-of-Four.Stanford.EDU, roach@Gang-of-Four.Stanford.EDU,
rpg@Gang-of-Four.Stanford.EDU, shankar@Gang-of-Four.Stanford.EDU,
simmons@Gang-of-Four.Stanford.EDU, simon@Gang-of-Four.Stanford.EDU,
trogon@Gang-of-Four.Stanford.EDU, tuminaro@Gang-of-Four.Stanford.EDU,
weening@Gang-of-Four.Stanford.EDU, yoram@Gang-of-Four.Stanford.EDU
Subject: passwords
Recent events indicate that several Stanford Unix systems may have
been targeted for password attacks. I would like to urge everyone to
change their password on Gang-of-Four and other systems. Passwords
should not be easy to guess, so they should be at least 6 characters
long, ideally with mixed upper/lower case or digits; not a person's
name or an English word, not a word spelled backwards, not the name of
a computer system, etc.
Your own password affects the integrity of everyone's account on the
system, so please choose it carefully and change it every few months.
∂13-Nov-88 1109 GKMARH%IRISHMVS.BITNET@forsythe.stanford.edu post-docs in cs
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 13 Nov 88 11:09:35 PST
Received: by Forsythe.Stanford.EDU; Sun, 13 Nov 88 11:08:48 PST
Date: Thu, 10 Nov 88 11:17 EST
To: jmc@sail.stanford.edu
From: steven horst <GKMARH%IRISHMVS.BITNET@Forsythe.Stanford.EDU>
Subject: post-docs in cs
Dear Professor McCarthy,
I am writing to inquire as to whether Stanford is offering
post-doctoral fellowships in cognitive science and related areas
for the 1989-90 academic year. I am a philosopher and currently
completing a dissertation consisting of an analysis and critique
of the "Computational Theory of Mind" as presented by Fodor and
Pylyshyn. (I shall gladly supply more details, but only if you
invite me to do so. Even a light teaching load has led me to
understand how crushing time demands can become.)
If any Stanford departments (or the research institute) are
offering post-docs, I should appreciate your letting me know whom
to contact and how and when to apply. Stanford seems a particularly
interesting place to me, not only because of the number and quality
of faculty, but because it seems to share my peculiar constellation
of interests in CS and AI, namely: (1) philosophical issues involving
intentionality, (2) the relationship of AI to understanding the
"life-word" (and hence its connections to phenomenology), and
(3) an interest in finding appropriate FORMALISMS for the
mathematization of psychology. (Not necessarily formal systems in
the narrow sense - what's wrong with eclecticism that includes
information theory and even statistics, if used in the kind of way
that my former teacher Stephen Grossberg uses stochastic methods?)
Thank you for your time. If you are not directly connected with
post-docs at Stanford, I apologize for the intrusion.
Sincerely,
Steven Horst bitnet: gkmarh@irishmvs
Department of Philosophy
Notre Dame, IN 46556
219-239-7458
∂14-Nov-88 1013 JK
John ---
Regarding EDI: I have now had contacts with both DEC and Boeing
regarding their interest with EDI. They both want to co-operate/or
support the research in some way. This needs to be defined. I sent
DEC a copy of your paper. You should expect email from Caroline
Dyer-Gunn from DEC some time in the near future.
Jussi
∂14-Nov-88 1157 SLOAN@Score.Stanford.EDU [Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>: My Vacation Time]
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Nov 88 11:57:45 PST
Date: Mon 14 Nov 88 11:55:19-PST
From: Yvette Sloan <SLOAN@Score.Stanford.EDU>
Subject: [Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>: My Vacation Time]
To: jmc@Sail.Stanford.EDU, clt@Sail.Stanford.EDU
Message-ID: <12446552577.15.SLOAN@Score.Stanford.EDU>
If you approve, I'll reinstate 45 hours of vacation time and 36 hours of sick
leave to his leave record. Please let me know. (NOTE: If he's laid off,
the 45 hours of vacation time will have to be paid to him.)
--Yvette
---------------
Return-Path: <pehoushe@Gang-of-Four.Stanford.EDU>
Received: from Gang-of-Four.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Fri 11 Nov 88 09:23:26-PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04899; Fri, 11 Nov 88 09:22:42 PST
Date: Fri, 11 Nov 88 09:22:42 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8811111722.AA04899@Gang-of-Four.Stanford.EDU>
To: sloan@score, mps@sail
Cc: pehoushek@Gang-of-Four.Stanford.EDU, jmc@sail, clt@sail
Subject: My Vacation Time
Somehow my vacation time got truncated. I would like this corrected,
as per our discussions. I was hired on July 17, 1987 and deserve full
credit for accrual of vacation time from that date. If sick leave
has also been truncated, please replace that too. Thanks.
-Dan
-------
∂14-Nov-88 1219 SLOAN@Score.Stanford.EDU
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Nov 88 12:19:17 PST
Date: Mon 14 Nov 88 12:16:53-PST
From: Yvette Sloan <SLOAN@Score.Stanford.EDU>
To: jmc@Sail.Stanford.EDU
Message-ID: <12446556503.15.SLOAN@Score.Stanford.EDU>
Professor McCarthy--
Carolyn told me that Pat could use some extra work and I've talked to Pat and
she agrees. We have two visitors here who need someone to go to when they
need things done. I've discussed this with Carolyn and Pat and they agree
on Pat's doing their work. Before making final arrangements with Betty as
to the percentage of Pat's salary the department will pay for her taking on
these two visitors, I wanted to make sure you were in agreement with this
arrangement. Please let me know how you feel about this. Thanks.
--Yvette
-------
∂14-Nov-88 1316 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Nov 88 13:16:28 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04807; Mon, 14 Nov 88 13:15:43 PST
Date: Mon, 14 Nov 88 13:15:43 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811142115.AA04807@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: crunch@well.UUCP (John Draper)
Newsgroups: comp.misc
Subject: Soviet Access to Usenet
Message-ID: <7649@well.UUCP>
Date: 13 Nov 88 21:17:16 GMT
Reply-To: crunch@well.UUCP (John Draper)
Organization: Whole Earth Lectroinic Link, Sausalito, CA
Lines: 155
Hi,
I recently returned from the Soviet Union, Met a LOT of programmers,
Educators, and their people of Technology. There is a LOT of amazing
changes goiong on over there right now. It's not the usual rhetoric of
Glasnost and Peristroika, it's more than that.
I was so inpired as the results the trip, that I'm just about to
publish my experiences while in the USSR. It was nothing short of
Amazing, and contains very useful information on setting up Joint
Ventures or study groups. It dissolves a LOT of myth about the
Soviet Union, and covers all the trivia with a Hackers eyes view of
the Soviet Union, and Soviet Hackers Lifestyles, which are very MUCH
similar to ours. Be looking for it HERE in "comp.misc" as soon as I finish
it.
Contrary to popular belief, Modems are NOT illegal in the USSR, instead
they are very much prized posessions. Anyone can own one, IF they can
get them.
Geeee!!! Lets start a Modem Drive...!!! Just kidding... But really!!
But can we DO this on Usenet. Hmmm Probably not. :-|
There are UNIX sites that exist in the Soviet Union, but only a FEW
are using UUCP. I have connections that can give me more information on
UNIX sites in the USSR, I just need to know what questions to ask.
I have heard a LOT of talk about adding Soviet Sites to the UUCP network
but have heard nothing but VAPORWARE. Does anyone out there in Net land
WANT to add Soviet sites?? I can think of a hundred reasons why!!
not to mention what it would do towards World Peace.
Imagine day-to-day communication with Soviet programmers, hackers
(I have met MANY), and Educators. They ALSO have virus problems,
software piracy (Mostly OURS), and most of all, Equipment Shortages.
They get payed MUCH less than we do, and have the Social status of
a clerk or secretary. But their style of programming is totally
amazing. Perhaps later, I can give you some examples. For instance,
when they got infected by the IBM-PC virus, they probably said...
Hmmmm!! Whats going on here?? Go into Debug, chase through the IBM-DOS
or operating system code, located it, and remove it. It's surprising
how MANY Soviet people who have PC's know how to do this. They think
NOTHING about going into the Machine code and patching commercial products.
Naturally, they have to be educated about the importance of Intellectual
property, and they would ALWAYS pay for American software if they were
ALLOWED to pay in Rubles. Unfurtunately, Soviets cannot pay for foreign
goods with Rubles, at least not Legally. Eventually, this will change,
as I was assured with my recent visit to the Soviet Union. Earlier, there
had been some publications mentioning that the Russians were stealing our
software. But each Soviet computer user I came in contact with,
expressed to me that they would Gladly pay for software licenses and support
if they were ALLOWED to pay in Rubles.
One IMPORTANT consideration and policy I'm adapting, is that if I see
an article worthey of sending to the Soviet Union, I will contact the origional
author FIRST and obtain permission. This would usually be for long and
informative articles and papers. However, I might NOT do this if I send
over "Idle chit chat" discussing important issues. Another equally important
consideration are the trade restrictions regulating the importation of certain
kinds of computer data to the Eastern block nations. I have ordered a copy
of the regs, and if anyone is interested, I'll summerize them. These were
enacted by the Expost Administration Act of 1979. Surly you all must remember
the Reagan Over-reaction, because the Russkies were obtaining Western
Technology. I leaned that the Soviets STILL get high tech parts from OTHER
countries. But WE should be careful, and take the responsibility to abide
by rules. I have so informed the Soviets, and will be getting copies of
THEIR rules and controls.
The current topics of discussion over this link should be:
a) Tips on setting up Joint Projects or ventures with the Soviets, such
as a list of American institutions wanting to work with Soviets, and
vice versa.
b) Soviet techniques for virus prevention and removal. They apply a
very **direct** solution to the problem.
c) New ideas for development tools from the Soviets, they are
Exceptionally good in this department. They are especially strong
in Natural Language development, AI, Object Oriented Programming,
and writing their OWN commercial quality programs. We have a LOT
to learn from them. I know I certainly did.
d) The Soviets are weak in free enterprise, and have NO experience,
largely because until just recently, were NOT Allowed to. They
ALSO want to start selling software, both within the Soviet Union,
and to Americans.
e) Soviets are also into Robotics, and factory automation. But MOST
importantly, regular Soviet citizens are snapping up PC's as soon
as they become available. Especially modems.
f) Soviets want to PAY for American software products, but currently
NO mechanism exists to allow this to happen easily. Comments and
suggestions for solving this problem are always welcome.
g) Reports on Soviet Trade shows to Americans.
h) Reports on American Trade shows to Soviets.
They have a 2 hr TV program in the morning that educates the public
about computers, and even have programming classes in 8086 assembly language,
Pascal and Basic. When I watched it, they were explaining how to patch
the BIOS so a Bulgarian printer will work with a PC. I don't completly
know Russian, but enough information was in English for me to get an
idea. It's amazing that material like this is broadcast over nation-wide
TV. Another show "120 minutes", also broadcast in the morning, constantly
informes the Soviet citizen about the importance of computers, and how they
help produce hight quality goods in their stores.
Their Cyrillic fonts of ascii characters above 0x80 are activated from
the keyboard by shift lock. The video driver is available from the
Academy of sciences for the asking.
In about 3 weeks, my SF/Moscow Data Teleport service will be firmly
established, enabling me to send and recieve Email from Moscow instantly.
This service is so inexpensive that I'm trying it for 6 months or so.
If anyone wants details, call (415) 931-8500 and ask for details. The
prices are $15/hr connect time (About the same as BIX I think), and $25/month
for BASIC service, and $75/month for Extended service, such as follow-up
for un-answered messages by phone calls, stimulation of timely responses
from your Soviet counterparts, technical training on the Soviet side, as
well as Email access FROM the Soviet Union TO the USA if you plan on traveling
there.
I have established an amazing list of contacts who ALSO will be getting
the teleport service on the Soviet side, dedicating towards setting up and
using a UUCP network, then I will have direct communication with them. I
also got the Extended service that provides me with phone call followup
messages to Soviets NOT connected to the system. This will enable them to
drop down to the local Teleport office and Email me a message. Or having
my Soviet contact call them on the phone to dictate a message to me.
If anyone here in NetLand wants to closly work with me, to establish
this UUCP network, please Email me, and let me know what YOU can do to
help facilitate the UUCP link. What we need is: A Unix site interested
in maintaining DIRECT connection to the Teleport, enabling Soviet users to
dial a LOCAL Moscow number, connecting DIRECTLY to your site. The
American site must make arrangements or provide a joint venture so that the
Soviet side maintains an office, accepts applicants for users, and sets
them up with an account.
So, what do you want to know about the Soviet Union, Please make your
resuests now, and flood my mail box. I'll gather up your requests and
Email them to my friend at the Academy of Sciences of the USSR, and lets see
what the Ruskies have to say. They are eagerly awaiting your questions.
Email me at: uunet!acad!well!crunch - Personal
or uunet!acad!crunch - If related to AutoDesk Business
Till later....
Crunch
∂14-Nov-88 1446 CN.MCS@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 14 Nov 88 14:46:51 PST
Date: Mon, 14 Nov 88 14:46:01 PST
To: jmc@sail
From: "Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>
Igor,
Complex Systems volume 1 is 30 days overdue. Unless this volume is
returned to the library by Wednesday, November 16, 1988 I will
consider suspending your privilege of borrowing materials from the
Math/CS Library.
Your record of not returning materials when needed by other
borrowers and not returning journals in a timely manner shows that
you are inconsiderate of others needs. This library, in general,
has only one copy of the items you check out. Other researchers
depend on these materials, it is your responsibility to return
materials when requested. As for journals, the Math/CS staff
expects you to keep a journal for only 24 hours. We should not have
to remind you constantly of the due date.
Rebecca Lasher
Head Librarian
Math/CS Library
cc: John McCarthy
∂14-Nov-88 1542 CN.MCS@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 14 Nov 88 15:41:57 PST
Date: Mon, 14 Nov 88 15:41:05 PST
To: jmc@sail
From: "Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>
Igor,
I hadn't realized that you had started returning materials more
promptly. I am somewhat removed from this day-to-day circulation
activity. So I should apologize for expressing my frustration in my
previous memo.
However, I do not consider a journal 30 days overdue to be a good
demonstration of promptly returning library materials. Journals are
different, they should be returned in the 24 hours. I realize that
you might not have known this. If we had to recall every journal
that left the library, research would slow down markedly. We try to
tell each journal borrower that the journal is expected back in 24
hours. We do not fine borrowers unless someone else wants the
material. But for journals we often call or request return of the
item without another borrower.
So now you know about journals and I know about your new leaf.
Great.
Rebecca Lasher
Math/CS Library
cc: John McCarthy
∂14-Nov-88 1635 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 14 Nov 88 16:35:06 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA05479; Mon, 14 Nov 88 16:33:55 PST
Date: Mon, 14 Nov 88 16:33:55 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811150033.AA05479@Gang-of-Four.Stanford.EDU>
To: qlisp@Gang-of-Four.Stanford.EDU
From: ray@think.com (David M. Ray)
Newsgroups: comp.parallel
Subject: TMC Connection Machine Network Server
Message-ID: <3540@hubcap.UUCP>
Date: 14 Nov 88 17:22:37 GMT
Lines: 81
A n n o u n c i n g . . .
The Connection Machine Network Server Pilot Facility
(available from an Internet host near you)
Thinking Machines Corporation has recently completed the installation
of a Connection Machine on the Arpanet for use by members of the network
community. The Connection Machine Network Server (CMNS) pilot facility
is being funded by DARPA contract DACA76-88-C-0012 for the purpose of
investigating the use of CM's over wide-area networks and developing the
software and environments required to support remote users.
The CMNS pilot facility hardware currently includes a 16K CM-2 (without
floating point), a 5 Gbyte DataVault, and a VAX 6210 front-end processor.
Upgrades to a 32K CM-2 with floating point and additional front-end
capacity are under consideration. The software configuration consists
of Release 5.0 of the CM system software running on Ultrix version 3.0.
The network server front-end is directly accessible from the Arpanet as
Internet host cmns.think.com (192.31.179.100). We are also actively
pursuing high speed connections to mid-level regional nets of NSFNet.
The pilot facility has a full-time staff including a facilities manager,
application engineer, and system software developer. System operator
and hardware support is also available during normal hours of operation.
The facility is currently 'open' weekdays from 10 a.m. to 6 p.m. EST.
This notice constitutes an open invitation to CM users, programmers,
enthusiasts, and other interested parties to actively participate in
this effort to make the CM into a true Network Machine. We at TMC have
lots of ideas about how to accomplish this but relatively little actual
experience (up to now) in supporting network-based CM users. We would
welcome outside participation at any of the following levels:
- if you simply reply to this mail message we will include you in the
mailing list for news about networking the CM and updates on the
CMNS project; we would also appreciate hearing which of the following
areas are of particular interest to you:
a) high-speed wide-area network connections for the CM
b) batch and time-sharing support for CM applications
c) system management aids and policies for CM environments
d) DataVault file server and data management utilities
e) on-line documentation, tutorials, source archives
f) remote CM graphical applications, X windows, user interfaces
g) remote procedure call (RPC) support for CM applications
h) all of the above
- if you would like an account on cmns.think.com and a chance to try
out new network-oriented features (or just a chance to use the CM)
please reply and include the following:
your name
organization
desired login
a phone # where you can be reached if need be
type of CM applications you're interested in
current level of familiarity with the CM
favorite CM language (*Lisp, C*, CM Fortran) if any
volume of data files (ball park) you will need
- if you would be interested in attending a TMC-sponsored workshop
on Networking the Connection Machine let us know; we will plan
according to the level of outside interest
If you have further questions or comments about the pilot facility or
the CMNS project in general please get in touch with me. I'm looking
forward to hearing from you.
David Ray
Connection Machine Network Server Project ray@think.com
Thinking Machines Corporation harvard!think!ray
245 First Street (617) 876-1111
Cambridge, MA 02142-1214 x600
∂14-Nov-88 1934 GLB meeting
Maybe there are some questions you want to ask about my work
before the exam next week and I'd be glad to talk about it.
Gianluigi
∂14-Nov-88 2152 DEK
To: JMC@SAIL.Stanford.EDU, JJW@SAIL.Stanford.EDU
∂14-Nov-88 2149 ME re: the big boat returns
∂13-Nov-88 0649 DEK the big boat returns
Congratulations on another long run (free of viruses etc)
ME - Thanks. I had a scare on Saturday afternoon, though, when I thought
we might not make it to the one month point. I had created a very big
file, which never got closed properly. Then when I tried to delete it,
the deleting job kept doing more and more diskops. It was doing one per
revolution of the disk (60/second). I was worried that it might be
trashing the disk, so I didn't dare let it continue for long. But I also
thought that in trying to stop it, I might crash the system (only hours
before the sailing ship was to appear). I even considered letting it run
for those few hours so I could see the ship, and then to try to kill it.
But, I studied the problem a bit and found that the deleted file simply
contained a circular list of disk blocks that were being freed. So all I
had to do was break the circle. And, whew, the diskops stopped quickly
when I did that. (With nary an error message on the CTY.)
This is only the 6th SAIL system to last a month or more. And in less
than 6 hours, it will (if still running) be the fourth longest lived SAIL
system.
Now I know what I want for Christmas. I just calculated that if this
system runs till then, it will become the longest lived SAIL system.
And a day or so before my birthday, it would pass 2000 hours. And on
about Zohar Manna's birthday, it would become the longest lived WAITS
system (S1-A ran one for almost 2294 hours). Then on 21 Jan 89, it would
reach 100 days. Well, that ought to be enough records to scare a system
to death, even if the power does stay on that long.
∂15-Nov-88 0955 VAL Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
THINGS THAT CHANGE BY THEMSELVES
Vladimir Lifschitz
Arkady Rabinov
Stanford University
Friday, November 18, 3:15pm
MJH 301
This talk is about the frame problem, which we understand as the problem of
formalizing the commonsense law of inertia. One interpretation of the law
of inertia is that a fluent doesn't change after an action is performed,
unless the action causes it to take on some value. We argue that, in many
domains, this interpretation is inadequate. First, some actions have effects
that are indirect "ramifications" of the changes that it causes. Second,
some fluents, such as time, change even after an action that is not assumed
to have any causal effects whatsoever, like "wait." We propose a more
flexible formalization of the commonsense law of inertia, that allows us
to describe some examples of these two types.
∂15-Nov-88 1513 debra@russell.Stanford.EDU EVENING SEMINAR
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 15 Nov 88 15:12:56 PST
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Tue, 15 Nov 88 15:14:51 PST
To: etch@russell.Stanford.EDU, sag@russell.Stanford.EDU, shoham@score,
der@psych, helen@russell.Stanford.EDU, peters@russell.Stanford.EDU,
barwise@russell.Stanford.EDU, HERB@psych, ECLARK@psych,
bratman@russell.Stanford.EDU, jmc@sail, amos@psych, nilsson@score,
latombe@coyote, genesereth@score, betsy@russell.Stanford.EDU,
john@russell.Stanford.EDU
Cc: debra@russell.Stanford.EDU, kuder@russell.Stanford.EDU
Subject: EVENING SEMINAR
Date: Tue, 15 Nov 88 15:14:50 PST
From: Debra Alty <debra@russell.Stanford.EDU>
REMINDER
The first EVENING SEMINAR will take place this Wednesday, November
16th @ 7:00 pm in the CSLI Cordura Conference Room.
Professor David E. Rumelhart, Psychology Department, will be leading
the first discussion.
The following will be served (so come hungry):
Cheese & Crackers Cognac, Courvoisier
Vegetable platter Wine, Chardonnay, Chablis
Fruit Calistoga
Chocolates Coffee
Tea
Hope to see you there.
∂15-Nov-88 1628 JANLEE%VTVM1.BITNET@forsythe.stanford.edu History of AI and Time Sharing
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 15 Nov 88 16:27:54 PST
Received: by Forsythe.Stanford.EDU; Tue, 15 Nov 88 16:27:03 PST
Received: by VTVM1 (Mailer X1.25) id 3110; Tue, 15 Nov 88 15:01:07 EST
Date: Tue, 15 Nov 88 14:06:18 EST
From: JAN <JANLEE%VTVM1.BITNET@Forsythe.Stanford.EDU>
Subject: History of AI and Time Sharing
To: John McCarthy <jmc@sail.stanford.edu>
Firstly my many thanks for the review of the Bloomfield book which
appears in the next issue of the Annals. If I had realized what an
extensive essay you had produced before I saw it in Bill Aspray's
manuscript for the issue, I am sure that I would have asked you to
turn it into a "real" article for the issue. I hope that our readers
will read it not just as a review of someone else's work but as a
overview of the history of AI for its own sake.
Turning to Project MAC, CTSS and the history of time sharing, let me
tell you of our plans. As you know we asked some of the principals
of both CTSS and the time sharing group from project MAC to meet with us
on the day following the seminars. I am sorry you were not able to
join us. However we had a reasonably successful day and hope to have the
transcripts of the meeting out by the end of the year, and intend to
ask for two things from that point: (1) for those who attended to
fix what they said and to add (in footnotes) what they wish they had said;
(2) for those who were unable to join us (like yourself, but also Dick
Mills, Bob Graham and others), to add your own notes and comments as
if you had been part of the oral sessions. I am not sure how we are
going to handle these extra comments; when I did the FORTRAN issue
I included the external comments as footnotes. We could include them
transparently as if you had been there, or we could keep them completely
separate.
We see the special issue of the ANNALS as being similar in some respects
to the Burroughs issue. I would like start out with two articles --
Chris Strachey's article from the UNESCO conference (1959) and your article
from Greenberger's book. Teh we need to select some seminal materials
from the various publications (and non-published materials) on CTSS and
Project MAC. Fortunately many of them were in AFIPS proceedings so
we will not have any difficulty with the copyrights. Gluing this
all together, and adding commentary which places the articles and
reports in context is the diffcult part for the editors (Bob Rosin
and myself).
If you have notes or comments that you can add now, and which might give us
some direction finding for the rest of the collection, we would be very
happy to receive them. Your suggestions for places to look for information
and items to seek would also be helpful.
Many thanks for your note at the seminar and my regrets that you were
unable to join us for Friday's activities. I look forward to hearing
from you.
My physical address is:
J.A.N. Lee, Editor-in-Chief
Annals of the History of Computing
133 McBryde Hall
Virginia Tech
Blacksburg VA 24061-0119
JAN
∂15-Nov-88 2226 harnad@confidence.Princeton.EDU BBS Call for Commentators: Are Species Intelligent / J. Schull
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 15 Nov 88 22:26:01 PST
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA11461; Wed, 16 Nov 88 01:24:39 EST
Received: by psycho.Princeton.EDU (3.2/1.64)
id AA01021; Wed, 16 Nov 88 01:03:00 EST
Date: Wed, 16 Nov 88 01:03:00 EST
From: harnad@confidence.Princeton.EDU (Stevan Harnad)
Message-Id: <8811160603.AA01021@psycho.Princeton.EDU>
To: epsynet@uhupvm1.BITNET
Subject: BBS Call for Commentators: Are Species Intelligent / J. Schull
Below is the abstract of a forthcoming target article to appear in
Behavioral and Brain Sciences (BBS), an international,
interdisciplinary journal providing Open Peer Commentary on important
and controversial current research in the biobehavioral and cognitive
sciences. To be considered as a commentator or to suggest other appropriate
commentators, please send email to:
harnad@confidence.princeton.edu or write to:
BBS, 20 Nassau Street, #240, Princeton NJ 08542 [tel: 609-921-7771]
____________________________________________________________________
ARE SPECIES INTELLIGENT?
Jonathan Schull
Haverford College
Haverford PA 19041
KEYWORDS: animal behavior; artificial intelligence; cognitive science;
evolution; intelligence; natural selection; parallel distributed
processing; punctuated equilibria; species
Plant and animal species are information-processing entities of such
complexity, integration and adaptive competence that it may be
scientifically fruitful to consider them intelligent. This
is suggested by the analogy between learning (in organisms) and
evolution (in species) and by recent developments in evolutionary
science, psychology and cognitive science. Species are now described
as spatiotemporally localized individuals in an expanded hierarchy of
biological entities. Intentional and cognitive abilities are now
ascribed to animal, human and artificial intelligence systems which
process information adaptively and exhibit problem solving abilities.
The structural and functional similarities between such species are
extensive, although these are usually obscured by
population-genetic metaphors (which have nonetheless contributed much
to our understanding of evolution).
In this target article I use Sewell Wright's notion of the "adaptive
landscape" to compare the performance of evolving species with those
of intelligent organisms. With regard to their adaptive achievements
and the kinds of processes by which they are attained, biological
species compare very favorably with intelligent animals in virtue of
interactions between populations and their environments, between
ontogeny and phylogeny, and between natural, interdemic, and species
selection. Whatever the answer, addressing the question of whether
species are intelligent could help refine our concepts of intelligence
and of species and could open new lines of empirical and theoretical
inquiry in many disciplines.
∂15-Nov-88 2251 harnad@confidence.Princeton.EDU Genetic Similarity Theory: BBS Call for Commentators
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 15 Nov 88 22:51:21 PST
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA12629; Wed, 16 Nov 88 01:50:08 EST
Received: by psycho.Princeton.EDU (3.2/1.64)
id AA01037; Wed, 16 Nov 88 01:24:43 EST
Date: Wed, 16 Nov 88 01:24:43 EST
From: harnad@confidence.Princeton.EDU (Stevan Harnad)
Message-Id: <8811160624.AA01037@psycho.Princeton.EDU>
To: epsynet@uhupvm1.BITNET
Subject: Genetic Similarity Theory: BBS Call for Commentators
Below is the abstract of a forthcoming target article to appear in
Behavioral and Brain Sciences (BBS), an international,
interdisciplinary journal providing Open Peer Commentary on important
and controversial current research in the biobehavioral and cognitive
sciences. To be considered as a commentator or to suggest other appropriate
commentators, please send email to:
harnad@confidence.princeton.edu or write to:
BBS, 20 Nassau Street, #240, Princeton NJ 08542 [tel: 609-921-7771]
____________________________________________________________________
GENETIC SIMILARITY THEORY
J. Philippe Rushton
Psychology Department
University of Western Ontario
KEYWORDS: Sociobiology; Inclusive Fitness; Kin Selection; Assortative
Mating; Dyad Formation; Ethnocentrism; Friendship; Behavior Genetics;
Altruism; Group Selection
A new thoery of attraction and liking based on kin selection suggests
that people detect genetic similarity in others in order to give
preferential treatment to those who are most similar to themselves.
Empirical and theoretical support comes from (1) the inclusive-fitness
theory of altruism, (2) kin-recognition studies in animals raised
apart, (3) assortative mating studies, (4) favoritism in families,
(5) selective similarity among friends, and (6) ethnocentrism.
Specific tests of the theory indicate that (a) sexually interacting
couples who produce a child together are genetically more similar to
each other in terms of blood antigens than they are to either sexually
interacting couples who fail to produce a child together or to
randomly paired couples from the same sample; (b) similarity between
marriage partners is greatest on the more genetically influenced sets
of anthropometric, cognitive, and personality characteristics; (c)
after the death of a child, parental grief intensity is correlated
with the child's similarity to the parent; (d) long term male
friendship pairs are more similar to each other in blood antigens than
they are to random dyads from the same sample; and (e) similarity
among best friends is greatest on the more genetically influenced sets
of attitudinal, personality and anthropometric characteristics.
Possible mechanisms are discussed. These findings may provide a
biological basis for ethnocentrism and group selection.
∂16-Nov-88 1320 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 16 Nov 88 13:20:34 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02262; Wed, 16 Nov 88 13:19:35 PST
Date: Wed, 16 Nov 88 13:19:35 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811162119.AA02262@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: crunch@well.UUCP (John Draper)
Newsgroups: comp.misc
Subject: A HACKERS view of the Soviet Union
Message-ID: <7662@well.UUCP>
Date: 15 Nov 88 20:53:06 GMT
Reply-To: crunch@well.UUCP (John Draper)
Organization: Whole Earth Lectroinic Link, Sausalito, CA
Lines: 878
After recieving no initial objections about my intention to post this
long comprehensive article on the Soviet Union, HERE IT IS. It's LONG
so you might want to dump it into a file for reading later, or capture
it, or whatever....
Join Ventures with the Soviets
(c) by John T. Draper
Programmers Network
***********************************
* Permission to redistribute this *
* article is granted as long as *
* the original copyright notice *
* is retained. *
***********************************
I had heard about Joint Ventures between Americans and Soviets before I left
for vacation, but I didn't know the level of interest was as great as it
currently exists in the Soviet Union (SU). It was the sincere eagerness on the
part of the Soviets that inspired me to learn more about joint ventures because
that's ALL they ever talked about. Not Really!! but Almost!! at least I get
a pretty good ear bending.
The economy of the USSR is currently very bad, as prices for goods are
about 20% higher than in the USA, and citizens are paid very low in comparison
with Western countries. In an attempt to boost their economy, Mikhail
Gorbachev is allowing citizens to engage in private enterprise. In my opinion,
this is good, but there are serious problems that still exist. Partly,
because while the SU has been stuck in an economic rut for a long time before
Perestroika was instigated, the Western countries have been pushing ahead in
high technology. The Soviets are very much behind in this department, and our
Government is not interested in allowing the Soviets to "Catch up", and has
enacted trade restrictions since 1979.
As we all know, by the news reports, tremendous changes are now taking
place in the USSR that can be VERY LUCRATIVE for American Businesses. There
are, of course, tremendous hurdles that have to be taken. The main problem
is their money. As far as the world market is concerned, the Ruble is not a
HARD currency, and cannot be taken out of the USSR. So, any business
relationship with the Soviets would have to result in Americans getting paid in
other Soviet goods, or services. You can actually take currency OUT of the
USSR if you pay their government 30 percent in taxes. So it's important to
keep the currency in the US. Relations like having Soviet programmers develop
software for the US market might not be so bad, and they would really LOVE to
get their hands on some recent equipment.
One might think that the Soviet goods aren't worth anything, this is
partially true, but the Soviets have more things to offer than Hard Currency.
I'll outline these things below, which have been suggested by some of the
Soviets I came in contact with.
The Office space for Computers offer
====================================
I talked with one person in Leningrad who uses PC's for the Hotel and
booking business. His current idea is to offer any American company free
office and hotel space for visiting executives in return for old IBM PC's. This
person uses the PC's connected to LANS for organizing and managing all the hotel
bookings, opera tickets, and other booking services, and is going in a very
big way to computerizing their business. They are writing their OWN software,
and almost ALL Soviets who have PC's can program them on just about ANY level.
This offers some rather interesting benefits, especially to those companies
that would want to establish offices in Leningrad.
When I asked how many PC's they were interested in, they said thousands.
I'm not sure what the US Laws are regarding exporting IBM-PC's to the Soviets,
but if someone who has an import license wants to pursue this, the
opportunities are amazing.
Leningrad has some very tight hotel and office space, and if any company
who has old IBM PC's, and wants to replace them with more modern systems,
one might benefit from such an arrangement.
I have the name of the Soviet attorney who is an expert on Soviet Law in
respect to arrangements like this, and if anyone is really interested in
following through on this, then contact me at the address mentioned below.
This arrangement gets around the Ruble soft currency problem, but involves
getting PC's over to the USSR. That's for the Attorneys to investigate. I'm
just taking these ideas down from my journal as I talked to the Soviets.
One thing really amazing about the Soviets is their discipline and lack of
"flakiness". I have heard reports on the contrary, but MY dealings with them
have been VERY POSITIVE.
Why the Soviets always pirate American software
===============================================
I had often thought that the Soviets always liked to pirate American
commercial software, but while talking to a group of Soviets at a dinner, one
person came up to me and said: "I would GLADLY pay for a legal copy of
FrameWork so I can get technical support", "we Soviet people are not allowed
to pay for our software in Rubles". He went on to mention; "It is illegal for
us to possess". If American software companies want to discourage Soviet
pirating, it would be in their best interests to set up joint ventures, and
field offices in the Soviet Union. An increasingly large number of reputable
software companies are doing this with great success.
Because the Soviets cannot pay in hard currency, one must examine what the
Soviets can offer us instead of money. Naturally, the SU manufacturers other
goods not obtainable in the USA, such as Vodka, Furs, and other non-tech
goods. A lot of the earlier joint ventures usually result in the Americans
getting paid in Soviet goods, who can turn around and sell them in the USA for
a mark-up.
Other than Soviet goods, the SU has a very rich supply of Programming
talent, much more disciplined than the American programmer. The average
Soviet programmer get paid anywhere between 250 - 500 rubles a month, but they
also have much less day-to-day expenses, as their average monthly rent for
government provided housing is around 3 - 6 rubles/month, and phone service is
essentially free or very in-expensive, except of course calling into the USA,
which is approximately 30 rubles for 5 minutes, which is prohibitive for 90% of
anyone.
When a Soviet person obtains a PC, they will learn EVERY aspect of the
computer, especially how to write Assembly Language programs for it. This is
necessary, because equipment is VERY HARD to come by, and the PC Owner will
have to write their OWN driver code to interface it with some printer obtained
in Bulgaria or anywhere else they find them. Almost EVERYTHING they do on a PC
is "Home Brew", but I did see a lot of commercial Software in use. Especially
the Norton Utilities.
The Academy of Sciences in Moscow had released (Free of charge) to all PC
users, a video driver providing the Cyrillic (Russian alphabet font). Usually
this means that if Shift lock is pressed, then typing will be in Cyrillic. The
Academy also offers key covers with Cyrillic characters.
Most of the languages used are Turbo Pascal, FORTH, C, and LISP in about
that order of importance. I was very surprised to find such a high interest
in FORTH. The FORTH system most often used in the Soviet Union is
"Astro-forth" from Project Learner in Leningrad. It's widely used in the SU,
and is very inexpensive. Igor Agamirzyan is the author, and works for the
Institute of Information and Automation of the Academy of Sciences of the USSR.
This system is used widely for factory automation and robot control. I also
met Sergei Baranoff, the author of the first book on FORTH written in Russian.
The most preferred C language is Turbo and Microsoft C, and Common LISP is
used among the University Crowd, and at the Academy of Sciences in Moscow. A
LOT of work is being done on natural languages, textual databases, and AI.
I was invited to the Academy of Sciences in a very uncommon move, I was
shown some fairly modern work stations of some unknown type. The monitors
were in Russian, and appeared to have 1200 X 768 Pixel resolution, and were the
very best systems available. No details were given on their operating system,
as they were used by the Scientific community, some of which were using them in
conjunction with the military. Just the fact that I was even allowed inside a
facility like that really gives me the impression of their openness.
The people at the Academy of Sciences in Moscow are VERY interested in
computer security, and the proliferation of computer viruses. They get BYTE
magazine, and a host of other American magazine, usually about 2-3 months
late. Most are brought in by foreigners, as there is a large network of
foreigners constantly bringing in consumables like diskettes, printer paper,
and other commodities. Mailing these things into the Soviet union,
especially diskettes, will usually get lost in the ozone. Either the
American officials will snatch it, or the Soviet officials. This explains
why the Soviets cannot "Mail order" any supplies from the West. Diskettes are
NOT OBTAINABLE for the private personal computer user. You CAN get them if
your Official status allows you to do so.
The social status of the typical Soviet programmer is that of a clerk or
secretary. Probably because children and EVERY citizen is exposed to a 2 hour
TV show aired in the morning. It's VERY educational, and features computers
and technology. When I watched it, someone was writing a flow chart for a
velocity and orbit program, then he was typing it in Basic, showing the
listing IN ENGLISH, then running the program. A bouncing ball appeared on
the screen, and something orbiting. He meticulously worked out the math and
was explaining what he was doing. About a half hour later, the person was
explaining how to patch a driver to work with some strange printer. All this
is on NATIONAL TV, and almost EVERYONE UNDER 16 had knowledge of Pascal and C.
Some teenagers I met through the Peace Committee were Obviously well educated.
The Soviets citizen is into peace in a very big way. Huge red signs
saying "Peace to the Earth" are posted everywhere, and Soviets have every
reason for wanting peace, as all throughout history, they've been bombed,
invaded, and harassed by outsiders.
In January of 1987, new laws were enacted by the Supreme Soviet allowing
its citizens to engage in setting up private and joint ventures. I explain
some of this later.
I had many discussions with Soviet programmers eager to gain experiences
setting up businesses. One such idea was proposed by Eduardo, the son of a
Soviet Diplomat I met on the train between Helsinki and Leningrad. Richard was
the diplomats name, and he's from Vienus, Lithuania USSR. He asked me when
I was going to be in Moscow, and sent his son to meet me in Moscow. We talked
for 2 days, hopping from coffee shop to coffee shop along the Arbot, a popular
hangout for artists and musicians. Eduardo made the following proposal idea:
Eduardo suggests having American Software firms invite Soviet programmers to
participate in a "Work Experience" program of about 6 months. USA companies
would pay the Soviet Programmer Competitive rates, and work on Software
products that are of interest to the American companies. After the 6 month
"Internship", the Soviet programmer would return to the USSR to finish his
share of the work. The Soviet programmer would take a PC back to the USSR,
provided by the American company as partial payment for the Software he/she is
to write. Keeping in mind that Soviet winters are hard, the work habits of
Soviet programmers (Who often work at home), are nothing short of amazing. This
offers some very good opportunities for American companies in a tight bind, that
could benefit from such an arrangement.
Soviet Programmers are allowed to travel to the USA, but they must be
invited, and the Americans must pay for their stay while in the USA, because
Soviets have no HARD currency. Soviets CAN however make it to either Wash DC
or New York on Aeroflot (Which takes Rubles). It costs the Soviet person
about 800 rubles for a "round trip" to NYC or Wash DC if the trip were planned
in advance, as they usually are. It also can take up to 5 - 6 weeks for all
the bureaucratic processes to grind to completion. Not only do we have to
deal with the local Soviet Consulate, but I suspect the State Dept might also
have to be involved. I've never had the experience of inviting a Soviet to
the States, so I don't know ALL the ropes, but it goes something like this:
Getting Soviets to USA involves first sending a letter to the Soviet
Programmer inviting them to the USA. There are 2 ways to do this.
A) A personal invitation, and
B) a Corporate or Institution invitation.
I'll explain the advantages and disadvantages of each. If we send a
Corporate invite, we cannot pick the Programmer we want to visit us, as when
working through the Official Organs, the institution usually will decide who
goes, by picking a person least likely to defect. If we send a PRIVATE
invitation, we CANNOT use company letterhead, as the officials empowered to
grant permission will usually deny it. Getting Lithuanians to the USA is
considerably easier than getting Russians from Leningrad or Moscow.
Lithuania, is almost considered another country, somewhat separated from
the Soviet Union, in fact, the Lithuanians are now in an Identity Crisis,
and are asking for more independence from the USSR.
The USSR is now relaxing their rules on foreign travel for their citizens,
so inviting good programmers to visit USA institutions is much easier, and will
become even more so in the next few months.
I met a fan of Ted Nelson while visiting the Academy of Sciences in Moscow.
who has been following the Xanadu movement since day 1, and has written an
IBM-PC program I managed to smuggle out of the USSR (I'll explain that part
later). This program is a HyperText style Dictionary program for aiding in
translating English to Russian. It has a rather unique look-up facility, and
was written in Turbo Pascal. The person's name is Arcady Borkowsky, and has
written several specialized word processors, full-text databases, and Natural
Languages understanding HyperText. Arcady was the person who taught Vadim how
to write programs. Vadim is one of the authors of TETRIS, a popular Mac
program now being marketed by Spectra Halobyte. Vadim is currently 17 years
old and is in his Third year at the University. When I visited him, he had
his nose in a copy of Goodman's HYPERCARD book. He knows C, Pascal, and
just about every Microprocessor Assembly Language.
An interview with the ORIGINAL author of TETRIS
================================================
Alexey Pazhithov, who currently works for the Academy of Sciences in Moscow
is the original author of TETRIS. I got the opportunity to meet with him and
learn a little about the origin of the game. Alexey first developed on an
LSI-11, and implemented TETRIS as a TEXT version. Everyone liked it at the
Academy of Sciences. This apparently happened a long time ago. Eventually,
IBM-PC's became available, and Vadim Gerasimov, a 16 year old Hacker who knows
a little English did the IBM implementation. The IBM version wasn't marketed,
but was distributed throughout the Soviet Union, and a few other Eastern Bloc
countries. It was written in Turbo Pascal. The TETRIS program was also used
widely in Hungary. Later on, Alexey met Robert Stein of Andromeda Software,
who worked the deal with Spectra Halobyte. Alexey told me he didn't like the
American version, because they changed it from the original idea. American
Mac programmers implemented it on the Mac to Alexeys specifications (Almost!!
anyway). Although not in the Mac "About box", Arcady Borkowsky also taught
Vadim. I met him 2 days later, and he was very knowledgeable about Object
oriented programming, and gets the latest copy of BYTE magazine. We talked
SERIOUS PROGRAMMING stuff, getting into all the gory details of linked list
data structures, and user-interfaces. Yow!! this guy knows a LOT. Arcady
also is into word processing, full text databases, natural language
understanding, and HyperText. He is a very big fan of Ted Nelson.
When I told Arcady that I was working for the same Company as Ted Nelson,
he opened up and got really excited. I made a note to see what I could do to
get Ted and Arcady together. In fact, Arcady was responsible for me getting
a visit to the Academy of Sciences. He asked me if I wanted to address the
Academy of Sciences people in a lecture at Moscow University. I agreed to
lecture as long as we could agree on the subject matter in advance. So we spent
the rest of the afternoon working on the topics of discussion, which were:
a) Early history of Apple
b) Telecommunication services, Packet Switching services, BIX, Well,
and Compuserve.
c) Viruses and their prevention.
d) The Next computer (What I knew of it).
e) SF/Moscow teleport.
Yes!! They have heard of John Draper, and have read about my past and
history. They get a LOT of books and magazines from the US.
I talked about private bulletin boards and public data systems like
Compuserve, BIX, etc. They didn't appear to know about any of this. All were
VERY interested, and nothing I could contribute regarding my Object Oriented
expertise would impress them, and in fact, they gave me a lot of useful tips
on some problems I've recently ran into before my trip. They get Byte
magazine (2 months late), and pointed out the article about Object oriented
techniques. I also addressed the virus problems, and gave them the same
advice I give to anyone else. I was really surprised at the openness and
their interest in telecommunications. I also talked about the private Bulletin
boards in the USA, and how easy it is to access information. I also turned
them onto the SF/Moscow teleport (Explained later) which they didn't appear to
know about.
The Soviets told me that some people have access to Dialog, an American
data service, and Telenet, and the Source, but the access is highly
restricted to Academic use. The Soviets have to go through a LOT of paperwork
and scrutiny to obtain access to these networks.
Apparently, information is still restricted, for example, the SF/Moscow
teleport was only advertised in the Intourist hotel on Gorkey St. Normal
Soviet citizens are NOT ALLOWED to enter that part of the hotel. SF/Moscow
teleport's offices are about a block from the Arbot. No signs, or anything
visible from the outside can be seen to give a clue as to what's inside.
Inside are Modems, and even a Xerox machine (Another Soviet Taboo). In
fact, a Soviet Policeman was in there inspecting the locks on the door to the
Xerox machine, because ALL Xerox machines must be under lock and key. Ooops I
mean (Soviet Police Approved Locks) and keys.
When I wanted to bring my Soviet friends to meet the SF/Moscow teleport, at
the Intourist Hotel, I had to go through a lot of paperwork to get them access
to the hotel. ALL soviets I brought to meet the Teleport setup were VERY
surprised to find it there. I couldn't believe the look on his eyes when he
saw that little Hitachi lap-top computer brought in from the States.
The Teleport people were VERY EAGER to help my Soviet friends, and promised
to help them get on the service.
Very few Soviets have and use Modems. Contrary to popular belief, they are
NOT ILLEGAL, and you Don't have to have permission from the Organs to use them,
you just have to have a LOT of patience to use them on their phone system.
Some precautions to take while dealing with Soviets
====================================================
If you plan on visiting the Soviet Union, do NOT plan on taking diskettes
containing software OUT OF the Soviet Union. Diskettes are what they call
"Non-Controllable" items, and are subject to confiscation. I managed to slip
ONE IBM disk out of the USSR, by stuffing it inside my dirty underwear. But
a Mac diskette of Cyrillic fonts I obtained from Leningrad didn't make it
through.
And now, more do's and Don'ts:
Do learn Russian, at least learn enough of the Alphabet to read road signs
and street names. Also learn how to pronounce numbers, so that when prices
are quoted, you can understand them. Russian isn't as hard of a language to
learn as Finnish, or the other Scandinavian languages, and really goes a long
ways to developing friends. It takes about 4 days to learn enough sayings to
go to stores and ask for things, such as prices, etc. Other good ways to
learn the language is by watching TV, Listening to the radio, and having
someone help you. Russians like getting questions about the language, as
that gives them the opportunity to be helpful, and shows them that you are
seriously interested in them, and their language.
If you meet ONE group of people, NEVER EVER try and introduce someone from
ONE group to one from ANOTHER group. The Soviet citizen is very sensitive and
has a very deep paranoid feeling about spontaneous meetings with other Soviets.
Soviets think Collectively, and don't really understand individualistic ideas.
So when speaking to them, always think "Collectively" or in those terms.
Because the Soviet citizen doesn't have the concept of Individualism, this
sometimes makes it harder for them to understand private enterprise, and
often leads to communication difficulties.
REMEMBER!! Nothing will ever work out as planned, so if you have a
prescribed plan, agenda, or meeting, more than likely, it will be
invalidated fairly quickly. Always make contingency plans. Things tend to
happen spontaneously, and the very best contacts are met that way.
Especially in trains and subways.
If you plan on staying in the Soviet Union more than 2 weeks, plan on
getting used to the food. It is VERY BAD, starchy, and often very hard to
chew. The food at even the BEST HOTELS can be almost indigestible. If you're
a vegetarian, you can survive, but fresh vegetables are VERY RARE.
Soviet people are incredibly easy to meet, and will usually go out of
their way to make American friends. They are very disciplined, and will
ALWAYS be on time for meetings, and will expect their American friends to show
the same reliability. They like ANYTHING American, so if you plan on
visiting the Soviet union, bring LOTS of "T" shirts, trinkets, pins, and
Frisbees, and Bruce Springstein records. They mostly have cassette tape
players and record players. I saw NO Compact disk players in the USSR, and
learned that VHS tapes use the SECAM video standard which is DIFFERENT than
ours, and don't expect ANY Soviet citizen to have one, they cost around
20,000 rubles. They also use 220 volts 50 Hz power, and that should be a
consideration if you plan to bring your computer. It's OK to bring in a
computer, just remember to Declare it, and take it with you when you leave.
If you plan on leaving it, don't declare it. If it don't show up on
your declaration card when you leave, none's the wiser. When you enter the
Soviet Union, you will be given a Customs Declaration card. In this card,
you Declare ALL your cash in negotiable travelers checks, jewelry, etc. Then,
when you change dollars to rubles at the hotel, or beriosky (Dollar store),
the customs declaration form is marked with that transaction. You must have
this form at all times, as it is your "Documents" and must be surrendered to
the authorities when asked for it. I have NEVER, nor has anyone else EVER
been asked for these papers.
When you LEAVE the Soviet Union, all the money you have is counted again,
and all the rubles you have left is then converted back to dollars (Again at the
.63 rubles/dollar rate). You will also be asked for the receipts you got when
you bought rubles. If you purchase anything on the Black Market, it usually
shows up in excess and unreasonable amounts of rubles. I spent rather lightly
there, despite the fact that I ran out of gifts, and had to go the Dollar
stores to get more. Soviets cannot buy goods at the dollar stores, as they
are not allowed to have dollars or any hard currency. In fact, the WORST
thing you could do to a Soviet cab driver or private car driver is to offer them
Dollars or American money. It makes them feel VERY uncomfortable and uneasy.
REMEMBER!! Dollars are Illegal for them to possess. That's like trying to
offer a joint to a politician in public. You would be surprised to find out
how many naive American tourists there are in Moscow. A lot of Americans
would try and hang out with me, but I would slip down an alley and lose them.
There is a very large "Black market" in the USSR, especially to change
Rubles into dollars. Be very careful about meeting "Traders" or people on the
street willing to change Dollars to Rubles. The price for dollars is about 5
- 7 rubles per dollar, and the "Official" rate is .63 rubles per dollar.
Exchanging dollars to Rubles is Illegal, but it's done all the time, and
usually the authorities will look the other way. NEVER make the exchange on
the street, instead enter a coffee shop and sit down at a table and do the
transaction that way. While I was there, I never had the need for Rubles,
as the Berioski's or Dollar stores had the best goods and much better prices.
However, if I ever wanted to make a phone call or purchase Services, or
Non tangible things, then the Black Market would be a good way to save money.
Phone calls to USA costs 25 rubles for 3 minute call. On black market,
that would be about 3 dollars. Almost worth it.
There are traders EVERYWHERE, all very eager to do Business with
Foreigners. Sometimes they can be a bother. One person in our group became
completely agitated by ONE persistent trader in Tbilisi, Georgia. I have to
admit, he WAS very persistent.
Getting around in Moscow and other Trivia
=========================================
Moscow has a RUSH HOUR and a traffic problem just like every OTHER city I've
been in, but at least the cars can MOVE. Often slowly, as old funky
trucks often break down and block traffic. Pedestrians are NOT allowed to
cross the street. Instead, there are underground passages that cross the
angled streets. Sometimes these go for thousands of feet, and all along the
sides are little stands selling ice cream, and hot dogs. There are LOTS of
little nooks and crannies and little alley ways one can explore, and contrary
to popular belief, escorts are NOT necessary, but it's always nice to have a
local person with you.
Moscow has streets radiating out of the Kremlin and Red Square like spokes
of a wheel. Most outbound and inbound streets are wide and one can go almost
from the Kremlin to the outskirts without stopping as long as it isn't RUSH
HOUR. Rush hour is between 5:30 and 6:30, and between 8:00 and 9:00 in the
morning.
The subway system is nothing short of amazing. For a measly 5 kopeks, one
can go ANYWHERE in Moscow. The only bummer is that the subway system closes
down at midnight. The system also looks like a museum. Huge marble
archways, and golden decorations. It's actually gold leaf. It's very
clean, and VERY VERY CROWDED.
If surface transportation is desired, then stand by the street in a
designated spot, and raise out your arm. Have a map handy to show your
desired destination to the driver. If the driver says "Da" you've got your
ride, if "Nyet" then hail another driver. Usually, all passenger car owners
will stop for people. Cabs will also stop, but often he will deny you a ride
if you're not going HIS way. Costs usually range from 2 to 5 rubles from
either cabs or private cars.
Another interesting thing I remember, is that at night, the cars use ONLY
their parking lights. It's ILLEGAL to use full headlights at night in the
city. Hmmm, and I remembered that Finnish drivers must use their headlights
in the DAY while in the country.
Soviet cars (The Lada) are pretty remarkable. I was in ones that were
only a year old. They ran pretty good, a little poor on the acceleration, and
their mileage is about 35 MPG. One thing amazing is that they started up
IMMEDIATELY in 15 below zero weather, and they didn't even warm them up before
pulling out into the traffic.
About the WORST thing you could do in Moscow as far as a traffic violation
is to block traffic. This means you MUST pull off the street in designated
areas. This often leads to driving 10 miles to go 200 feet. Your best bet
while visiting Moscow is to "Hitch a ride", use Subway, or cabs. Cabs cost
from 1.50 rubles to 5 rubles depending on time.
Gas costs about $1.65 to $2.40 depending on the grade of the stuff, and I
noticed there were long lines at some of the gas stations. Usually, Soviets
will purchase gas from black market dealers who make "House calls" and fill up
your tank for you at half the official gas rates. This is sort of GRAY MARKET
stuff.
Most of the time, while in the streets of Moscow, I was either with
Russians, or alone. I was amazed at the freedom I had. There were no
restrictions to where I could go.
Late at might, there are people EVERYWHERE despite the nasty cold weather.
Hoovering around -5 to -15C, and 3 inches of snow on the ground. I have
absolutely NO idea where they are going, but they are just walking around,
talking to each other in small groups of 3 - 4 people.
Doing Business with the Soviets
===============================
Caution!! The Soviets are very shrewd business men. It's important to be
very firm with them, and to make NO promises or commitments with them, as they
will hold you to your word. They have very little experience in business
ventures, and are thus extremely cautious. When negotiating with them always
impress the fact that YOU are the expert in business, and explain that "That's
NOT the way we do business" if something seems out of kilter.
Currently, the laws state that the Soviets must own 51% of the Joint
Venture, but pressure and the reluctance of American Businessmen to deal with
those terms has forced their government to reconsider the 51% - 49%
relationship, and new laws are being considered to allow the Americans up to
80% of Joint Venture ownerships. Try and work out the relationship in such a
way, as that MOST of the technology, and original ideas belongs on the US
side.
Naturally, the Ruble not being a hard currency is the MAIN stumbling block
for Joint Ventures, and will remain so, until the Soviet Union becomes a
financial force in the Common Market, the Soviets have only their goods, and
services to offer us. This is going to change, as news reports are
indicating that the Soviet Union is considering setting up a Stock Exchange. I
remember reading about this in a British Newspaper while flying back.
Excerpts from a pamphlet provided by the Soviet Government
=========================================================
Following are excerpts of a pamphlet I found on the train between Helsinki
and Leningrad. It was provided by the Soviet Government, and directed towards
American businessmen traveling to the USSR. It's titled:
"Joint ventures: Benefit for all".
I'll summarize it here, and mention the key points.
The Legal Basis of the New Type of Business
-------------------------------------------
Under Soviet Legislation, Joint Ventures are considered "juridical
persons", which means they acquire legal rights and responsibilities, and can
be plaintiffs and defendants in court. The participants in a joint venture are
NOT responsible under its obligations, nor is the Soviet state.
An essential feature of a joint venture is the fact that its property is NOT
divided into shares. The law making instruments concerning joint ventures were
adapted in Jan 1987. Usually, the Soviet Government will take out 30 percent
of the profits in taxes, and leave the rest to be divided up in a way that
benefits the Soviet side. Usually they get 51 percent of the profits, but
according to news reports, this is going to eventually change.
The Decree of the USSR Supreme Soviet provided the necessary legal basis for
the setting up of joint ventures. It establishes the procedure for taxation
of joint ventures and their foreign participants and for supplying joint
ventures with natural resources and specifies who is to settle disputes in which
a joint enterprise is involved. Alongside the legislative measures are the
departmental regulations and guidelines, regulated and controlled by the USSR
Ministry of Finance and Ministry of Foreign trade. Also, the USSR Bank for
Foreign Economic Affairs. These organs usually specify and determine
Instructions on Taxation.
The organ responsible for regulation of Material resources and technical
supplies are the USSR State Committee for Material and Technical Supplies.
The Pamphlet tends to "Beat around the bush" as it tries to explain the
complexities of these so called Joint Ventures. It goes on to talk about all
the organs one must know about to set them up. Usually, the Soviet side has
to deal with all of these "Organs". It goes on to say what one CAN and CANNOT
do. One such rule stipulates that if a joint venture has NOT met its
liability, only the people designated as the "Arbitrators" can decide what to
do about it, and NOT the organs responsible for making the initial rules.
Also, the pamphlet points out that the first 2 years of a Joint Venture is
exempt from the 30 percent tax. A revision of the law now provides that the 2
year tax deferral starts when the Joint venture actually starts to make a
profit, and NOT when the Joint venture is initially set up.
Another guaranteed and privilege is the provision that the equipment,
materials and other property a foreign participant brings into the USSR as his
investments in the joint venture are exempt from customs duties. This actually
means that an American Software company and a Soviet Programmer can work on a
joint venture software product on CURRENT technology equipment. It might be
prudent to make sure the contract covered the use of the computer AFTER the
project.
Foreign participants in a joint venture are guaranteed the transfer abroad
in foreign currency of their share of the profits. And if the joint venture
is liquidated, only after the Soviet Participants get their share, then the
foreign portion is guaranteed the return in foreign currency. There is a 20
percent transfer tax when profits are transferred FROM the USSR to the American
side. This is NOT applied in cases where there is an agreement on the
elimination of double taxation between the USSR and the foreign participants
country of origin.
The Process of Setting up Joint Ventures
========================================
The law makes NO exception for any individual industry. The process of
setting up joint ventures comprises of 3 stages from the Soviet side.
1. Those parties agree on a draft charter, such as objective, goals,
location, and its participants, and the size of the funds available, as well
as the structure, composition, and competence of the management bodies, and
the procedure of decision making, etc.
2. This stage involves securing permission from the competent body, and
making sure the joint venture charter conform to the legislation currently in
force in the USSR.
3. The third and last stage involves the registration of joint ventures.
This is handled by the Ministry of Finance of the USSR. It keeps registers of
various types of joint business ventures. No subjects of civil law are
permitted to enter into contractual relations with a joint venture before it's
entered into the Register.
One must remember, the reason why the Soviet Government is allowing its
people to set up Joint ventures, is because the Soviet Union is really hurting
economically. The government basically wants Americans to bring in convertible
hard currency at .63 rubles/dollar. ALL assets in the joint venture is based
in rubles only, and ONLY at the official .63 rubles/dollar exchange rate. This
is naturally a bad ripoff for the Americans, but the Soviet brochure states
that "This procedure is quite fair", so don't believe EVERYTHING you read from
the Soviet Government.
When setting up programming projects with the Soviets, we also must provide
a provision that the Source code be available on the American side to allow for
modifications and changes, and other contingency. And, as a benefit to the
Soviet programmers, they should be given the latest piece of equipment and
development tools to perform the programming work, if some relationship were to
be set up. ALWAYS staying within the guidelines of the 1979 Technology act,
forbidding certain American Technology from the Soviets. I have inquired about
those restrictions, but no responses yet. I was hoping they would get here
before publishing this, but I can always publish another article including
more details on the rules. One would think that the government would make a
better effort towards educating it's people about the rules, but getting them
has been very frustrating. I encourage other people who know the rules to
write up a short summary in "non-legal" terms. I'm sure American Businesses
would eventually want to know them.
If just ONE Soviet programmer is needed for a project, it might be less
expensive to bring him to the USA for the initial portion of the software
project, to allow him to become familiar with the operations of your project,
purpose, etc.
In either case, communication is vitally necessary, and should be included
in the costs of the venture. There exists only ONE system of Email and
electronic communication with the USSR. And as far as I know, there exists no
other data services that provide Email communications such as UUCP. There are
UNIX machines used at the Academy of Sciences and the University, but not many
people use UNIX in the USSR.
Data Communication to and from the Soviet Union
===============================================
There IS a service that recently became available that provides
communication between SF and Moscow by direct satellite link. This link
connects San Francisco callers DIRECTLY to a computer center in Moscow. Special
accounts have been set up to store and retrieve Electronic Mail and computer
data.
I suspect that this joint venture formed as a result of the relationships
that developed between the people working on the Space Bridges that have been so
popular in the past. The name of this joint venture is SF/Moscow Teleport.
On the Moscow side
==================
There are 2 basic offices. One on the Arbot which is very discreet and has
all the Modem equipment. And another one at the Intourist hotel on Gorkey
Street, a very short distance from the Kremlin. The Intourist location is for
American travelers to send Email by TWX, FAX, or special "Overnight delivery"
for fees ranging from $6 to $25 a page, depending on how fast it has to get
there. They are doing a "brisk business" and are serving the American
tourists communication needs.
Both of these places have only Modems and connect to a Moscow research and
data exchange facility that links computers to their OWN internal network and
provide data store and forward services.
Moscow side pays for the use of the National Center for Automated Data
Exchange of the USSR (VNIIPAS), which has satellite facilities to the States.
While in Moscow, I was told that the Soviets pay in Rubles. Costs were
undergoing revision, but this was what they were when I visited Moscow.
100 Rubles setup
5 rubles/month plus 1.33 rubles/minute connect charge for NON COMMERCIAL
100 rubles/month plus 1.33 rubles/minute connect charge for COMMERCIAL
Also a "Per 1000 characters" charge but nobody knew what it was.
ACCESS:
Through the Soviet institution in Moscow, The Soviet person dials in either
2400 baud or 1200 baud modems. A local number in Moscow is used (In Pulse Dial
mode only).
The Moscow phone system is mostly old step by step switches. It's
incredibly primitive and no-doubt massive, with just about Everybody having a
phone, and 2 - 5 public phones per block. Even individual rooms in MOST of
the Intourist hotels have separate lines to the Phone station. Their phone
system is a hodge podge system of old funky in-band equipment mixed with new
modern digital service, usually between the major cities. An elaborate system
of accounting is sitting between your hotel phone and the rest of Moscow to make
sure you get your phone bill when you leave your favorite Intourist hotel.
Apparently all the lines are run through ducts under the streets, and a lot of
moisture gets in there. It's especially bad when it rains, but while I was
there, everything was frozen.
Lines are incredibly noisy, and in MANY cases, I've had to call the party
back because of bad and noisy connections. Phone calls from the hotels into
Moscow are free, but calls to distant cities are charged to your room, and
are VERY inexpensive. Public phone calls are only 2 kopecks, and 15 kopeks
will get you a 1 - 3 minute call ANYWHERE within the Soviet Union, depending on
distance. When I tried to Access BIX from Moscow, it took us an HOUR to get a
line good enough to use for 1200 baud. Once we got in, the quality was
pretty good using a Robotics Modem. Not knowing if BIX was on telenet, I
tried in vain to access BIX, even if they HAVE telenet access, or even getting
the address. Soviets will have LOTS of problems getting modems. And worst yet,
they will have to get "Error correcting" modems to even GET to the Teleport
system. I suspect that the SF/Moscow teleport people will eventually get data
quality lines for their users through the Ministry of Telecommunications who
maintains and controls the phone system.
On the American side
====================
On the American side, I suspect a local number is set up to connect to
modems and data processing equipment eventually being sent over the satellite to
the Moscow computer.
$100 setup
$200/month for COMMERCIAL use + $15/hr
$75/month for NON-COMMERCIAL Enhanced services + $15/hr
$25/month for NON-COMMERCIAL Basic service + $15/hr
Enhanced service: Includes having the Moscow side making hard copies of
your messages, and getting it to Moscow Clients. Follow up phone calls to
Soviet clients, and other "Cattle prodding" services designed to aid the Soviet
side in getting connected to the Teleport. For instance, you can send a note
to the Moscow side of the teleport asking them to call your client so you can
get a vital piece of information. Same going the other way, Soviet clients
can send short messages to the States.
ACCESS:
A local number in SF will probably connect you to the Soviet computer.
SF/Moscow teleport now has DIRECT satellite service to the Moscow computer.
There is ALSO direct-data services provided by the Teleport, and prices
vary depending on circumstances.
A few comments on Intourist, and the Intourist hotels
=====================================================
Always bring your OWN toilet paper. In combination with the bad food,
and making a lot more appointments with "nature", one doesn't want to use what
the Soviets call Toilet paper. I would rather use the Pravda for my
"Important paper work", instead of using that sandpaper.
Don't expect to find a clock in your room, or the fancy conveniences found
in Western hotel rooms. For instance in Soviet Georgia in Tbilisi, the hotels are
NOT heated until the END OF OCTOBER. The weather in Tbilisi was in the mid
50's, dropping down to 45 F at night.
All the plates, dishes, and table ornaments all look the same, and very
often, the plumbing doesn't work. The rooms have color TV, which I use for
a clock, because the stations display a test pattern which also displays the
time.
If you go to Leningrad, DON'T drink the water. There are a lot of little
nasty bacteria that can cause serious problems. Moscow water is FINE to drink,
and it's better than in Oakland or in NYC.
The Intourist guide WE had was very helpful, informative, and offered us
total freedom. I often neglected to tell my guide where I was going, nor did
she care. One VERY IMPORTANT thing to remember about Intourist, is that it's
a STATE RUN organization, or often referred to as an "Organ". It's not
advisable or required that you volunteer any information such as who you have
met, visited, or what you have seen, to your local Intourist guide.
Intourist guides are NOT required to accompany you EVERYWHERE you go. They are
at your disposal, should you prefer to use them. Costs of their services
vary, depending on what you want to do, and the size of your group.
On ones FIRST visit to the USSR
===============================
GO WITH A TRADE GROUP!!! It's the BEST way to visit the USSR, because the
group leader has been to the USSR many times, knows the procedures, and
spends LOTS of time haggling and begging the Intourist agencies for nice hotel
space, dealing with the visas and arranging the contacts on the Soviet side.
Go with groups of 5 to 15 people, but try and avoid groups catering to more
than 15 people.
The Projects for Planetary peace in Monterey, California. was the group I
went with. They promote citizen diplomacy, and provide very informative
Plenary sessions and classes dedicated towards making one more informative about
the Language, customs, and philosophies of the average Soviet person. If you
plan on going, you will be given an opportunity to fill out a questionnaire so
a Soviet counterpart can be located and contacted. Usually, you will meet
with the Soviet peace committee, who makes the arrangements for you to meet
your counterparts. It's an incredibly efficient system.
The trip starts off with a flight to Helsinki, for a 4 day stay in a plush
country manor (Heikko Manor). During this phase, you will meet your leader,
and the rest of the group. There are 3 meetings a day, intensive training in
learning the NECESSARY Russian, and customs. At the end of the training, a
city tour of Helsinki is planned.
After the 4 day sessions, the group busses to the train station, to board
a Soviet train for the trek across the border into the USSR. Soviet customs
will usually NOT inspect your luggage upon entry, but DON'T depend on it.
Once within the Soviet union, you will meet your Intourist agent who will
escort your group to your hotel. Each trip is different, and usually 3 cities
are visited. All expenses while within the USSR is taken care of. The
average costs are $2500 to $3500 per person, and includes EVERYTHING except
gifts, phone calls, and laundry.
During the process of meeting my counterparts, I was given the opportunity
to visit them again at THEIR expense. All I have to do, is to get to an
Aeroflot city (NYC or Wash DC), and the rest of the expenses are covered. I
will probably not take them up on the offer, because I suspect they would
expect to get something substantial from me, which I could NOT deliver. Before
I would accept an offer like this, I would be very careful and clearly explain
to my Soviet contact what I CAN and CANNOT do.
Please plan on spending a lot of time prior the trip to:
a) Learn Russian, or enough words to count, buy things, and get around.
ESPECIALLY, learn to read and pronounce Russian like "Excuse me", "Thank you",
"Good morning" etc.
b) Study and read about the Soviet Union before visiting, check out
special PBS TV programs, or meet a Soviet American group.
c) Never call them "Russians", as there are MANY "Nations" within the USSR
like Lithuania, Georgia, Ukraine, etc Each one has their own language, and
ALL can speak the National Language of Russian.
I hope this information will be useful for those who are interested in
developing some type of Business relationship with the Soviets.
If there are any points I have left out, Please feel free to Email me, and
I'll be glad to give you the answers.
John D
======
Programmers Network
Email: WELL: crunch
BIX: crunch
UUCP: uunet!acad!well!crunch
Please post this onto other systems like the Arpa Net, BITNET, and
the other networks. And I hope this dissolves a few Myths about
the Soviet Union.
∂16-Nov-88 1330 ginsberg@polya.Stanford.EDU formfeed to meet tomorrow!
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 16 Nov 88 13:30:50 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA28519; Wed, 16 Nov 88 13:22:42 PDT
Date: Wed, 16 Nov 88 13:22:42 PDT
From: Matthew L. Ginsberg <ginsberg@polya.Stanford.EDU>
Message-Id: <8811162122.AA28519@polya.Stanford.EDU>
To: feed@polya.Stanford.EDU
Subject: formfeed to meet tomorrow!
... after the month's hiatus. See you all at 12.00 or 12.15 ...
Matt
∂16-Nov-88 2207 VAL Special seminar on default reasoning
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
Tomorrow (November 17) Irina Gerasimova from the Institute of Philosophy
(Moscow, USSR) will speak on
"Unless"-Norms and Default Reasoning.
Time: 4:15pm. Place: Ventura Hall.
∂17-Nov-88 0140 JMC Expired plan
Your plan has just expired. You might want to make a new one.
Here is the text of the old plan:
I will be in Japan till Nov 15 and Dallas until Nov. 17.
I will mostly be at the Miyako Hotel in Kyoto.
011 81 75 771-7111
∂17-Nov-88 0937 ginsberg@polya.Stanford.EDU Formfeed to meet today!
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Nov 88 09:37:46 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA12878; Thu, 17 Nov 88 09:37:13 PDT
Date: Thu, 17 Nov 88 09:37:13 PDT
From: Matthew L. Ginsberg <ginsberg@polya.Stanford.EDU>
Message-Id: <8811171737.AA12878@polya.Stanford.EDU>
To: feed@polya.Stanford.EDU
Subject: Formfeed to meet today!
Don't forget -- 12.15 in MJH 252. See you then!
Matt
∂17-Nov-88 1313 @Score.Stanford.EDU:ball@polya.Stanford.EDU CSDCF Meeting Discussion Item...Cost Analysis
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 17 Nov 88 13:13:11 PST
Received: from polya.Stanford.EDU by SCORE.STANFORD.EDU with TCP; Thu 17 Nov 88 13:10:20-PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA27650; Thu, 17 Nov 88 13:12:45 PDT
Date: Thu, 17 Nov 88 13:12:45 PDT
From: Jim Ball <ball@polya.Stanford.EDU>
Message-Id: <8811172112.AA27650@polya.Stanford.EDU>
To: Facil@Score
Subject: CSDCF Meeting Discussion Item...Cost Analysis
****Discussion Item for the next Facilities Committee meeting****
1. Are CF's rates correct? CF has been criticized for billing at
rates that are out of proportion to the services performed.
Arguments about what the "correct" rates should be have abounded
on the BBoards. Some, if not most of the arguments are appealing
and seem to make good common sense. However, the findings out-
lined in this report do not support these contentions.
2. The following summerizes my conclusions regarding Computer
Facilities' (CF) workload, cost and rate structure;
2.1. For CF to function effectively it must be directed by
formal goals established by CSD.
2.2. CF's current goals are informal and the product of react-
ing to the demands and perceived needs of CSD.
2.3. CF should not be expected to justify its own existence.
2.4. CF's workload is reasonable heavy due to responding to the
demand for quality service.
2.5. CF's current cost and rate structure is appropriate for
the established goals of the group.
3. Resource rates are not just for hardware
3.1. Depreciation, which reflects the cost associated with
hardware, is only 15% of the total cost of operations.
This percentage is for the total of all systems. For
systems that are older, like Sail, the number is much
lower. Sail's hardware percentage is only 2% of its total
cost. Score's hardware percentage is 25.6% of its total
cost. Rates are established to recover total costs includ-
ing hardware costs.
4. Ratio of CPU to Disk Storage charges
4.1. A key question is, what is the correct ratio between CPU
and Disk charges? On initial examination the answer seems
obvious; relate the charge to the ratio between hardware
cost differences. Unfortunately the reality is that the
hardware costs only account for 15% (on average) of the
total system costs.
4.2. After careful analysis of all the data the ratio is ap-
proximately 67.5% for CPU related costs and 32.5% for Disk
Storage costs. This model works well for Polya and Score
whose rate structure follows very closely to this model.
Sails rates were not as close to the model with the cur-
rent rates divided into a 50/50 structure.
5. Overall Budget Breakdown
Salaries $506,667 51%
Benefits $136,800 14%
Operating Expense $199,000 20%
Depreciation $146,404 15%
Total $988,871 100%
As can be seen from the breakdown listed above, approximately 65%
of the total budget is related to personnel expenses, 35% to
hardware, spare parts, outside repairs, materials and supplies.
System 88-89 budget
Score $252,062
Sail $177,286
Polya $295,713
Jeeves $ 35,040
Labrea $ 32,427
VAX Maint $ 52,771
Printers $ 75,387
Phototype $ 21,652
Network $ 46,533
∂17-Nov-88 1532 MPS
Call Tom Henriksen at 3-4255 - personal matter
∂17-Nov-88 1753 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
THINGS THAT CHANGE BY THEMSELVES
Vladimir Lifschitz
Arkady Rabinov
Stanford University
Friday, November 18, 3:15pm
MJH 301
This talk is about the frame problem, which we understand as the problem of
formalizing the commonsense law of inertia. One interpretation of the law
of inertia is that a fluent doesn't change after an action is performed,
unless the action causes it to take on some value. We argue that, in many
domains, this interpretation is inadequate. First, an action may have effects
that are indirect "ramifications" of the changes that it causes. Second,
some fluents, such as time, change even after an action that is not assumed
to have any causal effects whatsoever, like "wait." We propose a more
flexible formalization of the commonsense law of inertia, that allows us
to describe some examples of these two types.
∂17-Nov-88 1810 MGardner.pa@Xerox.COM AIJ Board Members Mtg
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 17 Nov 88 18:09:58 PST
Received: from Cabernet.ms by ArpaGateway.ms ; 17 NOV 88 17:41:48 PST
Date: 17 Nov 88 17:40 PST
From: Mimi MGardner <MGardner.pa@Xerox.COM>
Subject: AIJ Board Members Mtg
To: Amarel@Rutgers.edu, harryb%cvaxa.sussex.ac.uk@NSS.Cs.Ucl.AC.UK,
Berliner@K.GP.CS.CMU.EDU, bledsoe@cs.utexas.edu, boyer@cs.utexas.edu,
rjb@research.att.com, buchanan@vax.cs.pittsburgh.edu,
bundy@edinburgh.ac.uk, bundy@rutgers.edu, JGC@nl.cs.cmu.edu,
davis@wheaties.ai.mit.edu, deKleer.pa@Xerox.COM, duda@Polya.Stanford.edu,
LErman@TEKNOWLEDGE-VAXC.ARPA, Generserth@sumex-aim.stanford.edu,
grosz@harvard.harvard.edu, Hayes.pa@Xerox.COM,
hinton%ai.toronto.edu@relay.cs.net, Lehnert@CS.UMass.edu,
val@sail.stanford.edu, dwl@cs.duke.edu, jmc@sail.stanford.edu,
McDermott@CS.CMU.EDU, mcdermott%gent.DEC@decwrl.dec.com,
Tom.Mitchell@C.CS.CMU.EDU, adafd%icnucevm.bitnet@ICNUCEVM.CNUCE.CNR.IT,
BMoore@ai.sri.com, nagao%kuee.kyoto-u.junet%utokyo-relay.csnet@Xerox.COM,
Newell@CMU-IDA.ARPA, Nilsson@score.stanford.edu, Judea@CS.UCLA.edu,
rperrault@sri.com, zenon%uwo.cdn%ubc.CSNet@relay.cs.net,
reiter%ai.toronto.edu@RELAY.CS.NET, Sridhara@ai.cel.fmc.com,
Stefik.pa@Xerox.COM, walker@mouton.bellcore.com, Wahlster@seismo.css.gov,
Yorick%nmsu.CSNet@relay.cs.net, Winograd@csli.stanford.edu,
WWoods@harvard.harvard.edu
cc: MGardner.pa@Xerox.COM, Bobrow.pa@Xerox.COM,
jmb%sevax.prg.oxford.ac.uk@nss.Cs.Ucl.AC.UK
Message-ID: <881117-174148-3214@Xerox>
Dear Board Member:
This letter summarizes the meeting of the editorial board that was held in
St. Paul in August, asks your opinion on several proposed additions to the
editorial board, and proposes a new statement of interest for the AI
Journal which includes some changes with respect to areas of interest, sole
publication, manuscript length, and accessibility.
We currently have over 2500 instutional subscriptions, and at the end of
1987 had about the same number of personal subscriptions. The backlog of
papers for standard issues will fill two to three full volumes of the AI
Journal. In 1988 the AI Journal will have published three volumes of three
issues each (about 1000 pages in all). In addition to ordinary papers, we
have in the pipeline two special groups of papers. One is a group of
papers that make up a theme issue. A theme issue publishes together a set
of related articles, with a guest editor's introduction. Two such issues
coming up are "AI and Learning Environments" edited by Clancy and Soloway,
and "Computer Chess" edited by Hans Berliner. These are usually only
slightly larger than an ordinary issue of the journal.
We also have a tradition of booklength special issues. These take a whole
volume of the journal. MIT press simultaneously publishes the issue as a
book. Two such issues will be "Geometric Reasoning" edited by Mundy and
Kapor, and "Machine Learning" edited by Carbonell.
Because of the current full pipeline, and the existence of these special
issues, the board agreed to increasing the number of issues published a
year to 12 (four volumes of three issues). In line with this, North
Holland will raise the personal subscription price to $60 per year. The
board suggested limiting the number of full volume special issues to an
average of one per year.
We also discussed at the meeting the increasing reviewing load on board
members, and the subsequent increase in turnaround time. The board decided
it would be appropriate to change the original policy of the journal that
required all reviewers to be from the board; we will now make much more
extensive use of other reviewers. We will not continue the practice of
having a "Recommended By:" at the head of each article. Board members will
continue to receive some papers to review, and will be especially called on
when difficult decisions come up with regards to some papers.
We will continue to elect to the board people to ensure appropriate
representation for various fields, and whose judgment we want to call on
for critical areas of interest to the journal. Three nominees have been
received. Are there any objections to any of the following: Matt Ginsberg
of Stanford, who has been most helpful in reviewing papers in automated
reasoning and logic; Bill Clancey of the Institute for Research on
Learning, who has helped a great deal with papers in expert systems and
learning environments; and Bob Wielinga, Free University of Amsterdam, who
has been most helpful with papers in natural language and knowledge
representation. We will be glad to accept other nominations.
The board discussed getting AIJ submissions to North Holland in electronic
form. An experiment is currently underway. We also discussed getting AIJ
abstracts in electronic form for the electronic library experiment at CMU,
and received approval of both the board and North Holland.
We appreciate all the help that the editorial board gives us, the editors.
Most often you hear from us when we have something to carp about (where is
that review?). It is the hard work of all of you that has made the journal
the premier publication it is today. Thank you.
Cordially,
Daniel G. Bobrow
Michael Brady
!
Articial Intelligence Journal
1. Statement of Interest
The Artificial Intelligence Journal is published monthly by
North-Holland. It welcomes basic and applied papers describing mature
work involving computational accounts of aspects of intelligence.
Specifically, it welcomes papers on automated reasoning; computational
theories of learning; heuristic search; knowledge representation;
qualitative physics; signal, image, and speech understanding;
robotics; natural language understanding; and software and hardware
architectures for AI. The journal reports results achieved; proposals
for new ways of looking at AI problems must include demonstrations of
effectiveness. From time to time the journal publishes survey articles.
2. Sole publication and the review process
Papers submitted for publication must be original.
Manuscripts are accepted for review on the strict understanding that
the same work is not pending publication, or under review, by another
periodical journal; that it will not appear subsequently in another
periodical journal without the permission of the Artificial
Intelligence Journal; and that its submission for publication is
approved by all of its authors and by the institution where the work
was carried out. The Editors of the Artificial Intelligence Journal
advises reviewers to maintain confidentiality until the review process
is complete.
3. Manuscript length
There is no restriction on the length of submitted
manuscripts. However, authors should note that publication of lengthy
papers, typically greater than fifty pages, is often significantly
delayed, as the length of the paper acts as a disincentive to the
reviewer to undertake the review process. Unedited theses are
acceptable only in exceptional circumstances. Editing a thesis into a
journal article is the author's responsibility not the reviewer's.
4. Accessibility
The AI Journal caters to a broad readership. Papers that are
heavily mathematical in content are welcome but should be preceded
by a less technical introductory section that is accessible to a wide
audience. Papers that are only mathematics, without demonstrated
applicability to Artificial Intelligence problems may be returned.
----- End Forwarded Messages -----
∂17-Nov-88 2250 ramshaw@src.dec.com the history of the term "bag"
Received: from decwrl.dec.com by SAIL.Stanford.EDU with TCP; 17 Nov 88 22:50:25 PST
Received: from jumbo.pa.dec.com by decwrl.dec.com (5.54.5/4.7.34)
for jmc@sail.stanford.edu; id AA10727; Wed, 16 Nov 88 10:29:45 PST
Received: by jumbo.pa.dec.com (5.54.5/4.7.34)
id AA03022; Wed, 16 Nov 88 10:29:34 PST
From: ramshaw@src.dec.com (Lyle Ramshaw)
Message-Id: <8811161829.AA03022@jumbo.pa.dec.com>
Date: 16 Nov 1988 1029-PST (Wednesday)
To: jmc@sail.stanford.edu (John McCarthy)
Cc: ramshaw@src.dec.com (Lyle Ramshaw)
Subject: the history of the term "bag"
John, I am writing a book about splines in which I use the word "bag"
to refer to "a set with multplicities" or "a sequence without order".
I first learned this sense of the word "bag" in one of your courses
at Stanford, but I would like to assign credit properly. Did you
pioneer the use of "bag" in this sense, or do you know who did? (Don
Knuth credits N. G. de Bruijn with inventing the alternative term
"multiset".)
Lyle
∂18-Nov-88 0805 SLOAN@Score.Stanford.EDU Re: reply to message
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Nov 88 08:04:55 PST
Date: Fri 18 Nov 88 08:02:21-PST
From: Yvette Sloan <SLOAN@Score.Stanford.EDU>
Subject: Re: reply to message
To: JMC@Sail.Stanford.EDU
In-Reply-To: <4CDLp@SAIL.Stanford.EDU>
Message-ID: <12447558743.18.SLOAN@Score.Stanford.EDU>
Thanks.
--Yvette
-------
∂18-Nov-88 0813 CN.MCS@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 18 Nov 88 08:13:14 PST
Date: Fri, 18 Nov 88 08:12:22 PST
To: jmc@sail
From: "Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>
John,
As you know I have been having trouble getting materials back from
Igor Rivin. He has indicated to me that his attitude has changed
and that he has been returning material when requested.
However, I gave him until Wednesday to get Complex Systems volume 1
back or I would revoke his borrowing privileges. He never returned
it and now he claims it is lost. Two people are waiting on that
journal plus the cost to replace it is $100.
I would like to get all of the material he has checked out back into
the library and then revoke his borrowing privileges. The only
threats available are to hold his paycheck, which Yvette can do but
it is illegal. Or I can find out where he is going and send a
letter to that institution warning them. Or perhaps you can help me
by exerting pressure on him.
Alternatively, if you can convince me not to revoke his borrowing
privileges, I will consider that.
What do you think?
Rebecca Lasher
3-0864
∂18-Nov-88 1425 GLB oral exam of Gianluigi Bellin.
To: JK@SAIL.Stanford.EDU, CLT@SAIL.Stanford.EDU, JMC@SAIL.Stanford.EDU,
jcm@POLYA.Stanford.EDU, sf@CSLI.Stanford.EDU, der@PSYCH.Stanford.EDU
Subject: oral exam of Gianluigi Bellin.
It has not been possible to find an outside chair for the orals,
scheduled for Tuesday, Nov 22 at 2:30.
On the other hand Prof.D.Rumelhart, Psychology, has kindly offered
to be the chair, but cannot on Tuesday.
Now we need to find a date that would be good for everyone.
Prof.Feferman cannot Dec 5 and Dec.13,14.
∂18-Nov-88 1436 peters@russell.Stanford.EDU CSLI's Industrial Affiliates Program
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Nov 88 14:36:03 PST
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Fri, 18 Nov 88 14:38:20 PST
To: jmc@sail
Subject: CSLI's Industrial Affiliates Program
Cc: ingrid@russell.Stanford.EDU
Date: Fri, 18 Nov 88 14:38:17 PST
From: peters@russell.Stanford.EDU
John,
CSLI now has an Industrial Affiliates Program, which we hope many
companies will join, thereby supporting research at CSLI. I'm writing
to ask if you want to be listed in the program's brochure as a
participating faculty member. I hope you will.
Apart from you there are 18 faculty; the CS Department members include
Nils Nilsson, Yoav Shoham and Terry Winograd. If you want to see a
complete list, Ingrid will send it to you on request.
I hope you'll decide to have yourself listed as one of us. If you do,
please give us a three- or four-line list of your research interests.
Stanley
P.S. Are you just back from accepting the Kyoto prize?
Congratulations.
∂18-Nov-88 2101 @Gang-of-Four.Stanford.EDU:arg@lucid.com new new-qlisp
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 18 Nov 88 21:01:29 PST
Received: from LUCID.COM by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA08629; Fri, 18 Nov 88 21:00:00 PST
Received: from bhopal ([192.9.200.13]) by heavens-gate.lucid.com id AA02817g; Fri, 18 Nov 88 20:59:11 PST
Received: by bhopal id AA11436g; Fri, 18 Nov 88 20:57:58 PST
Date: Fri, 18 Nov 88 20:57:58 PST
From: Ron Goldman <arg@lucid.com>
Message-Id: <8811190457.AA11436@bhopal>
To: qlisp@go4.stanford.edu
Subject: new new-qlisp
A new version of new-qlisp is now available for use. Changes include:
1) THROW now works across processes
2) new Qlisp primitives:
SUSPEND-PROCESS, RESUME-PROCESS, KILL-PROCESS,
CURRENT-PROCESS, GET-PROCESSOR-NUMBER
3) new global variable: *cluster-size*
4) user control of process scheduling
5) various old bugs have been fixed
A new version of /qlisp/qlisp.doc incorporating the text below is also
available.
1) THROW
THROW has been extended so that it is now possible for a child process to
do a throw to a catch tag established by its parent or one of its parent's
parents. In brief all of the processes involved are stopped, all
UNWIND-PROTECT cleanup forms that had been established beneath the CATCH
being thrown to are executed by the process that established them, the
process that set up the CATCH being thrown to continues execution of the
code surrounding the CATCH, and any other processes are killed.
The processes always effected by the THROW are the process executing the
THROW, its parent, and its parent, et cetera, up to the process that set up
the CATCH being thrown to. Processes that these processes have spawned
will generally not be effected, unless the child process was spawned by a
construct like (QLET T ...) where (1) the parent process is waiting for the
child processes to finish, and (2) there was no way for the future being
computed by the child to have been stored away in some more global data
structure. Then it seems safe to kill the child processes.
When a process is created it inherits the chain of catch frames being used
by its parent. During the execution of the child process, a THROW to a
catch frame defined by the parent will result in the child process being
killed and the parent process continuing the processing of the THROW,
interrupting whatever it had been doing. If the parent process exits the
scope of a given catch frame, then it is no longer possible for any child
process to throw to that catch frame. Also any catch frames established
after a child process has been spawned are not part of the child process's
chain of catch frames. If the body of a QLAMBDA process closure does a
THROW, the catch frames of the process that called the QLAMBDA are searched
rather than those in the process that created the QLAMBDA. Note that if a
THROW goes through a qlambda process it is not killed, but will instead
proceed to execute its next set of arguments.
As mentioned when a THROW is executed, any UNWIND-PROTECT cleanup forms
beneath the CATCH being thrown to are executed. It is now an error for any
of these cleanup forms to attempt to do a second THROW to any catch tag
beneath the tag being thrown to. In other words one cannot retract the
target of the initial THROW. It is possible to extend the THROW to a tag
located outside of the initial target catch tag. It is also possible for a
cleanup form to throw to the same tag as the initial THROW, in which case
the value returned by the CATCH will be that of the second THROW. Finally
it is also possible to THROW to a new CATCH that is established during the
execution of the cleanup form. The following example may clarify this:
(catch 'outer
(catch 'middle
(catch 'inner
(unwind-protect
(throw 'middle t) ; start throwing
;; cleanup forms
(catch 'inside-cleanup
(throw 'inside-cleanup t)) ; ok
(cond
((...) (throw 'inner t)) ; no good
((...) (throw 'middle nil)) ; ok - changes value
((...) (throw 'outer t))))))) ; also ok
2) New Qlisp primitives
(KILL-PROCESS) ; kill current process
(KILL-PROCESS p) ; kill process p
(SUSPEND-PROCESS) ; suspend execution of current process
(SUSPEND-PROCESS p) ; suspend execution of process p
(SUSPENDED-PROCESS-P p) ; T if process p is suspended, NIL otherwise
(RESUME-PROCESS p) ; resume execution of process p
(CURRENT-PROCESS) ; returns a pointer to the current process
(GET-PROCESSOR-NUMBER) ; returns the number of the processor process
; is running on
KILL-PROCESS causes the specified process to stop, run any UNWIND-PROTECTS,
and go away. If the process was computing a future then attempts to reference
the value of that future will cause an error. Any processes waiting on the
future's value will wake up and get an error. The argument to KILL-PROCESS
can be a process, a future, or a QLAMBDA function. For a future created by
SPAWN or QLET, the process computing it is killed. If the future had been
created by a call to a QLAMBDA, then either (1) if the qlambda process is
currently computing the future it will stop, run any UNWIND-PROTECT's, and
proceed with the next set of arguments waiting for it, or (2) the set of
arguments for the future will be removed from the qlambda process's queue.
The only way to kill a qlambda process, so that it stops processing any
more arguments is to pass the function itself as the argument to KILL-PROCESS.
Note that if the process being killed is currently waiting for child processes
created with (QLET T ...) to finish, then the children will also be killed.
SUSPEND-PROCESS causes a process to stop running and remove itself from
any queues it may be on (e.g. waiting for a lock). RESUME-PROCESS may be
used to cause a suspended process to continue execution. For the moment
the argument for these functions must be a process. Soon futures and qlambdas
will also be allowed as arguments.
CURRENT-PROCESS returns a pointer to the current process. GET-PROCESSOR-NUMBER
returns the number of the processor that the current process is running on.
3) New global variable
There is a new global variable, *cluster-size*, which is set to the number
of processors available for use by Qlisp. This is the maximum number of
processors that may be used.
4) Process scheduling
There is a new global variable *QL-SCHEDULER-METHOD*, which can be used to
control how processes are scheduled. It initially has a value of :FIFO, which
selects a first in, first out scheduler (the same as the previous Qlisp
scheduler). If its value is set to :LIFO then a last in, first out scheduler
is used. Finally if the value of *QL-SCHEDULER-METHOD* is a function, then
that function will be called when (a) parallel mode is about to be entered, so
user data structures can be setup, (b) a process is to be scheduled for later
execution, (c) when a processor needs a new process to run, and (d) when a
process scheduled for later execution needs to be suspended. In the first
case the function will be called with a single argument, the keyword
:INITIALIZE. At this point any user data structures (e.g. locks) should be
properly (re)initialized. In the second case the function will be called with
two arguments, the first being the keyword :SCHEDULE and the second being a
pointer to the process to schedule. In the third case the function will be
called with a single argument, the keyword :GET-PROCESS and it should return a
process to run or NIL if there is no process available to be run (i.e. the
"run-queue" is empty). In the final case the function will be called with two
arguments, the first being the keyword :REMOVE-PROCESS and the second being a
pointer to the process to unschedule.
There are several restrictions on what can be done in the user scheduler
function. When it is called interrupts are deferred and any process passed
as an argument to it has been locked, so no other process will be able to
modify it. When removing a process, no THROW's should be done. (This includes
RETURN's and RETURN-FROM's that go out of an UNWIND-PROTECT.) Note the need
for locks on user data structures since several processes might call the user
scheduling routine simultaneously.
Here is a simple example implementing a LIFO, N-QUEUE user scheduler:
(defglobalvar *processor-run-queue* (make-array *cluster-size*))
(defun init-user-scheduler ()
(dotimes (i *cluster-size*)
(setf (aref *processor-run-queue* i)
(make-lock :type :spin :value nil)))
(setf *ql-scheduler-method* #'user-scheduler))
(defun user-scheduler (what &optional process)
(case what
(:schedule
(let ((local-run-queue (aref *processor-run-queue*
(get-processor-number))))
(with-lock (local-run-queue)
(push process (lock-value local-run-queue)))))
(:get-process
(dotimes (i *cluster-size*)
(let ((local-run-queue (aref *processor-run-queue*
(mod (+ i (get-processor-number))
*cluster-size*)))
(process-to-run nil))
(with-lock (local-run-queue)
(setq process-to-run (pop (lock-value local-run-queue))))
(unless (null process-to-run)
(return-from user-scheduler process-to-run)))))
(:remove-process
(dotimes (i *cluster-size*)
(let ((local-run-queue (aref *processor-run-queue*
(mod (+ i (get-processor-number))
*cluster-size*)))
(done nil))
(with-lock (local-run-queue)
(when (member process (lock-value local-run-queue))
(setf (lock-value local-run-queue)
(delete process local-run-queue :count 1))
(setq done t)))
(when done
(return-from user-scheduler t)))))
(:initialize
(dotimes (i *cluster-size*)
(release-lock (aref *processor-run-queue* i) :ok-if-not-owner t
:ok-if-not-locked t)
(setf (lock-value (aref *processor-run-queue* i)) nil)))))
(defun user-qemptyp ()
(null (lock-value (aref *processor-run-queue* (get-processor-number)))))
****************************************************************************
Previews of coming attractions:
GC changes:
1) Very soon when a GC needs to take place the system will forcibly (but
politely) interrupt other running processes so the GC can proceed. It
will no longer be necessary to wait until other processors notice that it's
time for a GC.
2) Memory will be zeroed at the end of each GC rather than the current way
of doing it bit by bit as needed.
∂19-Nov-88 1525 ME SAIL
∂19-Nov-88 1446 JMC
Is SAIL or SU-AI the preferred name to give people?
ME - SAIL. For electronic mail or other network connections, it is
SAIL.Stanford.EDU.
∂19-Nov-88 1947 weening@Gang-of-Four.Stanford.EDU Inquirer article
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 19 Nov 88 19:47:34 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02441; Sat, 19 Nov 88 19:46:40 PST
Date: Sat, 19 Nov 88 19:46:40 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811200346.AA02441@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Inquirer article
I've put a copy of the Philadelphia Inquirer article on your desk.
∂20-Nov-88 1701 hoffman@csli.Stanford.EDU about meeting
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 20 Nov 88 17:01:52 PST
Received: by csli.Stanford.EDU (4.0/4.7); Sun, 20 Nov 88 15:14:55 PST
Date: Sun 20 Nov 88 15:14:53-PST
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: about meeting
To: jmc@SAIL.STANFORD.EDU
Message-Id: <596070893.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Professor McCarthy,
Sorry for taking so long to get back to you. Above all else, I do not want to
waste your time, so I have been waiting till I finished reading more of your
papers. However, as I must also attack other problems/readings
simultaneously, my progress has been slow and it would be very useful to
gain your advice and insight on a number of things (both in specific
examples and in scope of project). For example, in terms of specifics, I am
unsure of what exactly Minsky offers as the procedural alternative in the
procedural/declarative controversy or even a solid idea for distinguishing
procedural from declarative and vice-versa, e.g. if a machine has "knowledge
how", then it also has "knowledge that" it has "knowledge how". If the
machine has knowledge that, then must not it also have "knowledge how" to
use its "knowledge that"? Also, in terms of the scope of the project, I have
been finding my original conception far too broad and therefore difficult to
argue in a sophisticated way. So, I need to narrow down to more focus. So, I
will call you Monday morning and see if you have time.
I will bring a copy of the fragmented lecture notes which I have from von
Neumann's archive and I will also bring a copy of a book which I have
stumbled across that you might find interesting.
reid
-------
∂20-Nov-88 2043 ME NIC/DDN registration
∂20-Nov-88 0337 JMC Defense Data Network
Is this something I should be interested in?
ME - Probably. On SAIL, type READ NIC to see why you might want to be
registered.
∂20-Nov-88 2349 BXR Wow!
Congratulations on your richly deserved rich award. All the best!!!!!
∂21-Nov-88 0124 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Bessel, yo is my worry now
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 21 Nov 88 01:23:45 PST
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 336078; Mon 21-Nov-88 04:19:01 EST
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 76707; Mon 21-Nov-88 01:13:44 PST
Date: Mon, 21 Nov 88 01:13 PST
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Bessel, yo is my worry now
To: rwg@RUSSIAN.SPA.Symbolics.COM
cc: math-fun@RUSSIAN.SPA.Symbolics.COM, "ilan@score.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "gasper@nuacc.acns.nwu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <19881110101606.7.RWG@TSUNAMI.SPA.Symbolics.COM>
Message-ID: <19881121091315.9.RWG@TSUNAMI.SPA.Symbolics.COM>
Date: Thu, 10 Nov 88 02:16 PST
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
From: rwg
Date: a couple of days ago
Can anybody give >1 term of an expansion at x = 0 of
J[1+1/x](1/x)/J[1/x](1/x) ?
A physicist at BU had a continued fraction which he thought was
1 - k x↑(1/3) + . . ., but it came out to this Bessel ratio.
Turns out we were both right. With the help of the formulas on p 232
of Watson's (804 page) Treatise on the Bessel Function,
1 1 1/3
J (-) (- -)! (6 X)
1 1 + 1/X X 3
------------------------------------ = ----------- = 1 - --------------- - . . .
1 1 2
2 X + 2 - -------------------------- J (-) (- -)!
1 1/X X 3
4 X + 2 - ----------------
1
6 X + 2 - ------
. . .
(I was (stupidly) unprepared for the branchpoint at 0.)
This was the small X case of what he really wanted:
1
-------------------------------------------------------
X 1 1
C e - Y - - - ----------------------------------------
Y 2 X 1 1
C e - Y - - - -----------------------
Y 3 X 1 1
C e - Y - - - ------
Y . . .
(N + 1) X
N - ---------
==== /===\ 2
\ ! ! e
> ! ! --------------------------------
/ ! ! - K X - K X X
==== K = 1 (1 - e ) Y (e Y - C e )
N>=0
= ---------------------------------------------------- .
(N + 1) X
N - ---------
==== /===\ 2
X \ ! ! e
(C e - Y) > ! ! -----------------------------
/ ! ! - K X - K X
==== K = 1 (1 - e ) Y (e Y - C)
N>=0
Notice the nonobvious insensitivity of the rhs to reciprocating Y.
E.g., for Y = i, the lhs is neat, and the rhs looks imaginary.
Viz. (now with positive numerators),
==== N
\ (- 1)
(C + i) > ---------------------------------------
/ N
==== /===\
N ≥ 0 ! ! 1 K - 1
! ! (1 - ------) (1 - i C (- Z) )
! ! K
1 K = 1 (- Z)
C + -------------------------- = -----------------------------------------------------
1
C Z + -------------------- ==== N
2 1 \ (- 1)
C Z + ------------- > -----------------------------------
3 1 / N
C Z + ------ ==== /===\
. N ≥ 0 ! ! 1 K
. ! ! (1 - ------) (1 - i C (- Z) )
. ! ! K
K = 1 (- Z)
This is easy to test numerically, as both sides converge rapidly for c,z>1 (or
|c|,|z|>>1). It sure seems like the rhs is nontrivially real.
Set z=e↑x and compare with my earlier spazzing: (hope your screen is wide!)
==== - N ((N - 1) Y + X)
\ e
C SINH(X) > --------------------------------------
/ N
==== /===\
N ≥ 0 2 N ! !
C ! ! SINH(K Y) SINH((K - 1) Y + X)
! !
1 K = 1
C SINH(X) + ---------------------------------------- = ------------------------------------------------------
1
C SINH(Y + X) + ------------------------ ==== - N (N Y + X)
1 \ e
C SINH(2 Y + X) + ------ > ----------------------------------
. / N
. ==== /===\
. N ≥ 0 2 N ! !
C ! ! SINH(K Y) SINH(K Y + X)
! !
K = 1
Imagine trying to get Lim (y→0) of the rhs without knowing the left!
Since both of these identities were limiting cases of a more general
result, it should be possible to get the regular c.f. with fibonacci
denoms, e.g. (This sinh one gives you alternating numerator signs.)
∂21-Nov-88 1000 JMC
lunch
∂21-Nov-88 1035 CLT Knuth dinner party
Thursday evening is the time Richard and I usually meet
and we will have already missed three weeks due to Kyoto,
Dallas, and Thanksgiving.
If you will take over Timothy after supper this Wednesday
and on Dec 7 then we could go to Knuths Dec 8. Otherwise
I will have to decline the Knuth event, but you could still go.
∂21-Nov-88 1051 CN.MCS@Forsythe.Stanford.EDU
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 21 Nov 88 10:51:15 PST
Date: Mon, 21 Nov 88 10:51:08 PST
To: jmc@sail
From: "Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>
November 21, 1988
Igor,
As of today I am revoking your privilege to borrow from the Math/CS
Library. Because of your previous borrowing record, our e-mails
last week and the loss of Complex Systems volume 1 I am convinced
that you are not a responsible borrower.
In a visit to your office today I picked up all of the library books
on your desk, bookshelf, and floor. I am aware that you have more
materials out. Please return any library materials you still have
out as soon as possible. You will continue to received recall,
overdue notices and bills from the library and will be expected to
return the remaining material in a timely manner.
Rebecca Lasher
Math/CS Library
cc: John McCarthy
Carolyn Talcott
Pat Simmons
∂21-Nov-88 1157 PHY
∂21-Nov-88 1146 JMC
There will be two of us for Don's dinner December 8.
--
thank you.
∂21-Nov-88 1159 pehoushe@Gang-of-Four.Stanford.EDU Throw Problems
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Nov 88 11:59:15 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA05767; Mon, 21 Nov 88 11:51:50 PST
Date: Mon, 21 Nov 88 11:51:50 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8811211951.AA05767@Gang-of-Four.Stanford.EDU>
To: qlisp@Gang-of-Four.Stanford.EDU
Subject: Throw Problems
Using the following definition of Fibonacci and evaluating (fib 8)
yields a fine display of the new debugger. I know its a stupid
definition of fib, but it should work. It works in serial mode, but
not in parallel.
(defun fib (n)
(if (< n 2) (if (= n 0) (throw 2 1) 1)
(catch n
(qlet t ((a (fib (- n 2)))
(b (fib (- n 1))))
(+ a b)))))
∂21-Nov-88 1204 CLT Vacation Time
To: sloan@SCORE.STANFORD.EDU, pehoushek@GANG-OF-FOUR.STANFORD.EDU
CC: JMC@SAIL.Stanford.EDU, MPS@SAIL.Stanford.EDU
Yes, reinstate the missing vac and sick time for Dan
∂21-Nov-88 1209 JANLEE%VTVM1.BITNET@forsythe.stanford.edu History of Time Sharing
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 21 Nov 88 12:09:04 PST
Received: by Forsythe.Stanford.EDU; Mon, 21 Nov 88 12:09:07 PST
Received: by VTVM1 (Mailer X1.25) id 3141; Mon, 21 Nov 88 14:01:08 EST
Date: Mon, 21 Nov 88 13:29:41 EST
From: JAN <JANLEE%VTVM1.BITNET@Forsythe.Stanford.EDU>
Subject: History of Time Sharing
To: John McCarthy <jmc@sail.stanford.edu>
Many thanks for your response. We were aware of the video tapes
but were told that there is a total of 27 hours which has never been edited
or transcribed!! The best thing we thought was to start again.
I was of course well aware of your contribution to the HOPL materials,
since I was the Administrative Chairman of that conference with Jean
Sammet -- incidently we are considering HOPL II to look at the later
generation of languages, but nothing is solid on that yet. I would however
like to keep on top of the 30th Anniversary of LISP so that we might record
that in Annals if possible.
I would be grateful if you could send me (us) your reminiscences of your
work on time-sharing through the e-mail. That should be a very good starting
place for our work. We have been looking for the January 1, 1959 memo
to Philip Morse and so would also appreciate getting a copy of that one.
We are also looking (and have the MIT archives people looking) for
a copy of the Thansksgiving 1963 memo from Fano to Licklider on
starting project MAC. Apparently this predates the actual proposal
written the following January.
We will keep you informed of our progress, and will look forward to
seeing the above two items.
JAN
∂21-Nov-88 1326 weening@Gang-of-Four.Stanford.EDU Meeting with Alliant
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Nov 88 13:26:47 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA06263; Mon, 21 Nov 88 13:25:45 PST
Date: Mon, 21 Nov 88 13:25:45 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811212125.AA06263@Gang-of-Four.Stanford.EDU>
To: jmc@sail, clt@sail
Subject: Meeting with Alliant
Some people from Alliant would like to come by and get an update on
the status of Qlisp. (Bob Nikora, Ben Passarelli and one other whose
name I forget.) I've tentatively scheduled this for 1:00 p.m. next
Monday; please let me know if you want to come but can't make it at
that time.
∂21-Nov-88 1542 hoffman@csli.Stanford.EDU sorry about this morning
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Nov 88 15:42:16 PST
Received: by csli.Stanford.EDU (4.0/4.7); Mon, 21 Nov 88 15:41:12 PST
Date: Mon 21 Nov 88 15:41:11-PST
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: sorry about this morning
To: jmc@SAIL.STANFORD.EDU
Message-Id: <596158871.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
Jon Barwise sent me a very harsh note on the letter which I sent to the
faculty, and being impressionable youth, I got very discouraged about my
prospects in life, etc. and couldn`t bring myself to discuss important
intellectual things in what would be a very slipshod way. Barwise said
basically that I had gone through the wrong channels and that I was ignorant
in five points.
I`ll try to get in touch tomorrow.
reid
-------
∂21-Nov-88 1611 pullen@vax.darpa.mil [WASPRAY%UMNACVX.BITNET@CUNYVM.CUNY.EDU: final version of history circular]
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Nov 88 16:11:32 PST
Received: from sun30.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA23637; Mon, 21 Nov 88 17:38:13 EST
Posted-Date: Mon 21 Nov 88 17:37:50-EST
Received: by sun30.darpa.mil (5.54/5.51)
id AA17426; Mon, 21 Nov 88 17:37:51 EST
Date: Mon 21 Nov 88 17:37:50-EST
From: Mark Pullen <PULLEN@vax.darpa.mil>
Subject: [WASPRAY%UMNACVX.BITNET@CUNYVM.CUNY.EDU: final version of history circular]
To: ISTO-PI-LIST@vax.darpa.mil
Message-Id: <596155070.0.PULLEN@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
Dear PI,
The following was intended for distribution at the DARPA/ISTO PI Meeting
last week. Due to late receipt it did not get distributed.
We encourage all of you to assist Drs. Aspray and Norberg if they contact
you in the course of their research.
Mark Pullen
P.S. Thanks again for the enthusaistic participation which made the
meeting so successful!
---------------
Posted-Date: Fri, 11 Nov 88 16:53 CST
Received-Date: Fri, 11 Nov 88 17:59:56 EST
Message-Id: <8811112259.AA25613@vax.darpa.mil>
Received: from cunyvm.cuny.edu by vax.darpa.mil (5.54/5.51)
id AA25613; Fri, 11 Nov 88 17:59:56 EST
Received: from UMNACVX.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.1) with BSMTP id 9118; Fri, 11 Nov 88 17:57:39 EDT
Date: Fri, 11 Nov 88 16:53 CST
From: WASPRAY%UMNACVX.BITNET@CUNYVM.CUNY.EDU
Subject: final version of history circular
To: pullen@vax.darpa.MIL
X-Vms-To: IN%"pullen@vax.darpa.MIL",WASPRAY
THE ROLE OF DARPA/IPTO IN THE DEVELOPMENT OF
COMPUTER SCIENCE
A PROJECT OF THE CHARLES BABBAGE INSTITUTE
UNIVERSITY OF MINNESOTA
The phenomenal growth of computing after the second world war is primarily an
American success story, a fact which is frequently attributed to the timely
and generous support of agencies of the federal government. Among the most
prominently cited of these agencies is the Information Processing Techniques
Office (IPTO, now ISTO, the Information Science and Technology Office) of the
Advanced Research Projects Agency (ARPA, now DARPA, the Defense Advanced
Research Projects Agency). Many arguments have been put forth in its favor:
That the research IPTO sponsored has resulted in major advances cutting across
computer science and technology: artificial intelligence, networking and
timesharing, graphics, computer architecture, and VLSI design.
That IPTO support has enabled institutions like Berkeley, Carnegie-Mellon,
MIT, and Stanford to become world leaders in computer research and
education.
That IPTO succeeded far beyond all initial expectations in supporting the
critical mission requirements of the Department of Defense.
That IPTO rojects have spawned massive commercial developments in hardware,
software, and computer services.
And that both directly and indirectly DARPA has contributed to the
maturation of computer science and technology as professional disciplines
and to the computer industry as a major factor in the American economy.
In order to investigate these important and untested claims, we have engaged
in a two-year, three-part historical research project, sponsored by
DARPA/ISTO with administrative assistance from NASA-Ames and the Information
Sciences Institute of the University of Southern California. Our study will
examine the history and influence of IPTO from its beginnings in 1963
through the early 1980s. The final conclusions will be presented in a
report submitted to ISTO.
1. A management history of the DARPA/IPTO Office.
This study will allow us to investigate such questions as the development of
IPTO within DARPA and DoD; the evolution of IPTO's programs; the management
style of IPTO, and changes in it over time; the interactions of IPTO with
its principal investigators and more generally with the academic and
commercial sectors; and the evolution of its general objectives and its
effectiveness at meeting them.
2. Case Studies.
We will undertake detailed examinations of two areas (timesharing and
networking; artificial intelligence) in which IPTO has historically provided
major support. We will investigate the major projects sponsored by IPTO in
these two areas, and evaluate the overall effect of IPTO sponsorship on the
advancement of these research areas.
3. General Conclusions.
We will merge the findings of our management history with those from our two
case studies to reach general conclusions about IPTO and its programs. We
will also set our findings in the context of larger historical issues about
the overall development of computer science and technology, and more
generally about the role of technology in American society.
This is a complex historical task, because IPTO has been involved with many
different individuals at many different institutions. Our work will follow
the current professional standards for historical research and will involve
the use of many different kinds of sources: published literature, technical
reports, archival sources (like minutes and correspondence), interviews, and
site visits.
The apparent impact of IPTO on the development of computer science and
technology render it a worthy topic for careful historical examination.
However, it is our expectation that the study will also have larger
historical importance, adding to our knowledge of the role of government in
the development of cutting-edge technology, and of technology in modern
American life. The study is fundamentally of an historical nature, but it
may also have policy interest. By studying historically a case in which a
government agency has consciously set out to advance the frontiers of
science and technology, we may learn important lessons about the
organization of government research efforts, the role of government in the
stimulation of the economy, and the influence of the policy process on
effective support of frontier research.
Principal Investigators: Dr. William Aspray and Dr. Arthur L. Norberg
Charles Babbage Institute, 117 Pleasant St., SE,
Suite 103, Minneapolis, MN 55455
(612) 624-5050 or CBI@umnacvx.bitnet
November 1988
-------
-------
∂21-Nov-88 1650 scherlis@vax.darpa.mil PI MEETING -- DISCUSSION SUMMARY
Received: from vax.darpa.mil by SAIL.Stanford.EDU with TCP; 21 Nov 88 16:50:16 PST
Received: from sun45.darpa.mil by vax.darpa.mil (5.54/5.51)
id AA23932; Mon, 21 Nov 88 18:38:11 EST
Posted-Date: Mon 21 Nov 88 18:38:57-EST
Received: by sun45.darpa.mil (5.54/5.51)
id AA01757; Mon, 21 Nov 88 18:38:58 EST
Date: Mon 21 Nov 88 18:38:57-EST
From: William L. Scherlis <SCHERLIS@vax.darpa.mil>
Subject: PI MEETING -- DISCUSSION SUMMARY
To: SW-PI@vax.darpa.mil
Cc: ARCH-PI@vax.darpa.mil, DPSYS-PI@vax.darpa.mil, VLSI-PI@vax.darpa.mil,
squires@vax.darpa.mil, toole@vax.darpa.mil, boesch@vax.darpa.mil
Message-Id: <596158737.0.SCHERLIS@VAX.DARPA.MIL>
Mail-System-Version: <SUN-MM(217)+TOPSLIB(128)@VAX.DARPA.MIL>
TO: The Software PIs
CC: Systems PIs
Subject:(1) The PI Meeting.
(2) Engineering and consensus.
(3) The long list of Thursday discussion questions.
(4) Thursday discussion session summaries.
This purpose of this message is to report on the area meeting we held
Thursday afternoon. I hope it will provoke some additional
discussion.
Bill
----------------------------------------------------------------
1. THE PI MEETING. Questions always arise about the costs and
benefits of large meetings such as the PI meeting last week, and these
questions were debated at length both in ISTO and in the community in
the weeks prior to the meeting. Now that the meeting is over, the
general feeling in the systems part of ISTO is that the meeting was
productive and interesting. It is reasonable to expect that we will
have these meetings from time to time, but perhaps not more often than
every three years and not for a full week.
A major conclusion that I drew from the meeting is that we in the
Software community need to get together -- as a group -- at least
annually for two or three days to present and exchange new results, to
promote interactions among the projects, and to consider new
opportunities. Several other DARPA research communities (such as
speech, vision, microsystems, and others) have such regular meetings
and derive benefit from them.
We did not have much time at the meeting to discuss the issues raised
in the long note of several weeks ago, particularly those relating to
engineering and technology transfer. I did, however, receive a number
of responses in the mail, and I expect to send out a summary of the
conclusions sometime soon. (If you have further comments on the
issues raised in that earlier note, please send them to me in the next
10 days, and I'll make use of them in writing the summary.)
(This note is also going to the PIs for the groups that shared our
Thursday afternoon meeting, because it includes summaries of the two
of the three discussion sessions of that afternoon. This note also
corrects some misleading statements from my Friday report.)
----------------------------------------------------------------
2. ENGINEERING AND CONSENSUS. In our combined Thursday afternoon
session, we formulated a long list of questions to consider in the
group discussion sessions that were to follow. We sought questions
that were important for our combined community. From this long list,
a short list was selected of questions on which we felt progress could
be made in a short group discussion. The process provided everybody
with a way of learning what we all felt were the burning questions.
ISTO generally feels that the results indicate that we are progressing
in the directions indicated by Craig Fields in his talk, but that
there is much more to be done. In particular, Craig encouraged us to
make more explicit moves (1) to document and understand our past
successes, and (2) to apply effort to promoting real tech transfer in
the future.
In formulating the long list, the assembled group seemed to focus
strongly on systems issues rather than purely technical issues.
Whatever progress we can make on the systems issues is important. As
Craig said often in his talk (and more often to ISTO people directly!)
we are measured, as a community, by our ability to get our results out
and into practice. Doing this requires (1) figuring out who our
customers are, (2) learning from them what their needs are, and (3)
working with them to reduce the risks they undertake in adopting new
technology.
The people who build real large scale systems have a primary need to
get something working (which means controlling risks), and this
conflicts with the need to achieve function and performance by
incorporating the latest technology (which usually means increasing
risks).
The goal is to reduce risk for our customers. A good way to do this
is, when the opportunity exists, to make technical commitments within
our community -- rather than require our customers to take the risk of
making the technical commitments for us. Most of the major tech
transfer successes have this critical property of a priori consensus
in the research community. Doing this involves (1) working together
to determine where technical and engineering consensus is possible,
given the current state of development; (2) targeting research to lead
to consensus in these areas of opportunity, and (3) making the
compromises to achieve consensus in the these areas.
This is not to say that we must achieve some impossible utopian
harmony of thought and approach, but rather that it should be sought
in areas where it is possible, making the small compromises this
entails. It also does not mean sacrificing the mixed strategy -- but
it does mean at least actively seeking understanding of the substance
of differences in technical opinion in the community. Since our first
customers are usually ourselves, this can start to happen early in the
process as components get built and assembled into larger engineering
prototypes. The question is how to facilitate this consensus
building.
----------------------------------------------------------------
3. THE LONG LIST. The long list of questions we developed Thursday
afternoon are those questions judged by the group to be interesting,
important, AND approachable, even if we judged that we could not make
progress on any or all of them in an hour of discission. I've tried
to recreate here the long list as best as I can reconstruct it from
the short phrases on the slides. The questions came from two sources:
(1) a discussion including Squires, Toole, Boesch, and Scherlis a few
days earlier, and (2) the Thursday afternoon meeting. (I have tried
to avoid a software bias in this rendering, but perhaps it is
inevitable.)
(A) What are better technology transfer models for CAD and for
software tools? Note successes in architecture, systems software, and
microsystems. [It was suggested from the floor that this question
would be better considered as a theme for later discussion.]
(B) Programming Language (1): An increasing number of programming
languages are being used in the community for building research
prototypes. Examples include Common Lisp, C, C++, Ada, Scheme and
variants, ML, Prolog and variants, and many other. Should ISTO work
with the community to develop next generation programming languages
to support our prototype engineering?
(C) Programming Language (2): What do we see in the future for Common
Lisp and Ada? What positive contributions can the research community
make?
(D) Programming Language (3): The emerging parallel computing systems
embody a large number of distinct models of computation. In the
theory community, a relatively small number of abstract models of
computation are used as the basis for theoretical work in developing
and analyzing new algorithms. In the programming language community,
new language approaches are being developed that embody a wide range
of models. Is progress being made in converging to a common set of
abstract computational models that fairly reflects what is going on in
the development of actual parallel systems?
(E) Object Store (1): A small number of data store architecture models
posed at various levels of abstraction have been tabled in the
community. The scale of the problem is such that multiple groups will
need to participate in the development of any realistic system. Are
we in fact moving towards a community consensus in this area
concerning some of the internal interfaces?
(F) Object Store (2): What trust and authentication models are needed
for large scale object management systems?
(G) Object Store (3): The potential now exists, and indeed experiments
are already underway, to develop national-scale shared data stores.
These developments inevitably require some conventionalization of
interfaces if we are to promote interoperability beyond the level of
FTP. What issues will we be facing, and how can we facilitate
experimentation and development of this kind of capability? Does
DARPA have a role here?
(H) As CAD systems in microsystems and manufacturing move to encompass
more support for dealing with function as well as geometry,
similarities begin to develop with software CAD (at least with the
software CAD *problem*). Does it make sense to coalesce some of the
efforts here, if only at the level of engineering infrastructure, or
is the state of development not sufficiently far advanced?
(I) Large scale time-constrained (a.k.a. real-time) systems
predominate defense software development and are beginning to become
more important in commercial applications such as control and robotics.
What can the research community do to develop technology in this area?
Are there standard real-time abstractions, for example?
(J) What can our community do to develop technology in support of the
sort of rich HDTV functionality described in earlier sessions? The
question expands to include other potential high-volume high-tech
products.
(K) What can the community do to accelerate the development of
sharable component technologies? Consensus/validation workshops are
of value, but are there other mechanisms? How can we ensure that
progress is not slowed as a result of the compromises that result?
(L) Strategic Computing (1): The Strategic Computing program has been
successful in producing a number of parallel computing systems that
are now in the commercial marketplace. Progress in systems software
has made these machines more easily accessible for experimentation.
The SC program has also contributed significantly to progress in AI
technology, with some demonstration applications systems being
developed. One of the great challenges we now face is how to better
connect the AI applications with the systems software and
architectures in a way that will permit (1) larger scale systems and
(2) hybrid systems in which AI components are incorporated into more
conventional systems. Could this be facilitated through the
development of SC "Applications Laboratories" for specific domains?
(M) Strategic Computing (2): What are the driving applications for the
TeraOps systems and for the National Data Store? Should
applications-focused work be indertaken?
(N) What value is added (if any) through DARPA (vs. NSF) sponsorship
of research in algorithms, theory, and programming language design?
In particular, (1) do these activities benefit from the scale at which
DARPA-sponsored activities can operate; (2) do they benefit from
interactions with the DARPA community, and (3) does the DARPA
community benefit from having them operate in our midst?
(O) Should we accelerate work in the area of formal specifications?
Where should the emphasis be?
(P) What can we do to develop better systems software and tools for
large scale parallel computing?
(Q) How can we promote simplicity of concept and approach in the
research we do and the systems we develop? [It was decided that this
question raised a theme that should be considered the discussion of
the specific technically focused questions.]
(R) The TeraOps plan (as presented) did not spell out details of
approaches to I/O and to front-end systems support. What are the
issue here, and what is a good set of strawman plans?
The prevailing opinion in ISTO is that this list of questions is
itself an important result. Please send us additions and comments.
----------------------------------------------------------------
4. THE SHORT LIST: DISCUSSION SUMMARIES. Looking over the summaries
of the Thursday afternoon discussion, it is clear that even in the
short afternoon discussions some good progress has been made. I
attempt here to summarize the results, as I understand them from the
material I have.
I had hoped on Friday morning to present this kind of summary and to
use the opportunity to highlight the aforementioned issue of
engineering and consensus, which is so often highlighted to us by
DARPA's customers. Unfortunately, in the effort to highlight this
issue, my talk seemed to have created the impression that there were
no real results of the discussion sessions. This is not the case, and
I'm sorry if I left this imporession. Indeed, some of these
discussions generated enough interest that they may continue over the
net and form the basis for new working groups.
Please contact the session moderators for details. Other ISTO program
managers will send out additional comments on these three areas, and
we hope to develop some discussion in these areas.
These three questions were developed from the Long List above through
a massively parallel communication process involving all present at
the Thursday afternoon session.
-- The first question embraced (B), (C), and (D) above, Programming
Language (1), (2), and (3).
-- The second question embraced (E) and (G) above, Object Store (1)
and (3).
-- The third question embraced (L) and (M) above, Strategic Computing
(1) and (2).
4.1. Programming Language (Al Despain, moderator). The report below
is a verbatim copy of the report submitted by Al.
The charge to the group was to discuss:
1. The fate of Lisp and Ada.
2. The development of a new language for by the DARPA community.
3. Models of computation.
Much animated discussion ensued and some limited consensus emerged:
1. "Lisp and Ada will be with us forever and we should develop
programming environments to accommodate them both."
2. "No development of a new language standard beyond CPL should now
be attempted by the DARPA community."
3. "DARPA should sponsor a language research and development program
that has [as] its goal the understanding of the problems of:
A. Programming in the large,
B. Creating safe software,
C. Covering a wide-spectrum of tasks that span meta-languages,
software programming, and hardware description."
Within this program, research projects should address the issues of:
A. Typing systems.
B. Declarative vs. Imperative languages.
C. Language relationship to proofs, specification, and parallelism.
D. Sets, logic, and transformation representation.
E. Models of computation.
The community should also develop and evaluate several different
experimental language systems.
[Comments from WLS: Al also submitted a set of notes, whch were very
interesting. I should say that the overall results of this discussion
are generally consistent with the approaches we are taking in ISTO.
Concerning number 1: DARPA invests, but only at a moderate level, in
this area. This is not to say that we don't recognize certain trends
concerning these languages. There is significant commercial activity,
of course, and DARPA investment thus focuses on higher-risk technology
issues (such as data interoperability, object base support for
environments, incorporation of formal methods tools in Ada
programming, and explicit process encoding).
Concerning number 2: The issue remains of what languages to use for
larger scale engineered components to be shared in the community. Are
RPC and data interoperability tools (e.g., MIT's Mercury) developing
to the point where we can sustain a heterogeneous approach here? If
so, then the original question becomes less urgent.
Concerning number 3: These are clearly important areas, and DARPA
already invests in most or all of them (if I interpret the
descriptions correctly). Proposals continue to be welcome (see
forthcoming office-wide BAA).]
4.2. Data Store (Carl Hewitt, moderator). Alas, these notes seem not
to have made it back from the meeting with me. Activity in this group
is continuing, however, and it is reasonable to expect that a report
will surface soon. (Carl, you might want to forward the notes on to
use for redistribution.)
4.3. TeraOp Applications (John Hennessy, moderator). The report
below is a verbatim copy of the slides provided.
What is TeraOp goal? (YES: 1, 3)
1. Increase productivity/competitiveness
2. Grand challenge problems
3. Spin-off technology
Big concern whether program really supports 1 and 2.
How can these be ensured?
Does Teraop technology trickle down into competitive US commercial
and military products?
OR is more directed goal required?
[I hope this does justice to the next slide:]
|* the TeraOp |
| \ | Any transfer?
A | \ V
$ | | \
| | \
| ******* \ workstations Can you put $ here to
| ********** \ PCs HDTV ... Dynabook impact Teraop?
|______________\
product volume ->
Goals for TeraOp (Need participation from these communities)
- Intelligent factory
- Dynabook
- Simulated reality
- Simulation for product development
[Comments from WLS: There are important technical questions raised in
this summary. It has been stated that the point of the TeraOps
program is not simply to create a TeraOps device, but to develop
scalable parallel technology that would ENABLE the TeraOps device, and
to use this development to leverage all sorts of results in
microtechnology, systems software, and parallel programming tools.
But there remains the question of how in fact to accomplish this.
This is an issue to explore further, and Squires will have more
comments.]
----------------------------------------------------------------
∂21-Nov-88 2231 weening@Gang-of-Four.Stanford.EDU Soviet access to USENET
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 21 Nov 88 22:31:44 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA08675; Mon, 21 Nov 88 22:30:45 PST
Date: Mon, 21 Nov 88 22:30:45 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811220630.AA08675@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Soviet access to USENET
Well, there is now quite a flame war going on in response to John
Draper's article, on exactly the issue you pointed out -- whether or
not giving USENET access to Soviet citizens would also help the Soviet
government and whether this is important.
∂22-Nov-88 0520 JANLEE%VTVM1.BITNET@forsythe.stanford.edu Quick Reply
Received: from forsythe.stanford.edu by SAIL.Stanford.EDU with TCP; 22 Nov 88 05:20:21 PST
Received: by Forsythe.Stanford.EDU; Tue, 22 Nov 88 05:20:22 PST
Received: by VTVM1 (Mailer X1.25) id 2202; Tue, 22 Nov 88 08:19:06 EST
Date: Tue, 22 Nov 88 08:17:53 EST
From: JAN <JANLEE%VTVM1.BITNET@Forsythe.Stanford.EDU>
Subject: Quick Reply
To: John McCarthy <jmc@sail.stanford.edu>
Just to let you know that I did get your four messages -- two with
your own comments, the 1959 memo and the proposal. I have not looked at
them in detail but wanted to ack. their receipt straight away.
Many thanks
JAN
∂22-Nov-88 0854 beeson@ucscd.UCSC.EDU tetrahedron
Received: from ucscd.UCSC.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 08:54:12 PST
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA02901; Tue, 22 Nov 88 08:57:09 PST
Date: Tue, 22 Nov 88 08:57:09 PST
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8811221657.AA02901@ucscd.UCSC.EDU>
To: jmc@sail.stanford.edu
Subject: tetrahedron
The regular tetrahedron is embeddable in three-space at
(1,0,0),(0,1,0),(0,0,1),(1,1,1). (observe the distance between
any two of these points is sqrt(2).) (pointed out to me by
Brian Peterson).
I still haven't had time to do more on the triangle problem, but
every night I let my computer extend the numerical bound on a possible
k such that a triangle embeddable in 4-space but not in 3-space must
have its tangents rational multiples of sqrt(k). It's up to 1042 as
of this morning.
∂22-Nov-88 0943 CLT nsf proposal
To: JMC, VAL
It needs a section called "results of prior NSF support" which should
include the following data:
--NSF award number, amount and period of support
--Title of project
--Summary of results of the completed work. To facilitate
review this summary should not exceed the equivalent of
four single-spaced pages.
--List of publications acknowledging the NSF award.
--If the proposal is for renewal of a grant, then describe
the relation of the completed work to the proposed work.
[I will get the first two items from Sharon, but you will need to
prepare the rest before proposal can be submitted.]
Also, the final report on the previous grant, 2-DMA480,
is due Nov. 30. Betty says me that NSF probably won't even review
this new proposal until they have received the final report.
∂22-Nov-88 1002 ghh@confidence.Princeton.EDU AI as CS and the scientific epistemology of the common sense world
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 10:02:45 PST
Received: from clarity.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA15396; Tue, 22 Nov 88 13:02:32 EST
Received: by clarity.Princeton.EDU (3.2/1.65)
id AA23539; Tue, 22 Nov 88 13:03:44 EST
Date: Tue, 22 Nov 88 13:03:44 EST
From: ghh@confidence.Princeton.EDU (Gilbert Harman)
Message-Id: <8811221803.AA23539@clarity.Princeton.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 17 Nov 88 2249 PST <hCD3S@SAIL.Stanford.EDU>
Subject: AI as CS and the scientific epistemology of the common sense world
I would appreciate very much your sending me a copy of this
paper. The Ringle book seems to have been lost by the
Princeton University Library. I look forward to your
collection of essays.
Gil
Gilbert Harman
Princeton University Cognitive Science Laboratory
221 Nassau Street, Princeton, NJ 08542
ghh@princeton.edu
HARMAN@PUCC.BITNET
∂22-Nov-88 1045 VAL re: nsf proposal
To: JMC
CC: CLT
[In reply to message from CLT rcvd 22-Nov-88 09:43-PT.]
John,
I'll be glad to help with writing the results section and the final report
on the previous grant, if you tell me what should be in there.
-Vladimir
∂22-Nov-88 1254 VAL Journal of Automated Reasoning
Larry Wos said he'd like to publish papers on nonmonotonic reasoning in JAR.
He invited me to join the editorial board, and I accepted.
∂22-Nov-88 1321 GLB
To: sf@CSLI.STANFORD.EDU, JMC@SAIL.Stanford.EDU, jcm@Polya.Stanford.EDU,
CLT@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU, lacey@CSLI.STANFORD.EDU
David Rumelhart's schedule next week makes orals possible
Wednesday Nov 30 from 3:00 on
Thursday Dec 1 from 4:00 on
Friday Dec 2 any time in the afternoon.
----------
You choose.
Gianluigi
∂22-Nov-88 1349 jonas@polya.Stanford.EDU cs323
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 13:48:55 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA25033; Tue, 22 Nov 88 13:48:54 PDT
Date: Tue, 22 Nov 88 13:48:54 PDT
From: Jonas Karlsson <jonas@polya.Stanford.EDU>
Message-Id: <8811222148.AA25033@polya.Stanford.EDU>
To: jmc@sail
Subject: cs323
i'm a junior majoring in CS, interested in AI. In order to get a
better feel for what AI is like, i'm considering taking your class
'non-monotonic reasoning' next quarter.
However, i'm wary of taking a 300-level course, after all they're
listed under 'primarily for graduate students'.
What previous knowledge is necessary for the class? How 'basic' a
'basic knowledge of logic' is necessary? (i have taken cs157) What
will the work be like? Do you feel the class would be worthwhile for
me to take, or should i wait 'til i'm a graduate student (as the
course guide suggests)?
I am currently enrolled in cs221 (intro to AI) but don't feel that
it's giving me a good representation of the field (because of the way
it is run) and thus want to take some other AI course before deciding
whether i want to continue my studies in that area.
I would much appreciate your answers to my questions and any
other suggestions you may have (such as other classes that would
better suit my present needs).
thank you,
-j
∂22-Nov-88 1351 MPS computer chess
I typed that letter that you are going to add some
words on the subject you will be talking about.
Tony Marsland. It is called aldridge.1
Pat
∂22-Nov-88 1441 GLB
To: sf@CSLI.STANFORD.EDU, JMC@SAIL.Stanford.EDU, jcm@Polya.Stanford.EDU,
CLT@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU, lacey@CSLI.STANFORD.EDU
We are getting closer to a fixed point:
Proposed by David Rumelhart: no day, bad day, better day for:
Wednesday Nov 30 from 3:00 on jmc jcm
Thursday Dec 1 from 4:00 on jcm
Friday Dec 2 any time in the afternoon. clt,jcm
∂22-Nov-88 1455 drb@cscfac
Received: from cscosl.ncsu.edu by SAIL.Stanford.EDU with TCP; 22 Nov 88 14:50:56 PST
Received: from cscfac.ncsu.edu by cscosl.ncsu.edu (5.59/1.00)
id AA00979; Tue, 22 Nov 88 17:44:18-3591
Received: by cscfac.ncsu.edu (1.2/Ultrix2.0-B)
id AA00428; Tue, 22 Nov 88 17:41:28 est
Date: Tue, 22 Nov 88 17:41:28 est
From: drb@cscfac (Dennis R. Bahler)
Message-Id: <8811222241.AA00428@cscfac.ncsu.edu>
To: jmc@sail.stanford.edu
Dear Prof. McCarthy:
I am writing to invite you to be one of four speakers next fall in an
interdisciplinary symposium series entitled
Toward a Science of Mind: Problems and Prospects for the Computational
Approach.
The series is scheduled for September/October, 1989, at North Carolina
State University in Raleigh, NC.
Although we would like to have all four speakers appear within a 3-4 week
period, your appearance could be virtually any time between Sept. 20 and the
end of November, 1989.
It is hoped that you will present an evening lecture suited to a general
audience, followed by a social reception.
Either earlier that day or the next day you are invited to present material
of a more advanced nature to interested faculty and students.
The honorarium for this is $1000 plus all expenses.
Two participants have already agreed to appear. They are Prof. Noam Chomsky
of MIT (Sept. 19) and Prof. Gerald Fodor of CUNY Graduate Center and Rutgers.
We are also planning to invite a speaker in psychology of comparable stature.
The series is cosponsored by the departments of Computer Science, Psychology,
and Philosophy at North Carolina State.
North Carolina State is located in Raleigh, in the Research Triangle area
of North Carolina. Among the universities in the immediate area are
Duke University, the University of North Carolina at Chapel Hill, and
numerous smaller institutions.
In addition, the nearby Research Triangle Park is the site for many large
industrial installations.
The area is served by Raleigh-Durham International Airport.
We are interested in attracting the highest caliber speakers and your name
was first on our concensus list. We hope you can come.
Please do not hesitate to write or call me if you have any questions.
Incidentally, I met you at a symposium on Logic and AI at the University
of Maryland in 1984, though I do not expect you to remember me.
Dennis Bahler
Assistant Professor
(919) 737-3369
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
∂22-Nov-88 1537 weening@Gang-of-Four.Stanford.EDU Qlisp meeting
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 15:36:43 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01010; Tue, 22 Nov 88 15:34:59 PST
Date: Tue, 22 Nov 88 15:34:59 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8811222334.AA01010@Gang-of-Four.Stanford.EDU>
To: qlisp@Gang-of-Four.Stanford.EDU
Subject: Qlisp meeting
There will be a Qlisp meeting Monday, November 28, at 1:00 p.m. in
MJH 301. The purpose of this meeting is to give a status report to
some people from Alliant who will be visiting and who would like to
learn what we're up to.
∂22-Nov-88 1549 air@Gang-of-Four.Stanford.EDU abstract from India
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 15:49:03 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01101; Tue, 22 Nov 88 15:48:00 PST
Date: Tue, 22 Nov 88 15:48:00 PST
From: Arkady Rabinov <air@Gang-of-Four.Stanford.EDU>
Message-Id: <8811222348.AA01101@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: abstract from India
John,
The abstract you gave me does not make much sense to me.
I also found out that Vladimir is going to review the full paper. After peruse
of it I have to conclude that the full paper does not make much sense either.
∂22-Nov-88 1557 MPS PTO
I left at 4:00 today in order to finish my
Thanksgiving shopping. The stores will be a mess
tomorrow afternoon. See you in the am.
Pat
∂22-Nov-88 1601 VAL re: NSF report
To: JMC, CLT
[In reply to message from JMC rcvd 22-Nov-88 14:35-PT.]
report[1,val] is report[f88,jmc] with some expository material added
at the beginning. The grant number still has to be substituted for "xxx".
Please tell me if there is anything else I can do.
∂22-Nov-88 2008 GLB
To: der@PSYCH.Stanford.EDU, sf@CSLI.STANFORD.EDU,
JMC@SAIL.Stanford.EDU, jcm@Polya.Stanford.EDU,
CLT@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU, lacey@CSLI.STANFORD.EDU
Proposed by David Rumelhart: no day, bad day, better day for:
Wednesday Nov 30 from 3:00 on jmc jcm, sf
Thursday Dec 1 from 4:00 on jcm
Friday Dec 2 any time in the afternoon. sf clt,jcm
-------------
∂22-Nov-88 1326 CLT (Carolyn Talcott)
Fri dec 2 is best for me.
Others might be work, but I would have to see if
arrangements can be made.
-------------
∂22-Nov-88 1412 JMC (John McCarthy)
It looks like Nov 30 is not possible for me. The other times are
possible.
-------------
∂22-Nov-88 1428 jcm@ra.stanford.edu (John Mitchell)
To: GLB@SAIL.Stanford.EDU
Tentatively, Wed or Friday is better.
I prefer later in the afternoon.
-------------
∂22-Nov-88 1951 sf@csli.Stanford.EDU (Solomon Feferman)
Wed Nov 30 is best for me; all clear at present, all afternoon.
Thurs Dec 1 possible, but less preferable (I might want to go to the
Math Colloquium). Fri Dec 2 is not possible in the afternoon, I'm
supposed to go to Berkeley.
-------------
GLB: (Gianluigi Bellin)
It looks like the only possible day for everyone is thursday dec 1,
at 4:00 pm. Of course, there may be other suggestions.
However, it seems hard to switch to the following week, because
Prof.Feferman cannot Dec 5 and C.Talcott and S.Feferman cannot on dec 6.
If there are no other suggestions, I'm going ahead and fix it for
Thursday dec 1 at 4 pm.
Thank you
Gianluigi
∂22-Nov-88 2132 hoffman@csli.Stanford.EDU i have printed out
Received: from csli.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 21:32:00 PST
Received: by csli.Stanford.EDU (4.0/4.7); Tue, 22 Nov 88 21:30:52 PST
Date: Tue 22 Nov 88 21:30:51-PST
From: Reid Hoffman <HOFFMAN@CSLI.Stanford.EDU>
Subject: i have printed out
To: jmc@SAIL.STANFORD.EDU
Message-Id: <596266251.0.HOFFMAN@CSLI.Stanford.EDU>
Mail-System-Version: <SUN-MM(242)+TOPSLIB(128)@CSLI.Stanford.EDU>
the last third of the book which was sent to you on disc: it looks like
it has very little to do with AI or Computer Science, but is instead
a political argument for the right. Are you interested in
(1) having the rest printed out?
(2) looking at the last third (arguments about SDI, etc.) and
deciding whether you wish to see the rest?
(3) trashing it?
thanks
reid
-------
∂22-Nov-88 2227 crew@polya.Stanford.EDU new mailing list mtc@polya now exists
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 22:27:24 PST
Received: from LOCALHOST by polya.Stanford.EDU (5.59/25-eef) id AA26218; Tue, 22 Nov 88 22:25:25 PDT
Message-Id: <8811230625.AA26218@polya.Stanford.EDU>
To: mtc@polya.Stanford.EDU
Reply-To: mtc-request@polya.stanford.edu
Subject: new mailing list mtc@polya now exists
Date: Tue, 22 Nov 88 22:25:22 -0800
From: Roger Crew <crew@polya.Stanford.EDU>
This mailing list (<mtc@polya>) is for discussing various MTC-student
logistical concerns (e.g., getting to conferences, faculty searches,
advising/funding situation, courses, the qual, whatever strikes anyone
as being an MTC logistical concern...) and any marginally related issues.
This is basically the union of all of the lists that I've seen Tom use,
all of the students I could find on the LOP and LOGMTC lists, and
various faculty members often associated with MTC.
If you've gotten this message, you're on the list. If you'd prefer not to
be on this list, let me know and I'll take you off. ... and conversely
for if you know of someone that should be on this list and isn't.
Roger (a.k.a <mtc-request@polya>)
the current list:
alur@polya.stanford.edu (Rajeev Alur)
arean@polya.stanford.edu (Luis Arean)
mb@polya.stanford.edu (Marianne Baudinet)
alex@polya.stanford.edu (Alex Bronstein)
rtc@sail.stanford.edu (Ross Casley)
crew@polya.stanford.edu (Roger Crew)
fernando@csli.stanford.edu (Tim Fernando)
galbiati@polya.stanford.edu (Lou Galbiati)
grove@polya.stanford.edu (Adam Grove)
tah@linz.stanford.edu (Tom Henzinger)
bhoward@polya.stanford.edu (Brian Howard)
howard@polya.stanford.edu (Howard Wong-Toi)
vasilis@polya.stanford.edu (Vasilios Kallistros)
katiyar@polya.stanford.edu (Dinesh Katiyar)
lincoln@polya.stanford.edu (Patrick Lincoln)
lowry@coyote.stanford.edu (Mike Lowry)
mcguire@polya.stanford.edu (Hugh McGuire)
nowick@polya.stanford.edu (Steven Nowick)
pieper@geode.stanford.edu (Karen Pieper)
martin@polya.stanford.edu (Martin Rinard)
roach@score.stanford.edu (Kelly Roach)
kar@polya.stanford.edu (Ken Ross)
traugott@polya.stanford.edu (Jonathan Traugott)
weening@gang-of-four.stanford.edu (Joe Weening)
eswolf@polya.stanford.edu (Elizabeth Wolf)
rdz@score.stanford.edu (Ramin Zabih)
zm@sail.stanford.edu (Zohar Manna)
jmc@sail.stanford.edu (John McCarthy)
jcm@ra.stanford.edu (John Mitchell)
pratt@polya.stanford.edu (Vaughan Pratt)
clt@sail.stanford.edu (Carolyn Talcott)
∂22-Nov-88 2251 harnad@Princeton.EDU BBBS Call for Neuroscience Nominations
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 22 Nov 88 22:51:26 PST
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA27495; Wed, 23 Nov 88 01:49:27 EST
Received: by psycho.Princeton.EDU (3.2/1.65)
id AA03972; Wed, 23 Nov 88 01:32:42 EST
Date: Wed, 23 Nov 88 01:32:42 EST
From: harnad@Princeton.EDU (Stevan Harnad)
Message-Id: <8811230632.AA03972@psycho.Princeton.EDU>
To: srh@elbereth.rutgers.edu
Subject: BBBS Call for Neuroscience Nominations
To: BBS Associates
In the ebb and flow of submitted material across the BBS seascape, the
tide occasionally happens to favor one of the B's (behavioral/cognitive)
or the other (brain/biological).
Right now, a temporary preponderance of the former over the latter
seems to have accumulated, and although time would no doubt remedy
this, I thought I'd speed things up by intervening and canvassing the
Associateship for burning issues in neuroscience that they would like
to see accorded open peer commentary. Please let me know the names and
addresses of authors the editorial office should approach to encourage
them to prepare a target artcle, and on what topic. (Of course
nominations of non-neuroscience topics are welcome too.)
Stevan Harnad
∂23-Nov-88 0010 rdz@Gang-of-Four.Stanford.EDU What I'm up to
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 00:10:30 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01983; Wed, 23 Nov 88 00:09:17 PST
Date: Wed, 23 Nov 88 00:09:17 PST
From: Ramin Zabih <rdz@Gang-of-Four.Stanford.EDU>
Message-Id: <8811230809.AA01983@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: What I'm up to
I haven't run into you for a while, so I figured I should send you a
brief note about what I've been doing. Basically, I've decided that I
need to go and learn some more mathematical logic. As you've probably
noticed, it's an area that I don't know all that much about, and I
want to rectify this situation.
I've been taking Devlin's course on Set Theory this quarter (we're
currently doing the Constructible Hierarchy). I plan to take the
standard first-order logic course in the Winter. I've also been
trying to finish up a few papers on various search-related topics,
just to get them off my chest.
Anyway, that's what I've been doing of late.
Ramin
∂23-Nov-88 0442 @CUNYVM.CUNY.EDU:GOLUMBIC@ISRAEARN.BITNET
Received: from CUNYVM.CUNY.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 04:42:13 PST
Received: from ISRAEARN.BITNET by CUNYVM.CUNY.EDU (IBM VM SMTP R1.1) with BSMTP id 5002; Wed, 23 Nov 88 07:07:37 EDT
Date: Wed, 23 Nov 88 14:10:41 IDT
To: jmc@sail.stanford.EDU
From: GOLUMBIC%ISRAEARN.BITNET@CUNYVM.CUNY.EDU
Comment: CROSSNET mail via SMTP@INTERBIT
Date: 23 November 88, 14:04:00 IDT
From: Martin Charles Golumbic 972 4 296282 GOLUMBIC at ISRAEARN
To: JMC at SAIL.STANFORD
Dear Professor McCarthy,
I am writing to you as the chairman of a Symposium on the Foundations
of Artificial Intelligence which we are organizing for June 1989 in
Israel to be sponsored by the Research Institute for the Mathematical
Sciences at Bar-Ilan University. The Symposium will be international
in scope, with invited one hour lectures by several leading
researchers from Israel and abroad. Although we have in mind a small
meeting of 30 to 40 researchers, attendance will be open.
On behalf of the program committee, I would like to invite you to be
one of the Symposium guest speakers. Formally, you would be
considered a Bar-Ilan Distinguished Visiting Lecturer for that
week. All participants are asked to submit a full length paper
(research or expository) shortly after the conclusion of the Symposium
which will be refereed according the usual standard of the best
professional journals and published in a separate, special issue of
the "Annals of Mathematics and Artificial Intelligence" as a
permanent record of the Symposium.
The following paragraph will be included in the symposium announcement
which I plan to issue around the end of the month.
"The Bar-Ilan Symposium on the Foundations of Artificial
Intelligence is intended to become a bi-annual event which will
focus on a range of topics of concern to the scholars applying
quantitative, combinatorial, logical, algebraic and algorithmic methods
to areas as diverse as decision support, automatic deduction,
knowledge-based systems, machine learning, computer vision, and robotics.
These may include applied logicians, algorithms & complexity
theorists, and applications specialists using mathematical methods.
By sponsoring such symposia, we anticipate influencing the growth
potential of new areas of applied mathematics and computational theory
generated by this cross-fertilization."
The exact date of the Symposium will be decided in the next two weeks,
but the most likely date is June 19-21. If you accept, travel support
and local expenses during the symposium will be available.
In addition, guest speakers are encouraged to nominate additional
talks related to their lecture to be given by other colleagues. We
would like to have two such "focus areas" and those currently being
recommended are (a) applied logic in AI and (b) algorithmic
foundations of AI.
We hope you will accept our invitation. Your participation will
greatly enhance our symposium. Please let us know your decision as
soon as possible, any time conflict you have with the proposed dates,
and what level of travel support you would require. We would be
delighted to hear your suggestions. I am sure we can also arrange a
talk here at the Scientific Center in Haifa.
Please accept my hearty best wishes.
Marty Golumbic
Israeli Program Committee
Yaacov Choueka (Bar-Ilan)
Rina Dechter (Technion)
Martin Golumbic (IBM Israel)
David Harel (Weizmann)
Daniel Lehmann (Hebrew Univ.)
Micha Sharir (Tel Aviv and Courant)
Jonathan Stavi (Bar-Ilan)
∂23-Nov-88 0647 harnad@Princeton.EDU re: BBBS Call for Neuroscience Nominations
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 06:47:52 PST
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA10739; Wed, 23 Nov 88 09:47:40 EST
Received: by psycho.Princeton.EDU (3.2/1.65)
id AA04367; Wed, 23 Nov 88 09:48:54 EST
Date: Wed, 23 Nov 88 09:48:54 EST
From: harnad@Princeton.EDU (Stevan Harnad)
Message-Id: <8811231448.AA04367@psycho.Princeton.EDU>
To: JMC@SAIL.Stanford.EDU
Subject: re: BBBS Call for Neuroscience Nominations
John, your AI/NI topic is certainly within the BBS spectrum and I
encourage you to prepare such a target article. -- Cheers, Stevan
∂23-Nov-88 0825 tom@polya.Stanford.EDU Dover
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 08:25:42 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA15243; Wed, 23 Nov 88 08:25:39 PDT
Date: Wed, 23 Nov 88 08:25:39 PDT
From: Tom Dienstbier <tom@polya.Stanford.EDU>
Message-Id: <8811231625.AA15243@polya.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 22 Nov 88 1947 PST <lEaIW@SAIL.Stanford.EDU>
Subject: Dover
yep your right..I sent the message before it was ready to send,,it was an opps.
tom
∂23-Nov-88 0900 JMC
smog
∂23-Nov-88 0958 VAL NSF proposal
The following is from: AIList Digest, Wednesday, 23 Nov 1988, Volume 8 : Issue 130.
Should this program be referred to in the proposal, or will it go there
automatically?
Date: Tue, 15 Nov 88 11:26:05 -0500
From: "Henry J. Hamburger" <hhamburg@note.nsf.gov>
Subject: NSF Program in Knowledge Models and Cognitive Systems
NATIONAL SCIENCE FOUNDATION
---------------------------
PROGRAM in
----------
KNOWLEDGE MODELS and COGNITIVE SYSTEMS
--------------------------------------
Knowledge Models and Cognitive Systems is a relatively new name
at NSF, but the Program has significant continuity with earlier
related programs. This holds for its scientific subject matter
and also with regard to its researchers, who come principally
from computer science and the cognitive sciences, each of these
emphatically including important parts of artificial intelligence.
Many such individuals are also interested in areas supported by
other NSF programs, especially in this division -- the Division
of Information, Robotics and Intelligent Systems (IRIS) -- and in
the Division of Behavioral and Neural Sciences.
This unofficial message has two parts. The first is a top-down
description of the major areas of current Program support. There
follows a list of some particular topics in which there is strong
current activity in the Program and/or perceived future
opportunity. Anyone needing further information can contact the
Program Director, Henry Hamburger, who is also the sender of this
item. Please use e-mail if you can: hhamburg@b.nsf.gov or else
phone: 202-357-9569. To get a copy of the Summary of Awards for
this division (IRIS), call 202-357-9572
Many of you will be hearing from me with requests to review
proposals. To be sure they are of interest to you, feel free to
send me a list of topics or subfields.
MAJOR AREAS of CURRENT SUPPORT
------------------------------
The Program in Knowledge Models and Cognitive Systems supports
research fundamental to the general understanding of knowledge
and cognition, whether in humans, computers or, in principle,
other entities. Major areas currently receiving support include
(i) formal models of knowledge and information, (ii) natural
language processing and (iii) cognitive systems. Each of these
areas is described and subcategorized below.
Applicants do not classify their proposals in any official way.
Indeed their work may be relevant to two or all three of the
categories (or conceivably to none of them). In particular, it
is recognized that language is intertwined with (or part of)
cognition and that formality is a matter of degree. For work
that falls only partly within the program, the program director
may conduct the evaluation jointly with another program, within
or outside the division. Descriptions of the three areas follow.
FORMAL MODELS of KNOWLEDGE and INFORMATION:
-------------------------------------------
Recent work supported under the category Formal Models of
Knowledge and Information divides into formal models of three
things: (i) knowledge, (ii) information, and (iii) imperfections
in the two. In each case, the models may encompass both
representation and manipulation. For example, formal models of
both knowledge representation and inference are part of the
knowledge area.
The distinction between knowledge and information is that a piece
of knowledge tends to be more structured and/or comprehensive
than a piece of information. Imperfections may include
uncertainty, vagueness, incompleteness and abductive rules. Many
investigations contribute to two or all three categories, yet
emphasize one.
COGNITIVE SYSTEMS
-----------------
Four recognized areas currently receive support within Cognitive
Systems: (i) knowledge representation and inference, (ii)
highly parallel approaches, (iii) machine learning, and (iv)
computational characterization of human cognition.
The first area is characterized by symbolic representations and a
high degree of structure imposed by the programmer, in an attempt
to represent complex entities and carry out complex tasks
involving planning and reasoning. The second area may have
similar long-term goals but takes a very different approach. It
includes studies based on a high degree of parallelism among
relatively simple processing units connected according to various
patterns. The third area, machine learning, has emerged as a
distinct area of study, though the choice between symbolic and
connectionist approaches is clearly relevant. In all of the
first three areas, the research may be informed to a greater or
lesser degree by scientific knowledge of the nature of high-
level human cognition. Characterizing such knowledge in
computational form is the objective of the fourth area.
NATURAL LANGUAGE PROCESSING
---------------------------
Recent work supported under the category Natural Language
Processing is in three overlapping areas: (i) computational
aspects of syntax, semantics and the lexicon, (ii) discourse,
dialog and generation, and (iii) systems issues. The distinction
between the first two often involves such intersentential
concerns as topic, plan, and situation. Systems issues include
the interaction and unified treatment of various kinds of
modules.
TOPICS of STRONG CURRENT ACTIVITY and
-------------------------------------
OPPORTUNITY for FUTURE RESEARCH
-------------------------------
Comments on this list are welcome. It has no official status,
is subject to change, and, most important, is intended to be
suggestive, not prescriptive. The astute reader will notice that
many of these topics transcend the neat categorization above.
Reasoning and planning in the face of
imperfect information and a changing world
- reasoning about reasoning itself: the time
and resources taken, and the consequences
- use and formal understanding of
temporal and nonmonotonic logic
- integration of numerical and symbolic approaches
to uncertainty, imprecision and justification
- multi-agent planning, reasoning,
communication and coordination
Interplay of human and computational languages
- commonalities in the semantic formalisms
for human and computer languages
- extending knowledge representation systems to
support formal principles of human language
- principles of extended dialog between humans
and complex software systems, including
those of the new computational sciences
Machine Learning of Classification,
Problem-Solving and Scientific Laws
- formal analysis of what features and parameter
settings of both method and domain are
responsible for successes.
- reconciling and combining the benefits of
connectionist, genetic and symbolic approaches
- evaluating the relevance to learning of AI
tools: planning, search, and learning itself
∂23-Nov-88 1055 scales@polya.Stanford.EDU qlisp
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 10:55:22 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA26947; Wed, 23 Nov 88 10:55:19 PDT
Date: Wed, 23 Nov 88 10:55:19 PDT
From: Daniel J. Scales <scales@polya.Stanford.EDU>
Message-Id: <8811231855.AA26947@polya.Stanford.EDU>
To: jmc@sail
Subject: qlisp
Professor McCarthy:
I sent you a message before about possibly joining the Qlisp project.
I have already talked to Joe Weening and read about Qlisp, and I would
like to meet with you. Do you have time available next week? My best
times are Monday, Wednesday, Friday before 11am or after 2pm. I will
try to be at the Qlisp meeting on Monday, but I have a conflict that I
have to reschedule, so I might not be there.
Dan Scales
∂23-Nov-88 1152 drb@cscfac.ncsu.edu Re: reply to message
Received: from cscosl.ncsu.edu by SAIL.Stanford.EDU with TCP; 23 Nov 88 11:50:51 PST
Received: from cscfac.ncsu.edu by cscosl.ncsu.edu (5.59/1.00)
id AA05286; Wed, 23 Nov 88 14:49:55-3591
Received: by cscfac.ncsu.edu (1.2/Ultrix2.0-B)
id AA00860; Wed, 23 Nov 88 14:47:04 est
Date: Wed, 23 Nov 88 14:47:04 est
From: drb@cscfac.ncsu.edu (Dennis R. Bahler)
Message-Id: <8811231947.AA00860@cscfac.ncsu.edu>
To: JMC@SAIL.Stanford.EDU
Subject: Re: reply to message
Dear Prof. McCarthy:
I am pleased that you express interest in our cognitive science symposium.
As this is a lecture/seminar sort of affair,
I don't believe a written paper is even expected of the speakers.
You would of course be welcome to bring along whatever reprints, tech reports,
etc. you wish, but we are not asking that anything be specifically crafted for
this event.
I will be back in touch after Thanksgiving to see if we can/should set a
tentative date.
Dennis Bahler
(919) 737-3369
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
∂23-Nov-88 1157 VAL re: NSF report
To: JMC, CLT
[In reply to message from JMC rcvd 22-Nov-88 14:35-PT.]
The proposal apparently will be mailed today. The final report (almost
identical to the previous support section of the proposal) is report.tex[1,val];
I left a hard copy on John's terminal.
∂23-Nov-88 1240 CLT unrestricted funds
I believe the 57k is a debt not a credit
∂23-Nov-88 1245 peters@russell.Stanford.EDU Re: [peters: Re: [peters: CSLI-Japan collaborations] ]
Received: from russell.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 12:45:29 PST
Received: from localhost by russell.Stanford.EDU (4.0/4.7); Wed, 23 Nov 88 12:46:26 PST
To: masahiko@nuesun.ntt.jp (Masahiko Sato)
Cc: jmc@sail
Subject: Re: [peters: Re: [peters: CSLI-Japan collaborations] ]
In-Reply-To: Your message of Tue, 22 Nov 88 14:12:12 O.
<8811220512.AA18098@MECL.NTT.jp>
Date: Wed, 23 Nov 88 12:46:23 PST
From: peters@russell.Stanford.EDU
Dear Masahiko,
CSLI will be delighted to have your student Yukiyoshi Kameyama visit
from 27 November through 16 December. We can provide him with office
space and intellectual interaction as well.
Stanley
∂23-Nov-88 1415 MPS Partytime
There is a Thanksgiving party for the secretaries
today at 2:15. After that, Betty says we can leave
for the day. Gotta get that turkey cookin.
You and your family have a happy Thanksgiving. See
you on Monday.
Pat
∂23-Nov-88 1441 alex@jessica.Stanford.EDU TAing your class on NonMonotonic Logic Winter Quarter
Received: from jessica.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 14:40:10 PST
Received: by jessica.Stanford.EDU; Wed, 23 Nov 88 14:40:24 PST
Date: Wed, 23 Nov 1988 14:38:52 PST
From: Alex Bronstein <alex@jessica.stanford.edu>
To: jmc@sail.Stanford.EDU
Cc: alex@jessica.Stanford.EDU
Subject: TAing your class on NonMonotonic Logic Winter Quarter
Message-Id: <CMM.0.88.596327932.alex@Jessica.Stanford.EDU>
Prof. McCarthy,
I would like to be a TA for your class next quarter. I'm working in
more "classical" logic for my thesis, but back when I had more time for
"breadth" (before my Quals) I had taken a few more "AIish" classes so I
believe I could study it to whatever level of proficiency you require by next
January.
If you think my TAing your class is possible, then we could meet
in person at your convenience (my schedule is fairly open).
Thank you,
Alex Bronstein
(one of your co-advisees)
∂23-Nov-88 1508 bowers@Popserver.Stanford.Edu Winter Schedule Changes
Received: from sierra.STANFORD.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 14:56:33 PST
Received: from bowers@Popserver.Stanford.Edu (36.10.0.88) by sierra.STANFORD.EDU (3.2/4.7); Wed, 23 Nov 88 14:47:30 PST
Date: Wed, 23 Nov 88 14:47:30 PST
From: bowers@Popserver.Stanford.Edu
To: ag@amadeus, BBL@star, binford@whitney, bracewell@star, daniel@mojave,
drmac@sierra, ejm@sierra, feigenbaum@sumex, ferziger@score,
franklin@isl, genesereth@sumex, goodman@isl, gray@isl, guibas@navajo,
jmc@sail, kroo@ames-aero, latombe@whitney, linvill@sierra, M@sierra,
mitchell@score, moin@score, nanni@mojave, nix@sierra, pantell@sierra,
plummer@sierra, pmbanks@star, reid@glacier, Reynolds@score, rwf@sail,
spicer@sierra, VVA@isl, wiederhold@sumex, wooley@presto,
ak.cfc@forsythe, cannon@sierra, drmac@sierra, Franklin@ISL,
goodman@isl, hf.emf@forsythe, Jones@score, Kruger@sierra,
luenberger@sierra, marx@sierra, na.bmm@forsythe, nilsson@score,
shah@sierra, shankle@sierra, stager@score, amy@popserver,
cschultz@popserver, d1.h13@forsythe, d1.h57@forsythe, DCT@popserver,
jeb@popserver, m.marine@hamlet, na.adp@forsythe, na.aha@forsythe,
na.geo@forsythe, na.kin@forsythe, na.lin@forsythe, na.lth@forsythe,
na.stu@forsythe, na.umd@forsythe, na.vsk@forsythe
Cc:
Subject: Winter Schedule Changes
Next week we will be distributing the "SITN Program Information and Course
Offerings" catalog, which announces SITN's winter schedule of televised
classes. If you are aware of any changes to this schedule, please notify me
so that we can attempt to accommodate those changes in our broadcast schedule,
and notify our company students of any changes. As of Nov. 23, these are the
schedule changes I have received:
Cancelled Classes: EE 334, CS 500, CS 193E, and EE 246
Added Class: AA 244B Classtime: MWF 9:00-9:50 Sk Aud
Broadcast: MWF 5:45-6:35 Ch. E4
Problem Session: CS 106A will be M 8:00-8:50 in Terman 156 on Ch. E3.
Other: E 207 will have a required on-campus lab. HCP students may come on
campus for it, or a similar lab may be set up at the company site by
contacting the TA for the course (as of yet unassigned). If NCO students
would like to take the course for credit, they will have to wait until the
first day of classes to determine whether they wifirst day of classes to dete
the lab.
Have a great Thanksgiving!
Pam Bowers,
SITN Broadcast Operations Coordinator
(725-3003)
∂23-Nov-88 1727 ILAN@Score.Stanford.EDU [Ilan Vardi <ILAN@Score.Stanford.EDU>: Thanksgiving Theorem]
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 23 Nov 88 17:27:33 PST
Date: Wed 23 Nov 88 17:24:47-PST
From: Ilan Vardi <ILAN@Score.Stanford.EDU>
Subject: [Ilan Vardi <ILAN@Score.Stanford.EDU>: Thanksgiving Theorem]
To: jmc@Sail.Stanford.EDU
Message-ID: <12448971849.16.ILAN@Score.Stanford.EDU>
Mail-From: ILAN created at 23-Nov-88 17:16:59
Date: Wed 23 Nov 88 17:16:59-PST
From: Ilan Vardi <ILAN@Score.Stanford.EDU>
Subject: Thanksgiving Theorem
To: dke@Sail.Stanford.EDU
cc: rivin@Gang-of-Four.Stanford.EDU, ilan@Score.Stanford.EDU,
ilan@Gang-of-Four.Stanford.EDU
Message-ID: <12448970431.16.ILAN@Score.Stanford.EDU>
Igor Rivin and myself have been looking at following question:
How many ways can you put n nonattacking queens on an nxn chessboard?
Let Q(n) be the number of ways. As you no doubt are aware, Q(n)>0
if n>3. The question is to find either a formula or an asymptotic estimate.
It seems clear to us that no formula exists so we have been looking at
bounds. As far as we know the best previous bound was Q(p)> p(p-3) if
p>3 is a prime (by Lucas). Our first (proved) result is
(*) Q(n)> exp[ cn] if n= 5*q, (q,30)=1, c= (1/5) log 10
But in a couple of days it is likely that we will be able to get an
exponential lower bound for all numbers.
The truth seems to be that
log Q(n) \sim (1/3) n log n.
Here is a heuristic argument: Put down a queen, this has n choices.
The queen kills three squares in the next column and so forth.
So you might expect n(n-3)(n-6)... solutions on average. It seems
like this argument can be refined to get an upper bound.
The amazing thing is that the proof of the lower bound is incredibly
elementary. Here it is:
We look at toroidal solutions which, as you know, exist if and
only if (n,6)=1. Let p,q be two odd numbers with (p,6)=(q,6)=(p,q)=1.
Now assume that (i,f(i)) is a pxp toroidal solution and (j,g(j))
a qxq toroidal solution and let h:{0,...,p-1}=>{0,...,q-1} be
an arbitrary function. Then it is easy to check that
(pi+qj, p[ g(i)+ h(j)]+ q f(j))
is always a toroidal pqxpq solution. This implies that there are
at least q↑p solutions for a pqxpq board. This is how we got (*).
This method can be improved since g(j) above does not have to be a
toroidal solution (so (p,6)=1 is not necessary) and if g_1,...,g_{Q(q)}
are all Q(q) solutions for the qxq, then for any function
H:{0,...,p-1}=>{1,...,Q(q)} you get a solution
(pi+qj, p [ g_{H(j)}(i)] + q f(j)).
It looks like this method can be refined to get an exponential lower
bound for all numbers. The bad case of primes of the form 2↑a*3↑b+1
(we assume, for argument's sake, that there are an infinite number of
these). We have also found ad hoc solutions for primes p, the
simplest one is
(i, [a+1/(4 a↑2)] i↑{(p+1)/2}) (mod p)
These can be generalized, but you end up using Deligne's Theorem
(Riemann Hypothesis for surfaces (mod p)).
-------
-------
∂24-Nov-88 0841 GOODMAN%uamis@rvax.ccit.arizona.edu Happy Thanksgiving
Received: from rvax.ccit.arizona.edu by SAIL.Stanford.EDU with TCP; 24 Nov 88 08:41:22 PST
Date: Thu, 24 Nov 88 09:30 MST
From: GOODMAN%uamis@rvax.ccit.arizona.edu
Subject: Happy Thanksgiving
To: DUANE.ADAMS@c.cs.cmu.edu, BLUMENTHAL@venera.isi.edu, 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, KNEMEYER@venera.isi.edu,
MCHENRY@GUVAX.BITNET, OUSTER@ginger.berkeley.edu, Ralston@mcc.com,
CWEISSMAN@dockmaster.arpa
X-VMS-To: @NAS, MBBIT
to all of you and your families.
Marjory is still hopeful that our book will appear by early December.
Sy
∂24-Nov-88 0901 JMC
Bass re Moscow circus
∂25-Nov-88 1602 CLT circus
It runs Dec 27 - dec 31. I propose we go Dec 31 at 3pm
I will be surprised if Timothy will sit through much more
than an hour, no matter how much he likes it. So I think
we have to be prepared to leave if he gets fed up.
The tickets range from $10 to $25.
Shall I order 4 at $25?
You better check that Susie and Dan are available on the 31st first.
∂25-Nov-88 1700 JMC philooλsophy
J. Phil. Phenom. REsearch vol. xlix, no. 1 sept 88
Lawrence Foster, Strong Relativism revisited
"...Chris Swoyer argues against a strong relativist thesis
according to which somehting can be true nx in one framework
but false in another. He defends as opxx possible, however,
a weaker relativism according towhich something could be true
in one framework but inexpressible and hence neither true nor
false in another."
All this seems to assume that truth in a framework is a
natural kind about whose properties we can dispute. It
seems more lilexx likely to me that we can have it however
we wish,and choosing one or the other will be subject
only to criteria of utility.
This is to start a collection of citations of philosophical
views taht wrongly assume certain entitites to be natural
kins. Ed Zalta expressed this attitude to blxx beliefs.
∂25-Nov-88 2150 beeson@ucscd.UCSC.EDU triangle problem solved!
Received: from ucscd.UCSC.EDU by SAIL.Stanford.EDU with TCP; 25 Nov 88 21:50:47 PST
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA08155; Fri, 25 Nov 88 21:49:51 PST
Date: Fri, 25 Nov 88 21:49:51 PST
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8811260549.AA08155@ucscd.UCSC.EDU>
To: jmc@sail.stanford.edu
Subject: triangle problem solved!
I found a neat trick. Did you know that rotations in four-space can be
represented using quaternions? It's true: the most general member of
SO_4 can be represented in the form f(x) = axb where a and b are two
unit quaternions. Here, however, we don't need this representation theorem,
as we seek only a CERTAIN orthogonal transformation (it doesn't even have
to preserve lengths). Specifically, we are given that k is a sum of
three squares, say k=uu + vv + ww, so we have a solution of the
triangle equations in 4-space, viz (1,0,0,0) and (0,u,v,w) (two orthogonal
vectors one of which is sqrt(k) times as long as the other, that's what
the triangle equations say). Now we want to "rotate" these two vectors
until their fourth component is zero, thus producing a solution in
three-space. We do that by seeking a quaternion q such that
both (1,0,0,0)q and (0,u,v,w)q (where this means quaternion multiplication)
have zero in the fourth component. Once you see to look for the solution
this way, it's easy to find: one vector is (v-u,w,w), and the
other is (-uw -vw, -u↑2 + uv - w↑2, v↑2 + w↑2 -uv). Without quaternions
I wouldn't have guessed that! (You can type these things into
Mathematica to check the result works, or check it by hand on a page of
paper--I did both.)
I'm writing this up, I'll send you a copy of it pretty soon.
∂26-Nov-88 0739 CLT
please get Thomas's raisin muffins like the ones
we had. the others are very good
∂26-Nov-88 1924 pehoushe@Gang-of-Four.Stanford.EDU N-Queue Qlisp Demonstration
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 26 Nov 88 19:24:21 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02831; Sat, 26 Nov 88 19:23:12 PST
Date: Sat, 26 Nov 88 19:23:12 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8811270323.AA02831@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: N-Queue Qlisp Demonstration
At your convenience, I require about 1 hour of your time.
If you have any specific programs which you would like to see
please let me know ahead of time. Would Monday at 11AM be alright?
The system's main feature is that spawning is both cheap and
garbageless. It uses a scheduling data structure which, internally to a
processor, is a stack, and externally to other processors, is like a
queue, precisely like Halstead's Multi-Lisp. However, his system was
not powerful enough to experimentally investigate the nature of the
n-queue/stack setup. I have run many experiments that spawned
BILLIONS of tasks.
-Dan
∂27-Nov-88 0956 harnad@Princeton.EDU Explanatory Coherence: BBS Call for Commentators
Received: from Princeton.EDU by SAIL.Stanford.EDU with TCP; 27 Nov 88 09:56:05 PST
Received: from psycho.Princeton.EDU by Princeton.EDU (5.58+++/1.87)
id AA05090; Sun, 27 Nov 88 12:52:39 EST
Received: by psycho.Princeton.EDU (3.2/1.65)
id AA08252; Sun, 27 Nov 88 12:35:11 EST
Date: Sun, 27 Nov 88 12:35:11 EST
From: harnad@Princeton.EDU (Stevan Harnad)
Message-Id: <8811271735.AA08252@psycho.Princeton.EDU>
To: connectionists@cs.cmu.edu, epsynet%uhupvm1.bitnet@confidence.Princeton.EDU
Subject: Explanatory Coherence: BBS Call for Commentators
Below is the abstract of a forthcoming target article to appear in
Behavioral and Brain Sciences (BBS), an international,
interdisciplinary journal providing Open Peer Commentary on important
and controversial current research in the biobehavioral and cognitive
sciences. To be considered as a commentator or to suggest other appropriate
commentators, please send email to:
harnad@confidence.princeton.edu or write to:
BBS, 20 Nassau Street, #240, Princeton NJ 08542 [tel: 609-921-7771]
____________________________________________________________________
EXPLANATORY COHERENCE
Paul Thagard
Cognitive Science Loboratory
Princeton University
Princeton NJ 08542
Keywords: Connectionist models, artificial intelligence, explanation,
coherence, reasoning, decision theory, philosophy of science
This paper presents a new computational theory of explanatory
coherence that applies both to the acceptance and rejection of
scientific hypotheses and to reasoning in everyday life. The theory
consists of seven principles that establish relations of local
coherence between a hypothesis and other propositions that explain it,
are explained by it, or contradict it. An explanatory hypothesis is
accepted if it coheres better overall than its competitors.
The power of the seven principles is shown by their implementation in a
connectionist program called ECHO, which has been applied to
such important scientific cases as Lavoisier's argument for
oxygen against the phlogiston theory and Darwin's argument for evolution
against creationism, and also to cases of legal reasoning. The
theory of explanatory coherence has implications for artificial
intelligence, psychology, and philosophy.
∂28-Nov-88 0318 cracraft@venera.isi.edu this'll be news by morning...
Received: from venera.isi.edu by SAIL.Stanford.EDU with TCP; 28 Nov 88 03:18:47 PST
Posted-Date: Mon, 28 Nov 88 03:12:44 PST
Message-Id: <8811281113.AA03009@venera.isi.edu>
Received: from LOCALHOST by venera.isi.edu (5.54/5.51)
id AA03009; Mon, 28 Nov 88 03:13:35 PST
To: jperry@unix.sri.com, jim@rand.org, jhs%hpltbm@sde.hp.com, rg@ai.ai.mit.edu,
jmc@sail.stanford.edu
Subject: this'll be news by morning...
Date: Mon, 28 Nov 88 03:12:44 PST
From: Stuart Cracraft <cracraft@venera.isi.edu>
The following applies to Deep Thought, a chess-machine
project from CMU, manned principally by Hsu, Nowatzyk, Anarathaman.
Project completely unrelated to Berliner/Hitech.
Deep Thought wins Software Toolworks mega-Californament!
Shares first place honors with GM Tony Miles from England.
(Others in same section, GM Tal, GM Larsen, IM Shirazi,
IM McCambridge, GM Browne, IM Silman, many more IM's,
lots of FM's, regular masters, and experts.)
Deep Thought defeated GM Bent Larsen, IM Jeremy Silman, FM LeSierge,
FM Glicksman, FM Salgado, draws IM McCambridge. Only loss was
to GM Browne. This record clinched the first place honors with
GM Miles -- winning scores for Miles and Deep Thought were 6.5 each,
out of a possible 8 points.
Deep Thought's performance rating for the event: 2745.
Stuart
∂28-Nov-88 0744 MPS Filing
It may appear to you that I am far behind, but I just
checked what you have given me and it will only fill
two file folders after I get it typed into your Chron
file. Do you think we need to discuss this to see
if there is a better way for me to do the filing?
Pat
∂28-Nov-88 0900 JMC
800 876 smog
∂28-Nov-88 1425 nilsson@Tenaya.stanford.edu DARPA for lunch?
Received: from Tenaya.stanford.edu by SAIL.Stanford.EDU with TCP; 28 Nov 88 14:25:12 PST
Received: by Tenaya.stanford.edu (5.59/25-eef) id AA08991; Mon, 28 Nov 88 14:23:13 PDT
Date: Mon, 28 Nov 88 14:23:13 PDT
From: Nils Nilsson <nilsson@Tenaya.stanford.edu>
Message-Id: <8811282223.AA08991@Tenaya.stanford.edu>
To: ullman@score, jlh@sonoma, cheriton@score, feigenbaum@sumex, jmc@sail
Subject: DARPA for lunch?
Do any of you think that the recent DARPA PI mtg resulted in insights,
opportunities, etc. that would be important for all of us to hear
and talk about at a Tuesday faculty lunch?
-Nils
∂28-Nov-88 1457 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU,@RITTER.AI.SRI.COM:TYSON@AI.SRI.COM [geoff@wacsvax.OZ: Input refutations]
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 28 Nov 88 14:57:22 PST
Received: from AI.AI.MIT.EDU by SCORE.STANFORD.EDU with TCP; Mon 28 Nov 88 14:55:47-PST
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 28 Nov 88 18:00:30 EST
Received: from RITTER.AI.SRI.COM (TCP 30003002417) by MC.LCS.MIT.EDU 28 Nov 88 17:49:37 EST
Received: from ELCAPITAN.AI.SRI.COM by RITTER.AI.SRI.COM via CHAOS with CHAOS-MAIL id 35422; Mon 28-Nov-88 14:48:45 PST
Date: Mon, 28 Nov 88 14:49 PST
From: Mabry Tyson <TYSON@AI.SRI.COM>
Subject: [geoff@wacsvax.OZ: Input refutations]
To: theorem-provers@mc.lcs.mit.edu
cc: geoff%wacsvax.oz@uunet.uu.net
Included-msgs: The message of 21 Nov 88 23:46 PST from geoff@wacsvax.OZ,
The message of 21 Nov 88 23:46 PST from Geoff Sutcliffe
Message-ID: <19881128224902.0.TYSON@ELCAPITAN.AI.SRI.COM>
This was originally sent to AILIST but it seems most appropriate for
THEOREM-PROVERS@MC.LCS.MIT.EDU. (Make sure your replies go to the
originator (Geoff@wacsvax.oz) of the message, not to me!)
Date: Mon, 21 Nov 88 23:46 PST
From: Geoff Sutcliffe <geoff@wacsvax.OZ>
Subject: Input refutations
To: AIList@AI.AI.MIT.EDU
I have been searching (in the wrong places obviously) for a proof that
resolution & paramodulation, or resolution & paramodulation & factoring,
form a complete input refutation system for sets of Horn clauses, and
that the single negative clause in a minimally unsatisfiable set of
Horn clauses may be used as the top clause in such refutations.
Refutation completeness, without specification of the top clause, is
in "Unit Refutations and Horn Sets" [Henschen 1974]. If set-of-support
is compatible with input resolution,paramodulation,factoring then it
is possible to choose the negative clause as the support set, and the problem
is solved. Is this compatibility known?
Any help, with this seemingly obvious result, would be appreciated.
Geoff Sutcliffe
Department of Computer Science, CSNet: geoff@wacsvax.oz
University of Western Australia, ARPA: geoff%wacsvax.oz@uunet.uu.net
Mounts Bay Road, UUCP: ..!uunet!munnari!wacsvax!geoff
Crawley, Western Australia, 6009.
PHONE: (09) 380 2305 OVERSEAS: +61 9 380 2305
∂29-Nov-88 0706 Rich.Thomason@cad.cs.cmu.edu AI and Philosophical Logic Book
Received: from CAD.CS.CMU.EDU by SAIL.Stanford.EDU with TCP; 29 Nov 88 07:06:52 PST
Date: Tue, 29 Nov 1988 10:04:25 EST
From: Rich Thomason <thomason@CAD.CS.CMU.EDU>
To: jmc@sail.stanford.edu
Cc: thomason
Subject: AI and Philosophical Logic Book
Message-ID: <CMM.0.88.596819065.thomason@CAD.CS.CMU.EDU>
John,
I said in an earlier message that I could get your paper in the
version of the JPL issue on Logic & AI that will be bound as a book if you
could get me Tex formatted electronic copy by December 10. I'd really like
to have the paper. It would be a valuable addition to the volume, and I
hope that there has been enough extra time to make the project feasible.
But I need to do some planning now, and need information. Has the
project gotten to the top of your stack, and do you still think Dec. 10 is a
reasonable date? I had picked this date the last time I was in touch with
the publisher over the phone, and I think that it may be negotiable. But if
I try to stretch it I will need a really firm deadline from you, one that
I can count on.
Thanks,
--Rich
∂29-Nov-88 0937 LASHER@Score.Stanford.EDU ["Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>:]
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Nov 88 09:37:02 PST
Date: Tue 29 Nov 88 09:35:57-PST
From: Rebecca Lasher <LASHER@Score.Stanford.EDU>
Subject: ["Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>:]
To: jmc@Sail.Stanford.EDU
Message-ID: <12450459366.33.LASHER@Score.Stanford.EDU>
Return-Path: <CN.MCS@Forsythe.Stanford.EDU>
Received: from forsythe.stanford.edu by SCORE.STANFORD.EDU with TCP; Tue 29 Nov 88 09:33:54-PST
Date: Tue, 29 Nov 88 09:34:48 PST
To: lasher@score
From: "Rebecca Lasher" <CN.MCS@Forsythe.Stanford.EDU>
Igor,
Please return the book by Terras "Harmonic Analysis on Symmetric
Spaces and Applications I". QA403 T47 1985. It is needed by
another borrower.
Rebecca Lasher
Math/CS Library
cc: John McCarthy
-------
∂29-Nov-88 1147 ABRAIDO@Score.Stanford.EDU Surprise for Devika Subramanian
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Nov 88 11:46:52 PST
Date: Tue 29 Nov 88 11:42:43-PST
From: Leonor Abraido <ABRAIDO@Score.Stanford.EDU>
Subject: Surprise for Devika Subramanian
To: devika-friends: ;
Message-ID: <12450482442.35.ABRAIDO@Score.Stanford.EDU>
Devika Subramanian will be leaving soon to start her new career as an assistant
professor at Cornell. As a going-away present, I'd like to give her a photo
album containing pictures of her friends and colleagues at Stanford.
If you would like to be included, you can either give me a photograph of
yourself, or let me take your picture. In either case, you should get in touch
with me by Monday, 5 December: 415-493-8784 (with answering machine), or email
to abraido@score.stanford.edu.
If you are giving me a photo, it should be larger than wallet size, but no
bigger than 5 inches by 7 inches. You can send it to me at:
4292 Wilkie Way, Apt. M
Palo Alto, CA 94306
Note that I must receive it by 14 December.
If you'd like me to take your picture, I'll be in MJH 232 on Wednesday, 30
November, at about 1:30 p.m. If you can't drop by then, let me know when would
be a convenient time (but please get in touch with me before Monday, 5
December).
Thanks for helping me prepare this present for Devika!
Leonor M. Abraido-Fandino
-------
∂29-Nov-88 1547 MPS
I am going to the library to pick up book.
Pat
∂29-Nov-88 1551 rpg@lucid.com US/Japan Workshop on Parallel Lisp
Received: from lucid.com by SAIL.Stanford.EDU with TCP; 29 Nov 88 15:51:33 PST
Received: from challenger ([192.9.200.17]) by heavens-gate.lucid.com id AA02214g; Tue, 29 Nov 88 15:49:16 PST
Received: by challenger id AA00766g; Tue, 29 Nov 88 15:44:15 PST
Date: Tue, 29 Nov 88 15:44:15 PST
From: Richard P. Gabriel <rpg@lucid.com>
Message-Id: <8811292344.AA00766@challenger>
To: jmc@sail.stanford.edu, clt@sail.stanford.edu, jsw@sail.stanford.edu,
nii@sumex.stanford.edu, arg@lucid.com, rhh@ai.ai.mit.edu,
ran@vx.lcs.mit.edu, gifford@xx.lcs.mit.edu, tk@ai.ai.mit.edu,
gls@think.com, Kessler@cs.utah.edu, pierson@multimax.arpa,
allen@bbn.com, zippel@stony-brook.scrc.symbolics.com
Subject: US/Japan Workshop on Parallel Lisp
Colleagues
This is the first and only announcement for the Joint US/Japan
Workshop on Parallel Lisp, to be held in Sendai, Japan. We expect to
hold a second Parallel Lisp workshop in the US within a year of the
first workshop. Here are the important facts:
Date: June 5, 6, 7, and 8
Place: Aoba Memorial Building
School of Engineering
Tohuko University
Sendai, Japan
Organizers: <US> Dr. Richard P. Gabriel (Stanford University and Lucid, Inc.)
<Japan> Professor Takayasu Ito (Tohoku University)
Major Topics:
Parallelism and Concurrency in Lisp
Parallel Lisp Languages
Parallel Machines for Lisp and Parallel Lisp
Object-oriented Systems for Parallel Lisp
Applications
Right now there are 16 senior researchers from Japan who are coming,
and I would like the US to have a good showing.
At this time I would like to get a list of those people who will
definitely be able to attend. There is limited funding available for
the workshop. Namely, Professor Ito is able to pay for train travel
between Tokyo and Sendai, and possibly housing and meals, for
University folks.
Please send me a note stating whether you will be able to come and a
short abstract of what you would like to say. There will be both long
and short talks by selected individuals.
Please send mail to: rpg@sail.stanford.edu
-rpg-
∂29-Nov-88 1603 GLB
To: der@PSYCH.Stanford.EDU, sf@CSLI.STANFORD.EDU,
JMC@SAIL.Stanford.EDU, jcm@Polya.Stanford.EDU,
CLT@SAIL.Stanford.EDU, JK@SAIL.Stanford.EDU
GLB oral exams:
Thursday dec 1 at 4:00 pm room 301.
∂29-Nov-88 1617 weening@Gang-of-Four.Stanford.EDU Re: US/Japan Workshop on Parallel Lisp
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Nov 88 16:17:53 PST
Received: from LOCALHOST by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01161; Tue, 29 Nov 88 16:16:32 PST
Message-Id: <8811300016.AA01161@Gang-of-Four.Stanford.EDU>
To: Richard P. Gabriel <rpg@lucid.com>
Cc: jmc@sail, clt@sail
Subject: Re: US/Japan Workshop on Parallel Lisp
In-Reply-To: Your message of Tue, 29 Nov 88 15:44:15 -0800.
<8811292344.AA00766@challenger>
Date: Tue, 29 Nov 88 16:16:30 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
I would like to go to the workshop in Japan, if there is funding for
the travel. I can talk about either my thesis work, or the Stanford
part of the Qlisp project.
∂29-Nov-88 1650 CLT US/Japan Workshop on Parallel Lisp
To: RPG@SAIL.Stanford.EDU, JMC@SAIL.Stanford.EDU,
JSW@SAIL.Stanford.EDU
I would not plan to go.
We have 4k in the proposed budget for foreign travel for next year.
∂29-Nov-88 1714 VAL Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
A THEORY OF CONCURRENT ACTIONS
Michael Gelfond
University of Texas at El Paso
Vladimir Lifschitz
Arkady Rabinov
Stanford University
Friday, December 2, 3:15pm
MJH 301
We propose an extension of the situation calculus and of the causality-
based approach to reasoning about action that can be used for describing
concurrent events. Concurrency is represented by an addition operation
on the set of actions. In the absence of information to the contrary,
the causal effect of the sum of several actions is assumed to coincide
with the union of the causal effects of the summands. Mechanisms are
provided for overriding this default. Examples illustrate the use of the
formalization for temporal projection and temporal explanation.
∂29-Nov-88 2011 VAL Re: Commonsense and Nonmonotonic Reasoning Seminar
Received: from Xerox.COM by SAIL.Stanford.EDU with TCP; 29 Nov 88 20:11:42 PST
Received: from Semillon.ms by ArpaGateway.ms ; 29 NOV 88 20:03:15 PST
Date: 29 Nov 88 20:02 PST
From: hayes.pa@Xerox.COM
Subject: Re: Commonsense and Nonmonotonic Reasoning Seminar
In-reply-to: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>'s message of 29 Nov
88 17:14 PST
To: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
cc: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
Message-ID: <881129-200315-1366@Xerox>
If I cant get to Michaels seminar, can someone ask him this question: if
the effect of the sums is the sum of the effects, how can he explain such
phenomena as vector addition of forces, two engines being able to pull a
heavier train than either can alone, two people carrying a table, and
similar instances of useful cooperation? They seem to be rather
important, especially in planning.
Pat
∂29-Nov-88 2121 BEDIT@Score.Stanford.EDU Summary of September computer charges.
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 29 Nov 88 21:21:18 PST
Date: Tue 29 Nov 88 20:48:05-PST
From: Billing Editor <BEDIT@Score.Stanford.EDU>
Subject: Summary of September computer charges.
To: MCCARTHY@Score.Stanford.EDU
Message-ID: <12450581725.17.BEDIT@Score.Stanford.EDU>
Dear Mr. McCarthy,
Following is a summary of your computer charges for September.
Account System Billed Pct Cpu Job Disk Print Adj Total
JMC SAIL 2-DMA807T 100 457.51 28.49 ***.** 6.70 5.00 2427.39
MCCARTHY SCORE 2-DMA807T 100 .00 .00 30.47 .00 5.00 35.47
jmc LABREA 2-DMA807T 100 .00 .00 109.31 .00 5.00 114.31
Total: 457.51 28.49 ***.** 6.70 15.00 2577.17
University budget accounts billed above include the following.
Account Principal Investigator Title
2-DMA807 McCarthy N00039-84-C-0211
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.
-------
∂30-Nov-88 0208 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM Pi
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 30 Nov 88 02:08:09 PST
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 338057; Wed 30-Nov-88 05:04:37 EST
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 77191; Wed 30-Nov-88 01:59:43 PST
Date: Wed, 30 Nov 88 01:59 PST
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: Pi
To: "dbailey@ew11.nas.nasa.gov"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
cc: rwg@RUSSIAN.SPA.Symbolics.COM, math-fun@RUSSIAN.SPA.Symbolics.COM, "jmc@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"dek@sail.stanford.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM, "r@tis-w.arpa"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"hen@bucs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM
In-Reply-To: <8811230130.AA00879@ew11.nas.nasa.gov>
Message-ID: <19881130095919.0.RWG@TSUNAMI.SPA.Symbolics.COM>
[Math-funsters: Dr. Bailey is the Crayoleer who recaptured the pi record
from me when he thought he was stealing it from Japan. Of course, the
Empire struck back.]
Date: Tue, 22 Nov 88 17:30:13 PST
From: David Bailey <dbailey@ew11.nas.nasa.gov>
Bill:
A while back one Roy North sent me a note describing a curious
phenomenon, which he had noted in using Gregory's series to evaluate
pi. I finally took a look at it today, and it does indeed seem
remarkable.
Consider the following. Let Pn be the sum of Gregory's series up to
the 1/(n-1) term:
n/2-1
Pn = 4 Sum 1/(2k+1)
k=0
When n = 1,000,000, the decimal results of this calculation are as
follows (the second line is pi):
3.1415906535897932404626433832695028841972913993751030509749446933498164008
3.1415926535897932384626433832795028841971693993751058209749445923078164062
X XX X XXX XXX X X XX XX
Ordinarily, we think of Gregory's series as impossibly slow in
convergence. However, note that with the exception of some errors,
many digits are correct. In fact, it appears that
Pi - Pn = a1 / n + a3 / n↑3 + a5 / n↑5 + a7 / n↑7 + ...
where the ai are INTEGERS. In particular, from the above it appears
that
a1 = 2
a3 = -2
a5 = 10
a7 = -122
a9 = 2770
a11 = -101042
I have tried to mathematically derive this fact. I did succeed in
finding a1 and a3, but after then the algebra got out of hand. Have
you ever seen any result like this? It seems too basic to be an
unknown fact, particularly given the generations of mathematicians
that have worked with various series for pi.
Any leads would be appreciated.
David H. Bailey
NASA Ames Research Center
415-694-4410
dbailey@ew11.nas.nasa.gov
OK, I just got your (nicely typesot!) physmail. (Even though it was addressed
to a 12 months obsolete location. We moved to
Suite 120
700 East El Camino Real
Mountain View 94040
415-969-0955)
A&S := Abramowitz & Stegun. (I lost mine, but luckily smearoxed the relevant
pages.)
(C346) POWERSERIES(ATAN(Z)/Z,Z,0);
Time= 681 msecs
∞
==== I6 2 I6 + 1
\ (- 1) Z
> -----------------
/ 2 I6 + 1
====
I6 = 0
(D346) ------------------------
Z
(C347) INTOSUM(NICEINDICES(%));
Time= 117 msecs
∞
==== I 2 I
\ (- 1) Z
(D347) > -----------
/ 2 I + 1
====
I = 0
(~ A&S 15.1.5). The nth tail of this is
(C348) (-1)↑N*4*F[2,1](1,N+1/2,N+3/2,-1)/(2*N+1);
Time= 753 msecs
1 3 N
4 F (1, N + -, N + -, - 1) (- 1)
2, 1 2 2
(D348) ------------------------------------
2 N + 1
by A&S 15.1.1. I.e.,
(C349) PARTFRAC(FFSTEP(%/(-1)↑N,3),N);
Time= 1624 msecs
7 9
4 FF(4, N + -, N + -, 4, - 1)
2 2 4 4 4
(D349) - ----------------------------- + ------- - ------- + -------
2 N + 7 2 N + 5 2 N + 3 2 N + 1
Where FF is a device for continuing the series,
(C350) REVERSE(ARGS(FFTRUNC(%)));
Time= 199 msecs
4 4 4 4
(D350) [-------, - -------, -------, - -------]
2 N + 1 2 N + 3 2 N + 5 2 N + 7
and FFTRUNC and FFSTEP are functions I just typed in:
(C351) DISPFUN(FFSTEP,FFTRUNC);
(E351) FFSTEP(EXP, N) := IF N = 0 THEN EXP ELSE FFSTEP(SUBST([F =
2, 1
LAMBDA([A, B, C, Z], FF(A, B, C, 1, Z)), FF =
A B Z FF(A + 1, B + 1, C + 1, D + 1, Z)
LAMBDA([A, B, C, D, Z], 1 + ---------------------------------------)], EXP), N - 1)
C D
(E352) FFTRUNC(EXP) := SUBST(FF = LAMBDA([[IGNORE]], 1), EXP)
Time= 837 msecs
(D352) DONE
By Euler's "linear transformation" A&S 15.3.4,
(C353) D348 = AS1534(D348);
Time= 820 msecs
1 3 N 3 1 N
4 F (1, N + -, N + -, - 1) (- 1) 2 F (1, 1, N + -, -) (- 1)
2, 1 2 2 2, 1 2 2
(D353) ------------------------------------ = ------------------------------
2 N + 1 2 N + 1
where
(C354) FUNDEF(AS1534);
Time= 1 msec
Z
F (A, C - B, C, -----)
2, 1 Z - 1
(D354) AS1534(EXP) := SUBST(F = LAMBDA([A, B, C, Z], -------------------------),
2, 1 A
(1 - Z)
EXP)
E.g., taking 3 terms of Gregory's and 11 terms of Euler's:
(C355) 4*(1-1/3+1/5)+SUBST(3.0,N,FFTRUNC(FFSTEP(RHS(D353),11)));
Time= 2342 msecs
(D354) 3.1415927
Euler's looks like
(C355) FFTRUNC(FFSTEP(RHS(D353),3));
Time= 550 msecs
3
--------- + 1
7
2 (N + -)
2
------------- + 1
5
N + -
2 N
2 (----------------- + 1) (- 1)
3
2 (N + -)
2
(D355) --------------------------------
2 N + 1
or in partial fractions:
(C356) PARTFRAC(%/(-1)↑N,N);
Time= 681 msecs
1 5 11 15
(D356) - ----------- + ----------- - ----------- + -----------
4 (2 N + 7) 4 (2 N + 5) 4 (2 N + 3) 4 (2 N + 1)
Taking 14 terms of Euler's, and expanding about n=infinity
(C357) TAYLOR(FFTRUNC(FFSTEP(RHS(D353)/(-1)↑N,14)),N,inf,15);
Time= 13749 msecs
1 1 5 61 1385 50521 2702765 199360981
(D357)/T/ - - ---- + ----- - ----- + ------ - -------- + -------- - --------- + . . .
N 3 5 7 9 11 13 15
4 N 16 N 64 N 256 N 1024 N 4096 N 16384 N
(I just thought to replace the Infinity character with inf on line (C357). I
hope I haven't forgotten any other exotic characters you might not have.)
Anyway, I'll bet it took you more than 14 seconds to get a[11]. Also notice that
if you choose n = half a power of 10, the digit perturbation is even milder.
I have a formula resembling A&S 15.3.15, based on taking terms alternately from
Gregory's and Euler's, which, I think, explains why the a[even] are 0, and which
converges (asymptotically) at three times the rate of Euler's 1 bit/term.
This is, of course, pathetic next to the *modern* Gregory's series. Gregory
Chudnovsky, that is!
∂30-Nov-88 0743 MPS Conferences
I am not sure when the Turing Institute meeting is, but I
think I remember it is not going to be until Mar. or April. I
can't find the original invitation. The England trip to
Sunderland is the 29th thru 31st of March. The dates for the
trip to Tallin are Dec 12-16.
Pat
∂30-Nov-88 0940 MPS
Do you want me to call A. Spector and find the date that
you speak? I do not remember any correspondence that
specifies the actual time. Only the dates of the conference.
Pat
∂30-Nov-88 1326 ginsberg@polya.Stanford.EDU Formfeed to meet on 12/1
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Nov 88 13:26:43 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA22387; Wed, 30 Nov 88 13:24:55 PDT
Date: Wed, 30 Nov 88 13:24:55 PDT
From: Matthew L. Ginsberg <ginsberg@polya.Stanford.EDU>
Message-Id: <8811302124.AA22387@polya.Stanford.EDU>
To: feed@polya.Stanford.EDU
Subject: Formfeed to meet on 12/1
Don't forget! 252 as usual, and 12.15 as usual ...
Matt
∂30-Nov-88 1352 pehoushe@Gang-of-Four.Stanford.EDU Patenting a scheduling system?
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Nov 88 13:52:23 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA04571; Wed, 30 Nov 88 13:50:54 PST
Date: Wed, 30 Nov 88 13:50:54 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8811302150.AA04571@Gang-of-Four.Stanford.EDU>
To: jmc@sail, clt@sail
Cc: pehoushek@Gang-of-Four.Stanford.EDU
Subject: Patenting a scheduling system?
Would it be unreasonable to attempt to patent the garbageless nstack
scheduling system? It is not necessarily specific to Lisp, and may in
fact be well suited to general purpose multi-processor parallelism.
But I know very little about intelellectual property, or even how much
of my system is "mine", since I work for Stanford.
Do either of you have any advice or experience on this matter?
-dan
∂30-Nov-88 1525 scales@polya.Stanford.EDU qlisp
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Nov 88 15:25:32 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA01502; Wed, 30 Nov 88 15:25:30 PDT
Date: Wed, 30 Nov 88 15:25:30 PDT
From: Daniel J. Scales <scales@polya.Stanford.EDU>
Message-Id: <8811302325.AA01502@polya.Stanford.EDU>
To: jmc@sail
Subject: qlisp
Prof. McCarthy:
Did you get this message? Can we meet this week or next?
> From scales Wed Nov 23 10:55:00 1988
> To: jmc@sail
> Subject: qlisp
> Professor McCarthy:
> I sent you a message before about possibly joining the Qlisp project.
> I have already talked to Joe Weening and read about Qlisp, and I would
> like to meet with you. Do you have time available next week? My best
> times are Monday, Wednesday, Friday before 11am or after 2pm. I will
> try to be at the Qlisp meeting on Monday, but I have a conflict that I
> have to reschedule, so I might not be there.
> Dan Scales
∂30-Nov-88 1627 MPS Files
Hi
I have all your files and have added the date to the ones you keyed
in. I have the ones for October ready to key in tomorrow morning.
They are next to my computer if you need to find something.
Pat
∂30-Nov-88 1807 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Nov 88 18:06:58 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA06039; Wed, 30 Nov 88 18:05:37 PST
Date: Wed, 30 Nov 88 18:05:37 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8812010205.AA06039@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: mjv@edsel.UUCP (M Valvo)
Newsgroups: rec.games.bridge
Subject: Computer Olympiad
Message-ID: <456@edsel.UUCP>
Date: 30 Nov 88 13:16:33 GMT
Organization: AT&T Bell Laboratories, Liberty Corner
Lines: 223
1st
COMPUTER OLYMPIAD
Park Lane Hotel, London
August 9th-15th 1989
* WHAT IS THE COMPUTER OLYMPIAD?
The Computer Olympiad is a kind of Olympic Games for computers
and computer programs. The programs will be playing against each
other at a variety of well known strategy games, such as chess, bridge,
backgammon, GO and Scrabble, and will compete for gold, silver and
bronze medals.
* WHY ORGANIZE A COMPUTER OLYMPIAD?
Ever since 1970 there have been regular computer chess tournaments
in which all of the competitors were computer programs. These
tournaments include national championships, European and North American
Championships, and World Championships, and they have done much to
stimulate interest in writing chess programs. As a result, the best
chess programs have now reached master strength.
In more recent years there have been computer bridge competitions, and
tournaments for playing Reversi (Othello), Go and Go-Moku.
There are now so many strategy games which have been programmed, and
there is so much interest in programming them, that it is high time to
organize a world calibre event combining all of these games.
* WHO WILL BE ELIGIBLE TO TAKE PART?
Any individual, or a company, school or other educational institute,
that has written a computer program to play one or more of these games.
* HOW MANY GAMES MAY EACH COMPETITOR ENTER?
As many as they wish. You may enter a program to play just one game, or
you can enter for every single game with a different program for each.
* WHERE WILL IT BE HELD?
The Park Lane Hotel, London, situated in Picadilly directly opposite
Green Park. This deluxe hotel was the location for the first half of
the 1986 World Chess Championship match between Gary Kasparov and
Anatoly Karpov.
* WHAT ELSE WILL THERE BE IN ADDITION TO THE TOURNAMENTS?
There will be a conference on computer games, at which at which
enthusiasts will present papers on various aspects of programming
strategy games.
There will be exhibition games by some of the best programs against
human experts.
* WHO IS ORGANIZING THE EVENT?
David Levy, an International Master and President of the International
Computer Chess Association.
* HOW WILL THE OLYMPIAD BE ORGANIZED?
There will be a different competition to cater for each game. The exact
format of each competition may vary according to how many entries there
are. In some tournaments it may be possible within the space of seven
days for all of the competitors to play each other. If there are too
many competitors for an all-play-all tournament the Swiss system will
probably be used.
* WHO WILL BE THE REFEREE?
There will be a different arbiter for each game, someone who is an
expert at that particular game.
* WHAT ARE THE PRIZES?
There will be gold, silver and bronze medals awarded for the top places
in each game. In addition there will be special prizes for the
programming team, school, college and company which win the most medals.
* WHAT GAMES WILL BE PLAYED?
Awari (Kalah) Go (9x9)
Backgammon Go (19x19)
Bridge Go-Moku
Chess Mah Jong
Chinese Chess Nine Men's Morris (Muhle)
Connect Four Poker
Cribbage Renju
8x8 Draughts (Checkers) Reversi (Othello)
10x10 Draughts Scrabble
Gin Rummy Shogi (Japanese Chess)
* HOW DO I GET MORE INFORMATION ABOUT THE COMPUTER OLYMPIAD?
Write to: David Levy
Computer Olympiad
11 Loudoun Road
London NW8 OLP
England
Telephone: (01) 624 5551
Telex: 939002 ICHESS G
All correspondence must be in English.
RULES FOR THE COMPUTER OLYMPIAD
Most of the rules apply to all games. These are:
(1) Each entry is a computing system and one or more human operator(s). Each
entry requires at least one full time operator (i. e. one operator cannot
assist in more than one entry).
(2) All computing systems must be on site at the Park Lane Hotel in London.
Entrants are responsible for making their own arrangements to have suitable
computers brought to the tournament site. In the case of special difficulties,
the organizers may decide to assist overseas entrants in obtaining the use of
suitable computers for the duration of the Olympiad.
(3) Unless otherwise specified, rules of play are identical to those in human
play.
(4) Most games will be played with clocks to record the thinking time of each
program. An operator may request the tournament director to stop his program's
clock at most twice in any game because of technical difficulties. The clock
must be restarted each time after no more than 15 minutes. An operator who
experiences technical difficulties is permitted to change to another computer
during the course of a game.
(5) Each entry must be made by one or more of the programmers of that program.
No entry will be allowed without the agreement of at least one member of the
programming team. An exception may be made in the case of a commercially
available program or game playing machine, which may be entered at the
discretion of the organizers.
(6) Each programming team is allowed only one entry in each tournament.
(7) A program may be altered between games but not during a game.
(8) Program parameters may not be changed by the human operator during a game.
(9) The time showing on a program's clock may only be communicated to a program
if it asks for such information.
(10) All programs must have the facility for taking back moves (and bids or
bets), one by one, as far as the start of the game, to allow for the
correction of human operator error. Any such error will be corrected, and the
human who erred may be penalized by the tournament director, for example by a
reduction in the program's remaining time allocation.
(11) If possible a program should record the whole game on disk or printer. In
any event, the operator of each program must provide the tournament director
with a written or printed record of the moves of the game at the end of each
game.
Some games will have their own specific rules. The most important ones, which
may affect your programming plans, are given here:
AWARI (KALAH) Each player will have 6 pits plus his own Kalah. At the start of
the game each pit contains 6 stones.
BACKGAMMON Programs should be able to decide whether or not to accept a double
made by the opponent. Any program which is unable to decide, will be assumed to
accept any double. Beavering is not allowed. Each match between two programs
will be won by the first player to reach a certain number of points. The
current status of the match (i. e. how many more points each program needs to
win the match) may be input to a program at the start of any game or when
recovering from a hardware fault. Each program should be able to play at the
rate of 30 "moves" in every 30 minute period. A "move" includes decisions about whether to double or accept a double.
BRIDGE Each team must use two computers for the bidding and play of a hand.
There will be no direct communication between any of the computers in a hand.
Any bidding system or convention may be used by a program and any leading
conventions. A program may ask at the start of a hand for yes/no answers to
any questions regarding bidding and leading, for example "Do you play ACOL?"
Any such questions must be answered as accurately as possible by the operator
of the opposing program. Programs will be expected to make each bid and play within 30 seconds.
Scoring will be IMP scoring (not match point scoring), i. e. the main objective
will be to make the contract and overtricks are of minor importance.
GO in the 9x9 tournament each program shall have 45 minutes to make all of its
moves. The komi for 19x19 will be 6.5. A program may pass at any time.
Chinese rules will be followed, apart from a few changes: All dead stones must
be removed by capture. The game ends after three successive passes or when of
the programs resigns. All stones on the board are alive. Only completely
surrounded territory is counted. Single stone suicide is exactly the same as a
pass.
POKER The game will be Hold-Em poker which is played in the World Championships
every year in Las Vegas. The stakes will be limit raise, with one unit ante by
each player. The first player bets 5 blind. The second player may pass, call or
raise 5. Any remaining raises before the flop are 10 units. After the flop the
blind player must check or bet 10 units and any raises must be 10 units. After
the 4th up card and the 5th up card the blind player must check or bet 20, and
any raises are also 20 units.
There may be no more than 8 raises at each betting interval.
Programs must be able to make all betting decisions within 15 seconds.
Each round will involve a match between two programs. The match will be won by
the program which wins all its' opponents money (10,000 units) or by the
program which is ahead after 50 hands.
RENJU and GO-MOKU Each program will have 120 minutes in which to make all its
moves.
REVERSI and GO-MUKU Each program will have 30 minutes to make all of its moves.
SCRABBLE The "Official Scrabble Players" Dictionary will be used. Each program
will have 60 minutes to make all of its plays. A program may challenge a word--
if the challenge is upheld, the program which made the word loses its turn. If
the challenge is incorrect, the program which made the challenge loses its turn.
OTHER GAMES No program should take nore than 60 seconds over any decision in a
card game (i. e. a bid or the play of a card). In board games programs will be
required tp play at the rate of 20 moves per hour.
∂30-Nov-88 2145 ilan@Gang-of-Four.Stanford.EDU Computer chess breakthrough
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 30 Nov 88 21:45:11 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA06546; Wed, 30 Nov 88 21:43:52 PST
Date: Wed, 30 Nov 88 21:43:52 PST
From: Ilan Vardi <ilan@Gang-of-Four.Stanford.EDU>
Message-Id: <8812010543.AA06546@Gang-of-Four.Stanford.EDU>
To: jmc@sail
Subject: Computer chess breakthrough
Cc: ilan@Gang-of-Four.Stanford.EDU, rivin@Gang-of-Four.Stanford.EDU
Article 1702 of rec.games.chess:
Path: polya!labrea!rutgers!rochester!pt.cs.cmu.edu!unh.cs.cmu.edu!fhh
From: fhh@unh.cs.cmu.edu (Feng-Hsiung Hsu)
Newsgroups: rec.games.chess
Subject: Deep Thought co-winner in Toolworks
Keywords: GM, Fredkin Intermediate Prize
Message-ID: <3693@pt.cs.cmu.edu>
Date: 28 Nov 88 18:12:02 GMT
Organization: Carnegie-Mellon University, CS/RI
Lines: 94
Deep Thought tied for first place with GM Miles in the open section of the
Software Toolworks Chess Championship held this weekend in Long Beach, CA.
Over 500 players were attracted by the $130,000 total prize fund, 76 of them
(mainly rated over 2300) in the open section, including former world
champion GM Mikhail Tal, the "Great Dane" GM Bent Larsen, GM Samuel
Reshevsky, GM Walter Browne, GM Gurevich, GM Gruefeld, GM Tony Miles and
several IMs. Deep Thought ended with a score of 6.5 out of 8. DT's
opponents were :
Name USCF rating DT's score
LeSiege 2334 1.0
Glicksman 2388 1.0
Bent Larsen (GM) 2590 1.0
Walter Browne (GM) 2640 0.0
McCambridge (IM) 2599 0.5
Salgado (FM) 2388 1.0
Fishbein (FM) 2572 1.0
Silman (IM) 2507 1.0
Deep Thought achieved a performance rating of 2745 in this tournament,
breaking the previous record for computers (set by DT itself) by over 100
points. GM Bent Larsen had the misfortune to find his name becoming
the answer to the Trivial Pursuit question "Who is the first GM to lose
to a computer under regular time control in a tourament?" by his third round
loss to DT.
Deep Thought's new established rating is estimated at over 2545.
The 6-month old Deep Thought has now played 42 rated games. It played
against International Masters 7 times, and won 5, drew 2, no loss. It
played International Grand Masters 3 times, and won 1, lost 2 (the loss
against GM Lev Alburt was due to a bug that caused the machine to throw
away a repetition draw).
Deep Thought has now met the qualifications for the Fredkin Intermediate
Prize for the first computer with a rating of over 2500 in 25 consecutive
games.
The Deep Thought team would like to take this opportunity to thank all
those whose help and encouragement made this possible, particularly
Lawrence Butcher, Stuart Cracraft, Jim Gillogly, Peter Jansen, Larry Kaufmann,
Kau-Fu Lee, Tom Mitchel, Raj Reddy, Danny Sleator, Ken Thompson, Hide Tokuda,
John Zsarnay and our advisors, Roberto Bisiani, Ed Clarke, H. T. Kung, and
Bob Sproull.
We would also like to give our special thanks to Stuart Cracraft for spending
his Thanksgiving Weekend operating DT on site.
The Deep Thought team includes: Thomas Anantharaman, Mike Browne, Murray
Campbell, Feng-hsiung Hsu, and Andreas Nowatzyk, all with the Computer Science
Department at Carnegie Mellon University.
Some of the more interesting games follow.
DT vs. Glicksman (2388), round 2
1. e4,e6; 2. d4,d5; 3. Nc3,Bb4; 4. e5,Ne7; 5. Bd2,c5;
6. Nb5,B:d2; 7. Q:d2,Nf5; 8. dc5,a6; 9. Nd6,N:d6; 10. cd6,Nc6;
11. f4,f6; 12. Nf3,o-o; 13. Be2,fe5; 14. fe5,Rf5; 15. Qc3,d4;
16. Qd2,N:e5; 17. N:d4,Qh4; 18. g3,Qe4; 19. o-o-o,Rf2; 20. Rhe1,Bd7;
21. Qc3,Rc8; 22. Bc4,Qg2; 23. R:e5,b5; 24. B:e6,B:e6; 25. Rc5,R:c5;
26. Q:c5,Bg4; 27. Qc6,Rf1; 28. Qe8,Rf8; 29. d7,Qd5; 30. Rf1,resigns.
White announced mate in 19 moves (37 plies).
GM Bent Larsen (FIDE 2560) vs. DT, round 3.
1. c4,e5; 2. g3,Nf6; 3. Bg2,c6; 4. Nf3,e4; 5. Nd4,d5;
6. c:d5,Q:d5; 7. Nc2,Qh5; 8. h4,Bf5; 9. Ne3,Bc5; 10. Qb3,b6;
11. Qa4,o-o; 12. Nc3,b5; 13. Qc2,B:e3; 14. de3,Re8; 15. a4,b4;
16. Nb1,Nbd7; 17. Nd2,Re6; 18. b3,Rd8; 19. Bb2,Bg6; 20. Nc4,Nd5;
21. o-o-o,N7f6; 22. Bh3,Bf5; 23. B:f5,Q:f5; 24. f3,h5; 25. Bd4, Rd7;
26. Kb2,Rc7; 27. g4,hg4; 28. Rhg1,c5; 29. f:g4,N:g4; 30. B:g7,Rg6;
31. Qd2,Rd7; 32. R:g4,R:g4; 33. Ne5,N:e3; 34. Q:d7,N:d1; 35. Q:d1,Rg3;
36. Qd6,K:g7; 37. Nd7,Re3; 38. Qh2,Kh7; 39. Nf8,Kh8; 40. h5,Qd5;
41. Ng6,fg6; 42. hg6,Kg7; 43. Qh7,Kf6; 44. resigns.
The following game is the second time that DT won a R,B vs R ending.
DT vs. FM Alex Fishbein (2572), round 7
1. e4,e5; 2. Nf3,Nc6; 3. Bb5,a6; 4. B:c6,dc6; 5. o-o,f6;
6. d4,ed4; 7. N:d4,c5; 8. Ne2,Q:d1; 9. R:d1,Bd7; 10. Bf4,o-o-o;
11. c4,Ne7; 12. Nc3,Re8; 13. Be3,Nc6; 14. Rd2,Ne5; 15. b3,h5;
16. h3,b6; 17. Nd5,Bc6; 18. a4,a5; 19. Nc3,Kb7; 20. Nb5,B:b5;
21. ab5,Bd6; 22. f3,Ra8; 23. f4,Nf7; 24. Kf2,Rhe8; 25. Kf3,Bf8;
26. Nc3,Nd6; 27. Rd5,g6; 28. g4,hg4; 29. hg4,Bg7; 30. f5,gf5;
31. gf5,Rh8; 32. Bf4,Rh3; 33. Bg3,Rg8; 34. Rd3,Bf8; 35. Kg2,Rh4;
36. Re1,Rhg4; 37. Kh3,a4; 38. ba4,Rg7; 39. Bh4,N:c4; 40. B:f6,Rg8;
41. e5,Bh6; 42. Nd5,Bf4; 43. Rf3,B:e5; 44. R:e5,N:e5; 45. B:e5,R4g5;
46. Bg3,c4; 47. Nc3,Rh5; 48. Kg2,R8g5; 49. f6,Rf5; 50. Re3,R:f6;
51. Re7,Rf5; 52. R:c7,Ka8; 53. a5,R:b5; 54. N:b5,R:b5; 55. ab6,R:b6;
56. Bf4,Rb3; 57. Kf2,Rb7; 58. R:c4,Rb2; 59. Ke3,Kb7; 60. Ke4,Rb5;
61. Be3,Rh5; 62. Bd4,Rg5; 63. Be5,Kb6; 64. Kd5,Kb5; 65. Rc8,Rg6;
66. Rc5,Kb6; 67. Rc2,Rh6; 68. Rb2,Ka6; 69. Bd6,Rh5 (sealed move)
70. Kc6,Rg5; 71. Rb3,Rh5; 72. Rb1,Rg5; 73. Rb2,Rh5; 74. Rb8,Rh7;
75. Bf8,Ka7; 76. Rb3,Ka8; 77. Re3,Rb7; 78. Re4,Rf7; 79. Re8,Ka7;
80. Bc5,Ka6; 81. Ra8, resigns.
--
∂30-Nov-88 2244 @ELEPHANT-BUTTE.SCRC.Symbolics.COM:rwg@RUSSIAN.SPA.Symbolics.COM n log n
Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by SAIL.Stanford.EDU with TCP; 30 Nov 88 22:44:02 PST
Received: from RUSSIAN.SPA.Symbolics.COM by ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 338382; Thu 1-Dec-88 01:41:43 EST
Received: from TSUNAMI.SPA.Symbolics.COM by RUSSIAN.SPA.Symbolics.COM via CHAOS with CHAOS-MAIL id 77264; Wed 30-Nov-88 22:36:31 PST
Date: Wed, 30 Nov 88 22:36 PST
From: Bill Gosper <rwg@RUSSIAN.SPA.Symbolics.COM>
Subject: n log n
To: math-fun@RUSSIAN.SPA.Symbolics.COM
cc: rwg@RUSSIAN.SPA.Symbolics.COM, "hen@bu-cs.bu.edu"@ELEPHANT-BUTTE.SCRC.Symbolics.COM,
"r@tis-w.arpa"@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: <19881201063602.2.RWG@TSUNAMI.SPA.Symbolics.COM>
A source moves at velocity (1,0) out the positive x axis, emitting
particles with velocities (-1/2, 1/2) back toward the positive y axis,
where they are refracted (hang a left) into the 2nd quadrant at a new
heading of (-1/2, -1/2). This takes them to the negative x axis where
they again turn left, etc., so that each particle is in a square orbit.
Between their emission and first refraction, the particles for a straight
line (wave) which is not parallel to their individual direction of motion.
What is the slope of that wave? What is it after 1 refraction? After n?
After -1? (I.e., suppose the source was not a source, but just the moving
impact point of the 0th refraction.) What is a smooth curve(t,theta)
through all the refraction points?
Now suppose that each particle fissions in two upon crossing the x-axis,
and that one of the two simply sails off instead of refracting. What is the
growth rate of the particle population?
(Answers below.)
The particles are gliders, the source is a glider puffer. The axes are the
exhausts of spaceship and eater puffers. I have constructed the primary
components of this device, but would (will?) take several days to perfect
the axis puffers. Meanwhile, I cheated and simulated the exhausts with
REPEAT modules. And I actually thought it would be easier than a breeder.
This is how the top level looks in my Life CAD system.
(defun cheat
(&key (reps 1024) (d2-stretch -68) (d3-stretch -67) (ee-stretch 12) (k-stretch -32)
(e-stretch -16)
&aux (w (car (send life-io :windows)))
(d1 (send (make-instance 'mw-glider-doubler) :inverse-poise :poise-input-glider))
(eater (send (send (make-instance 'glider-eater) :inverse-poise :poise-glider)
:withdraw-wrt-glider e-stretch))
(e1 (send d1 :poise-odd-glider (clone eater)))
(doubler (send (make-instance 'mw-glider-doubler)
:inverse-poise :poise-input-glider))
(d2 (send d1 :poise-even-glider
(send (clone doubler) :withdraw-wrt-glider d2-stretch)))
(bouncer (send (send (make-instance 'mw-g-bouncer)
:inverse-poise :poise-input-glider)
:withdraw-wrt-glider k-stretch))
(d3 (send d2 :poise-even-glider
(send (clone doubler) :withdraw-wrt-glider d3-stretch)))
(e3 (send d3 :poise-odd-glider eater))
(k3 (send d2 :poise-odd-glider (clone bouncer)))
(d4 (send d3 :poise-even-glider
(send (clone doubler) :withdraw-wrt-glider d2-stretch)))
(k1 (send d4 :poise-odd-glider bouncer))
(doublers (send (make-instance 'doubler-eater :stretch ee-stretch)
:merge-wrt :noop
(make-instance 'repeat :dx 96 :dy 0 :n reps
:inferiors `(,(make-instance 'mw-glider-doubler)))))
(eaters (make-instance 'repeat :inferiors `(,(make-instance 'glider-eater))
:dx 0 :dy -48 :n (* 2 reps)))
(bouncer (make-instance 'terminated-mw-g-bouncer))
(bouncers (unite (send bouncer :poise-middleweight
(make-instance 'middleweight-eater :slip (* 2 ee-stretch)))
(make-instance 'repeat
:dx 96 :dy 0 :n reps :inferiors `(,bouncer))
(make-instance 'repeat :inferiors
`(,(send bouncer :poise-snuff-block
(make-instance 'fancy-block)))
:dx 96 :dy 0 :n (* 2 reps) :x 48))))
; (break "bouncer ~a, k1 ~a" bouncer k1)
(send w :set-cell (repeat (make-glider) 24 -24 reps :merge
(send d1 :externalize (clone doublers) :assemble) :merge
(send d2 :externalize (clone doublers) :assemble) :merge
(send e1 :externalize (clone eaters) :assemble) :merge
(send d3 :externalize (clone doublers) :assemble) :merge
(send k3 :externalize (clone bouncers) :assemble) :merge
(send d4 :externalize doublers :assemble) :merge
(send e3 :externalize eaters :assemble) :merge
(send k1 :externalize bouncers :assemble))))
The CLONE generic functions reflect an erroneous decision to have some of the low
level transformations act "destructively".
In effect, the innermost glider orbit is two glider guns of period 1536, and
successive orbits have periods increasing by 768. If the periods were equal,
this would be a "breeder" with growth rate ~ t↑2. You may be surprised by
concluding that arithmetically increasing periods knock the growth way down
to t log t.
After the nth refraction, the slope is (-)↑(n-1) (2n+1)/(2n+3), which = -1 for
n=-1. (I inject the initial glider particles with a slope -1 space-rake.)
Unfortunately, the mechanism is so large and sparse that, by the time the
innermost glider starts its second orbit, the first wave is off the screen.
Otherwise, it would be a very visual construction.
It seems to me one could colorfully realize this winding spiral with one of
Fredkin's CAMs. Color discontinuties would have two bits of propagation
direction, and a cell with the axisness bit would increment this direction.
The smooth curve is ~ t exp(i theta)/theta. Anybody remember the name of this
spiral?
There must be an easier n log n that, say, emits consecutive integers in binary.
∂01-Dec-88 0355 @Score.Stanford.EDU,@AI.AI.MIT.EDU,@MC.LCS.MIT.EDU:bundy%aiva.edinburgh.ac.uk@NSS.Cs.Ucl.AC.UK Research Job at Edinburgh
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Dec 88 03:55:04 PST
Received: from AI.AI.MIT.EDU by SCORE.STANFORD.EDU with TCP; Thu 1 Dec 88 03:53:44-PST
Received: from MC.LCS.MIT.EDU (CHAOS 3131) by AI.AI.MIT.EDU 1 Dec 88 06:59:59 EST
Received: from NSS.Cs.Ucl.AC.UK (TCP 20012204403) by MC.LCS.MIT.EDU 1 Dec 88 06:19:14 EST
Received: from aiva.edinburgh.ac.uk by NSS.Cs.Ucl.AC.UK via Janet with NIFTP
id aa06678; 30 Nov 88 21:51 GMT
Date: Wed, 30 Nov 88 10:25:09 GMT
Message-Id: <905.8811301025@etive.aiva.ed.ac.uk>
From: Alan Bundy <bundy%aiva.edinburgh.ac.uk@NSS.Cs.Ucl.AC.UK>
Subject: Research Job at Edinburgh
To: theorem-provers@mc.lcs.mit.edu
I would be grateful if you could post the following advert.
Alan Bundy
Department of Artificial Intelligence
University of Edinburgh
RESEARCH ASSOCIATE
(Mathematical Reasoning)
Applications are invited for an SERC supported post, tenable, as soon
as possible, on a mutually agreed date. Appointment will be to October
1 1989, initially, but with a strong possibility of renewal to at
least 1 October 1991. 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 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 on the AR1A scale in
the range 9,865 - 15,105 pounds p.a., according to age 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,
SCOTLAND.
as soon as possible. The closing date for applications is 16th
January 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 5613.
∂01-Dec-88 0720 CLT Umbrella
To: Nillson@SCORE.STANFORD.EDU, bscott@SCORE.STANFORD.EDU,
JMC@SAIL.Stanford.EDU
Here is what I have found out.
Both Pullen and Scherlis agree that the DARPA contracting agency
is currently overloaded and unlikely to take on the new Umbrella.
They also both agree that we should get a white paper/draft version
to the program managers asap and to get agreement on the
text and budget. Then we can haggle about contracting agencies.
Pullen seems pretty much in favor of umbrella mode and
gave the email address that consists of all program managers.
He suggests sending the draft via email to that list and said he would
see that some action was taken.
According to Scherlis we should send
Introductory letter -- here it is guys please act.
A summary document
current situation
summarize current main projects, $/fy, next event
summarize new efforts
thoughts as how to procede
The umbrella text
Budget
I suggest we do this ASAP. I will help with the summary and
introductory letter and text. Maybe JMC would also look
at the text and make some helpful suggestions.
I emphasized to Pullen our unhappiness with SPAWAR.
He agreed, but also said that (1) SPAWAR is one of the
few places this sort of contract can be done (DSSW is an
alternative) (2) SPAWAR has been losing a lot of DARPA
contracts and he is making clear to Machato that they
need to shape up. Tasks should take only a month once
the umbrella is in place (the current record min that
I know of is 3months and I told him so).
I am also sending him a few more point of ammunition.
He feels that pressure by him on Machato will help
and said the current action on the cost overruns
is due to his talking to Machato.
According to Englemore there is a new ISTO deputy director -- Morose
who will sign off task orders (instead of Schwartz). Probably
the umbrella would go him after the program managers have finished
with it.
∂01-Dec-88 1159 pehoushe@Gang-of-Four.Stanford.EDU Qlisp Special Variables Benchmark
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Dec 88 11:57:17 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA08515; Thu, 1 Dec 88 11:54:38 PST
Date: Thu, 1 Dec 88 11:54:38 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8812011954.AA08515@Gang-of-Four.Stanford.EDU>
To: rpg@Gang-of-Four.Stanford.EDU, qlisp@Gang-of-Four.Stanford.EDU
Cc: pehoushek@Gang-of-Four.Stanford.EDU
Subject: Qlisp Special Variables Benchmark
I tried a couple variants of the Gabriel STAK benchmark, which uses
special variables, using the nstack system.
The two main parallel variants use #! and #?:
#! always causes tasks to be spawned.
#? uses the QEMPTY-P predicate which Joe spoke of in his orals.
A pure serial version of this program, without spawning or predicate
testing takes 8317 milliseconds. The #! version yields a speed-up of
6.1//8, the #? version yields 7.4//8. There is another variant which
tests the depth of the stack, and works slightly better than qempty.
This experiment shows that the special variable binding works (the
answer is 7), that sometimes spawning is slightly better than spawning
all the time, and that special variables are expensive to use for
small tasks, and that spawning in the nstack system is cheap.
I tried to squeeze informative column headings on the output below.
The first column, PT, is the parallel time on 8 processors. The next 2
columns are the number of spawns followed by an estimate of the serial
cpu time required to perform those spawns. Next are an Idle counter
plus the estimated amount of serial temporal cost. All times are in
milliseconds.
The current implementation of qlisp can't even come close to these
results, so comparison is unfair.
;;; Begin code. Always-Spawn version, variant of Gabriel benchmark.
(defvar *x*)
(defvar *y*)
(defvar *z*)
(proclaim '(type fixnum *x* *y* *z*))
(defun stak (*x* *y* *z*)
(declare (special *x* *y* *z))
(stak-aux))
(defun stak-aux ()
(if (not (< *y* *x*))
*z*
(multiple-value-bind (x y z)
#!(values (stak (1- *x*) *y* *z*)
(stak (1- *y*) *z* *x*)
(stak (1- *z*) *x* *y*))
(let ((*x* x)
(*y* y)
(*z* z))
(declare (special *x* *y* *z))
(stak-aux)))))
;;; End code.
> (cpu (stak 18 12 6)) ;; run the experiment 10 times
ptime spawns cost idle-cntr cost (the costs are in total serial cpu milliseconds)
PT: 1353 Sp:31804 1822 Id: 15184 121
PT: 1367 Sp:31804 1822 Id: 6977 55
PT: 1357 Sp:31804 1822 Id: 8141 65
PT: 1351 Sp:31804 1822 Id: 7305 58
PT: 1353 Sp:31804 1822 Id: 10136 81
PT: 1354 Sp:31804 1822 Id: 10851 86
PT: 1343 Sp:31804 1822 Id: 8547 68
PT: 1345 Sp:31804 1822 Id: 8359 66
PT: 1368 Sp:31804 1822 Id: 10145 81
PT: 1358 Sp:31804 1822 Id: 11817 94
#P:8 (STAK 18 12 6)
CPU (min mean stddev): 1343 1354.9 7.7
Spawn (min mean stddev):31804 31804.0 11.3
NIL
>
Now, trying the #? version.
> (cpu (stak 18 12 6))
ptime spawns cost idle-cntr cost (the costs are in total serial cpu milliseconds)
PT: 1106 Sp: 732 41 Id: 12469 99
PT: 1100 Sp: 756 43 Id: 9653 77
PT: 1115 Sp: 742 42 Id: 13907 111
PT: 1143 Sp: 894 51 Id: 12768 102
PT: 1108 Sp: 790 45 Id: 14422 115
PT: 1136 Sp: 1064 60 Id: 17797 142
PT: 1118 Sp: 894 51 Id: 15143 121
PT: 1119 Sp: 662 37 Id: 10724 85
PT: 1122 Sp: 860 49 Id: 11757 94
PT: 1102 Sp: 734 42 Id: 11860 94
#P:8 (STAK 18 12 6)
CPU (min mean stddev): 1100 1116.9 13.4
Spawn (min mean stddev): 662 812.8 110.8
NIL
>
∂01-Dec-88 1256 beeson@ucscd.UCSC.EDU Question for circumscription theory
Received: from ucscd.UCSC.EDU ([128.114.129.2]) by SAIL.Stanford.EDU with TCP; 1 Dec 88 12:56:19 PST
Received: by ucscd.UCSC.EDU (5.59/1.28)
id AA27699; Thu, 1 Dec 88 12:55:06 PST
Date: Thu, 1 Dec 88 12:55:06 PST
From: beeson@ucscd.UCSC.EDU (20012000)
Message-Id: <8812012055.AA27699@ucscd.UCSC.EDU>
To: clt@ucscd.UCSC.EDU, jmc@sail.stanford.edu, val@sail.stanford.edu
Subject: Question for circumscription theory
How can you formalize this piece of commonsense knowledge without
immediate contradictions: Each individual person is normally honest
(here honest = non-thief). But normally in a crowd there is at least
one thief.
∂01-Dec-88 1350 VAL re: Question for circumscription theory
To: beeson@ucscd.UCSC.EDU, clt@ucscd.UCSC.EDU, JMC@SAIL.Stanford.EDU
[In reply to message from beeson@ucscd.UCSC.EDU sent Thu, 1 Dec 88 12:55:06 PST.]
> How can you formalize this piece of commonsense knowledge without
> immediate contradictions: Each individual person is normally honest
> (here honest = non-thief). But normally in a crowd there is at least
> one thief.
No problem--the conclusion will be that there is *exactly* one thief.
Unfortunately, we won't be able to prove about any particular individual
that he's honest. That is called the "lottery paradox", and apparently is
not captured by circumscription and similar formalisms.
--Vladimir
∂01-Dec-88 1554 MPS
Bert Sutherland caaled at 4:54
He may try again around 4:30.
Pat
∂01-Dec-88 1558 VAL Reminder: Commonsense and Nonmonotonic Reasoning Seminar
To: "@CS.DIS[1,VAL]"@SAIL.Stanford.EDU
From: Vladimir Lifschitz <VAL@SAIL.Stanford.EDU>
A THEORY OF CONCURRENT ACTIONS
Michael Gelfond
University of Texas at El Paso
Vladimir Lifschitz
Arkady Rabinov
Stanford University
Friday, December 2, 3:15pm
MJH 301
We propose an extension of the situation calculus and of the causality-
based approach to reasoning about action that can be used for describing
concurrent events. Concurrency is represented by an addition operation
on the set of actions. In the absence of information to the contrary,
the causal effect of the sum of several actions is assumed to coincide
with the union of the causal effects of the summands. Mechanisms are
provided for overriding this default. Examples illustrate the use of the
formalization for temporal projection and temporal explanation.
∂01-Dec-88 1628 MPS telephone call
Ed. Feig;enbaum called you and said he would call
you around 8:00 tonight our time.
Pat
PS at your home
∂01-Dec-88 1640 STAGER@Score.Stanford.EDU Spring CS309C
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Dec 88 16:40:26 PST
Date: Thu 1 Dec 88 16:38:23-PST
From: Claire Stager <STAGER@Score.Stanford.EDU>
Subject: Spring CS309C
To: jmc@Sail.Stanford.EDU
Office: CS-TAC 29, 723-6094
Message-ID: <12451060556.35.STAGER@Score.Stanford.EDU>
Do you happen to have a phone number or email address for Witold Litwin?
I need to get in touch with him about scheduling days and times for his
course.
Thanks.
Claire
-------
∂01-Dec-88 1648 BERGMAN@Score.Stanford.EDU NSF salary cap
Received: from Score.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Dec 88 16:48:54 PST
Date: Thu 1 Dec 88 16:41:37-PST
From: Sharon Bergman <BERGMAN@Score.Stanford.EDU>
Subject: NSF salary cap
To: jmc@Sail.Stanford.EDU
Message-ID: <12451061143.22.BERGMAN@Score.Stanford.EDU>
John, Pat just asked me about the reduction in your salary
on the NSF budget. Here is a statement from a memo that Robert
Byer (Vice Provost and Dean of Research) recently sent out concernig
this subject: "When Congress approved the National Science Foundation's
FY'89 Appropriations Bill, they placed a limit on the level of
compensation individuals may be paid as direct charges to those NSF
grants awarded with FY'89 appropriations. As a result, salaries
charged this year to NSF grants which are awarded on or after 10/1/88,
may not exceed a full-time rate of $95K for a 12 month period, or
$7917 per month. . ."
This does not mean that your salary is being reduced. It only
means that your salary charge on an NSF grant can go up to $7917/mo.,
and anything in excess of this amount must be charged elsewhere.
If you are interested in seeing a copy of Byer's memo, let me know.
-Sharon
-------
∂01-Dec-88 1853 gurevich@polya.Stanford.EDU
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 1 Dec 88 18:53:32 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA22580; Thu, 1 Dec 88 18:53:26 PDT
Date: Thu, 1 Dec 88 18:53:26 PDT
From: Yuri Gurevich <gurevich@polya.Stanford.EDU>
Message-Id: <8812020253.AA22580@polya.Stanford.EDU>
To: jmc@sail
Dear Professor McCarthy,
Did I offend you in any way? It would be stupid on my side to offend
one of my scientific heroes. In any case, I would like to have a
chance to hear what are working on these days.
Sincerely,
-Yuri Gurevich
∂01-Dec-88 2010 RFC Prancing Pony Bill
Prancing Pony bill of JMC John McCarthy 1 December 1988
Previous Balance 16.24
Monthly Interest at 1.0% 0.16
Current Charges 4.00 (bicycle lockers)
-------
TOTAL AMOUNT DUE 20.40
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 7/12/88.
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.
∂02-Dec-88 0546 Rich.Thomason@cad.cs.cmu.edu AI and Philosophical Logic Book Again
Received: from CAD.CS.CMU.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 05:46:36 PST
Date: Fri, 2 Dec 1988 8:45:40 EST
From: Rich Thomason <thomason@CAD.CS.CMU.EDU>
To: jmc@sail.stanford.edu
Cc: thomason
Subject: AI and Philosophical Logic Book Again
Message-ID: <CMM.0.88.597073540.thomason@CAD.CS.CMU.EDU>
John,
I'm resending this message. I'll need an answer soon to advise
the publisher.
I said in an earlier message that I could get your paper in the
version of the JPL issue on Logic & AI that will be bound as a book if you
could get me Tex formatted electronic copy by December 10. I'd really like
to have the paper. It would be a valuable addition to the volume, and I
hope that there has been enough extra time to make the project feasible.
But I need to do some planning now, and need information. Has the
project gotten to the top of your stack, and do you still think Dec. 10 is a
reasonable date? I had picked this date the last time I was in touch with
the publisher over the phone, and I think that it may be negotiable. But if
I try to stretch it I will need a really firm deadline from you, one that
I can count on.
Thanks,
--Rich
∂02-Dec-88 0812 pehoushe@Gang-of-Four.Stanford.EDU CTAK test
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 08:12:44 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA00979; Fri, 2 Dec 88 08:10:57 PST
Date: Fri, 2 Dec 88 08:10:57 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8812021610.AA00979@Gang-of-Four.Stanford.EDU>
To: RPG@sail, Qlisp@Gang-of-Four.Stanford.EDU
Cc: pehoushek@Gang-of-Four.Stanford.EDU
Subject: CTAK test
The following is a variant of the Gabriel benchmark, running on the
nstack system, testing catch and throw. This is not realy an adequate
test of qcatch and qthrow, but it is a start. Using the serial common
lisp version of Catch and Throw, the serial running time is 2236
milliseconds. Using Qcatch and Qthrow, the serial running time is
4761. Qcatch and Qthrow are expensive. Note that, in this program,
it is possible to use Catch and Throw, instead of Qcatch and Qthrow,
because there are no interprocess throws. The major cost, then, is in
allocating and deallocating a complex qcatch frame.
As in STAK (my previous message), we tried #! and #?. The
always-spawn version gives a speed-up of 5.3//8 over the Qcatch
serial, and the sometimes-spawn version yields 7//8. Speed-up over
the true serial version is 2.5 and 3.3, respectively.
The experimental output is identical in format to that used and
explained my the previous message.
;;; Variant of Gabriel CTAK benchmark, Qcatch/Qthrow.
(defun ctak (x y z)
(qcatch 'ctak
(ctak-aux x y z)))
(defun ctak-aux (x y z)
(declare (fixnum x y z))
(cond ((not (< y x)) ;xy
(qthrow 'ctak z))
(t #!(ctak-aux
(ctak (1- x) y z)
(ctak (1- y) z x)
(ctak (1- z) x y)))))
> (cpu (ctak 18 12 6))
PT: 905 Sp:31804 1822 Id: 9442 75
PT: 890 Sp:31804 1822 Id: 5263 42
PT: 890 Sp:31804 1822 Id: 6074 48
PT: 885 Sp:31804 1822 Id: 6377 51
PT: 900 Sp:31804 1822 Id: 5428 43
PT: 895 Sp:31804 1822 Id: 7299 58
PT: 894 Sp:31804 1822 Id: 6438 51
PT: 900 Sp:31804 1822 Id: 7785 62
PT: 890 Sp:31804 1822 Id: 7043 56
PT: 905 Sp:31804 1822 Id: 7992 64
#P:8 (CTAK 18 12 6)
CPU (min mean stddev): 885 895.4 6.5
Spawn (min mean stddev):31804 31804.0 11.3
;;; Sometimes spawn version, using #? (Qempty predicate)
> (cpu (ctak 18 12 6))
PT: 681 Sp: 656 37 Id: 9592 76
PT: 686 Sp: 676 38 Id: 8488 67
PT: 684 Sp: 768 44 Id: 11449 91
PT: 680 Sp: 728 41 Id: 7072 56
PT: 696 Sp: 826 47 Id: 11085 88
PT: 711 Sp: 604 34 Id: 6306 50
PT: 684 Sp: 888 50 Id: 8304 66
PT: 681 Sp: 674 38 Id: 8821 70
PT: 683 Sp: 740 42 Id: 9472 75
PT: 694 Sp: 912 52 Id: 11192 89
#P:8 (CTAK 18 12 6)
CPU (min mean stddev): 680 688.0 9.2
Spawn (min mean stddev): 604 747.2 96.6
∂02-Dec-88 0857 gurevich@polya.Stanford.EDU Lunch
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 08:57:26 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA12115; Fri, 2 Dec 88 08:57:22 PDT
Date: Fri, 2 Dec 88 08:57:22 PDT
From: Yuri Gurevich <gurevich@polya.Stanford.EDU>
Message-Id: <8812021657.AA12115@polya.Stanford.EDU>
To: JMC@SAIL.Stanford.EDU
In-Reply-To: John McCarthy's message of 01 Dec 88 2100 PST <4JbM$@SAIL.Stanford.EDU>
Subject: Lunch
Thank you for the message.
Today I am going to Berkeley (to speak on their Logic Colloquium),
but the next Friday, Dec. 9, would be perfect. According to the
Faculty/Staff Directory you are in room 356. Should I come there
about noon? This may be the default: If I do not hear from you,
then I will show up about noon, on Friday, Dec. 9, in room 356.
-Yuri
∂02-Dec-88 1020 chandler@polya.Stanford.EDU Faculty Repoirt
Received: from polya.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 10:20:38 PST
Received: by polya.Stanford.EDU (5.59/25-eef) id AA15362; Fri, 2 Dec 88 10:20:14 PDT
Date: Fri, 2 Dec 1988 10:19:58 PST
From: "Joyce R. Chandler" <chandler@polya.stanford.edu>
To: binford@coyote, cheriton@pescadero, rwf@sail, genesereth@score,
golub@patience, zm@sail, mayr@polya.Stanford.EDU, jmc@sail,
jcm@polya.Stanford.EDU, nilsson@tenaya, oliger@pride,
pratt@polya.Stanford.EDU
Cc: bscott@score
Subject: Faculty Repoirt
Message-Id: <CMM.0.87.597089999.chandler@polya.stanford.edu>
Whoops...that should be faculty "report".....was due 12/1. Please get it to
me as soon as possible. Thanks.
∂02-Dec-88 1129 weening@Gang-of-Four.Stanford.EDU
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 11:26:33 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01546; Fri, 2 Dec 88 11:25:09 PST
Date: Fri, 2 Dec 88 11:25:09 PST
From: Joe Weening <weening@Gang-of-Four.Stanford.EDU>
Message-Id: <8812021925.AA01546@Gang-of-Four.Stanford.EDU>
To: jmc@sail
From: crunch@well.UUCP (John Draper)
Newsgroups: comp.misc
Subject: More info from Moscow
Message-ID: <7736@well.UUCP>
Date: 22 Nov 88 22:42:54 GMT
Reply-To: crunch@well.UUCP (John Draper)
Organization: Whole Earth Lectroinic Link, Sausalito, CA
Lines: 102
Here is the first (of many) interesting things to come out of the
Soviet Union that are of interest to UseNet readers.
In December, a new computer club will be forming in the Soviet Union
called the International Computer Club. It is a non-profit organization
sponsored by Soviet Software companies. They are looking for American
Software organizations to sponsor the American side of the Club.
The Club is a Joint Venture arrangement between a Marin County
organization or Chapter and a Chapter on the Moscow side. On the
15th and 16th of December, in Moscow will be when the Club officially
becomes a reality. I was told that American software people will
be invited to it's "kick-off".
They want to create a number of Software "Journalists" as they call it,
to link up all the embassies and post a public information forum through
the use of BBS'es.
Soviet citizens of ALL walks of life are invited to become a member
and the costs are $15 monthly and $40 per year dues. The information
I have is currently sketchey at this time, as things are still being
set up. They will have an Email address on the SF/Moscow teleport,
but in order for Americans to use the Email setup, a written authorization
from the ICC would be necessary. I would suspect that the ICC would
want to get as many Americans on the Email link as possible.
The ICC would be building computer facilities for those not currently
posessing Personal Computers, and will have Modem equipment.
In Februrary, they are planning on sponsoring a Programming Skills
competition. I'm not sure if Americans will be invited to participate,
but they are going to be looking for American Judges to judge the contest.
Each of the Americans selected to become a judge, will be flown to
Moscow for a weeks worth of judging. This would give the Americans
a chance to evaluate the programming Skills of Soviet Programmers. This
should prove VERY INTERESTING INDEED.
They also want to establish a World Exibition Software library. Hmmm!!
(I wonder what kind of piracy opportunity this provides??). They want
to collect all the new software thats produced within the Soviet Union
and catalog it with American Products.
The Soviet winners of the Software Skills Competition will be allowed
to visit the USA and meet their American Software counterparts, tour
American Institutions, and companies.
I was invited to participate in this joint venture through the
Programmers Network. I suspect that other information exchange forums
would be built as a result of this. I was asked to help spread the
word and interest to American Programmers.
If anyone is interested in more information, Please Email me, and
I'll send your resuests for information on to Moscow, and relay the
information by posting it to the appropriate newsgroup. I suspect that
"comp.misc" might be appropriate, as the other newsgroups devoted to
the Soviet Union have little or no traffic. Perhaps we should set up
a new newsgroup called "sov.joint.ventures". But I have absolutly
NO idea how to create a new newsgroup.
My Link to the SF/Moscow service has been delayed yet another week,
pending permission and necessary paperwork from the Soviet side.
I also will be publishing an erratum or correction information on the
SF/Moscow teleport as soon as I recieve the corrections to some errors
and assumptions I made.
One such correction, is that the SF/Moscow teleport will NOT connect
an American institution to the network unless INVITED on by a Soviet
institution. Origionally, nothing was mentioned about that when I
phoned in my inquiry, nor was it mentioned on the Service agreement.
The Soviet Teleport user must issue a request in writing to the
Teleport offices and the National Center for Automated Data Exchange,
an organization the Soviet user must sign a seperate contract with, and
a duplicate letter must also be sent to the Moscow Teleport offices.
The letter should outline in detail the names of the American institutions
wishing to establish Email contact with the particular Soviet Teleport
user.
This serves 2 purposes:
1) CONTROL over who the Soviet can communicate with.
2) Prevents the American from paying for services that don't
connect to any specific recipiant.
I just recently recieved this information yesterday, and passing them
onto the many people who asked me about the Teleport Services.
Interested organizations can contact me at:
uunet!acad!well!crunch
Also, please indicate in your inquiry if you want me to pass it onto
Moscow. I understand that certain UseNet sites are not permitted to
communicate with the USSR because of their work, so by default, I will
NOT pass this information on to the USSR without explicitly requested
to do so.
John Draper
Programmers Network
∂02-Dec-88 1252 pehoushe@Gang-of-Four.Stanford.EDU CONS Contention
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 12:51:56 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA01909; Fri, 2 Dec 88 12:50:17 PST
Date: Fri, 2 Dec 88 12:50:17 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8812022050.AA01909@Gang-of-Four.Stanford.EDU>
To: rpg@Gang-of-Four.Stanford.EDU, arg@Gang-of-Four.Stanford.EDU
Cc: jmc@sail, clt@sail, weening@Gang-of-Four.Stanford.EDU,
pehoushek@Gang-of-Four.Stanford.EDU
Subject: CONS Contention
The Browse Benchmark only speeds up by a factor of 3 without
condition-memory and only by 5 with condition-memory, when it should
come close to 8//8. This is due to CONS contention. CONS contention
affects many other benchmarks, too, and it is not that hard to fix.
This was a problem in the initial implementation, and is still a
problem now. I'm not sure what your priorities are, but we (here at
Stanford, including Professor McCarthy) asked that fixing this bug be
moved to the top of your priority list a long time ago.
When will CONS contention be removed? -dan
∂02-Dec-88 1322 rpg@lucid.com CONS Contention
Received: from lucid.com by SAIL.Stanford.EDU with TCP; 2 Dec 88 13:22:06 PST
Received: from challenger ([192.9.200.17]) by heavens-gate.lucid.com id AA04764g; Fri, 2 Dec 88 13:19:50 PST
Received: by challenger id AA19011g; Fri, 2 Dec 88 13:15:59 PST
Date: Fri, 2 Dec 88 13:15:59 PST
From: Richard P. Gabriel <rpg@lucid.com>
Message-Id: <8812022115.AA19011@challenger>
To: pehoushe@Gang-of-Four.Stanford.EDU
Cc: jmc@sail.stanford.edu, clt@sail.stanford.edu,
weening@Gang-of-Four.Stanford.EDU, arg@lucid.com
Subject: CONS Contention
I apparently deleted the message from you that contained the proof
that CONS contention was the sole factor in Boyer not speeding up by a
factor of 8 on an 8 processor machine. Could you forward it to me?
This information will be useful when go to quantify what factors
influence performance, so it is important to retain it.
Thanks.
-rpg-
∂02-Dec-88 1358 pehoushe@Gang-of-Four.Stanford.EDU CONS Contention
Received: from Gang-of-Four.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 13:57:57 PST
Received: by Gang-of-Four.Stanford.EDU (5.59/25-eef) id AA02268; Fri, 2 Dec 88 13:56:21 PST
Date: Fri, 2 Dec 88 13:56:21 PST
From: Dan Pehoushek <pehoushe@Gang-of-Four.Stanford.EDU>
Message-Id: <8812022156.AA02268@Gang-of-Four.Stanford.EDU>
To: rpg@lucid.com
Cc: jmc@sail.stanford.edu, clt@sail.stanford.edu,
weening@Gang-of-Four.Stanford.EDU, arg@lucid.com
In-Reply-To: Richard P. Gabriel's message of Fri, 2 Dec 88 13:15:59 PST <8812022115.AA19011@challenger>
Subject: CONS Contention
I don't know what "proof" you are talking about. 8 out of 8? I can get
6.25 or so, using condition-memory, when the serial version takes 10.5
seconds. It's not clear how much cons contention affects this
program, but it is quite clear on others.
The program which demonstrates CONS contention most clearly is Hanoi,
although matrix multiply with floating point or bignums does a pretty
good job of showing lossage. The following program is Towers of
hanoi. It gets a speed-up of 1.8 out of a possible 8, when n=14.
Using good depth cutoffs, it might be possible to boost the speed-up
to a horrible 2//8.
;;; Serial
(defun hanoi (a b c n)
(cond ((= n 1) (cons a c))
(T (cons
(cons (cons a c)
(hanoi a c b (1- n)))
(hanoi b a c (1- n))))))
;;; Parallel
(defun phanoi (a b c n)
(cond ((= n 1) (cons a c))
(t #?(cons
(cons (cons a c)
(phanoi b a c (1- n)))
(phanoi a c b (1- n))))))
∂02-Dec-88 1356 P.EPSTEIN@GSB-WHY.Stanford.EDU re: Justifying Schultz's decision about Arafat
Received: from GSB-WHY.Stanford.EDU by SAIL.Stanford.EDU with TCP; 2 Dec 88 13:56:47 PST
Date: Fri 2 Dec 88 13:55:29-PST
From: David L. Epstein <P.EPSTEIN@GSB-WHY.Stanford.EDU>
Subject: re: Justifying Schultz's decision about Arafat
To: JMC@Sail.Stanford.EDU
In-Reply-To: <vJbw6@SAIL.Stanford.EDU>
Message-ID: <12451293043.44.P.EPSTEIN@GSB-WHY.Stanford.EDU>
Your point about Arafat is well-taken; it didn't even occur to me that the
agreement signed by the U.S. didn't apply to him.
-------
∂02-Dec-88 1557 perrie@sumex-aim.stanford.edu Call from Ed Feigenbaum
Received: from sumex-aim.stanford.edu by SAIL.Stanford.EDU with TCP; 2 Dec 88 15:57:02 PST
Received: by sumex-aim.stanford.edu (4.0/inc-1.0)
id AA00539; Fri, 2 Dec 88 15:57:25 PST
Date: Fri, 2 Dec 1988 15:57:24 PST
From: Michelle Perrie <perrie@sumex-aim.stanford.edu>
To: JMC@sail.stanford.edu
Subject: Call from Ed Feigenbaum
Message-Id: <CMM.0.88.597110244.perrie@sumex-aim.stanford.edu>
Ed called you today at about 3:45. He was at the Airport. He said that he
would call you again this weekend or on Monday.
∂02-Dec-88 1600 JMC
Pfeffer 201 794-8502
∂02-Dec-88 1706 tony%pembina.alberta.cdn@relay.ubc.ca 6th World Computer Chess Championship
Received: from relay.ubc.ca by SAIL.Stanford.EDU with TCP; 2 Dec 88 17:05:56 PST
Received: by relay.ubc.ca (5.59/1.14)
id AA08563; Fri, 2 Dec 88 17:08:49 PST
Date: 2 Dec 88 15:02 -0700
From: "T.Anthony Marsland" <tony%pembina.alberta.cdn@relay.ubc.ca>
To: John McCarthy <jmc@sail.stanford.edu>
Message-Id: <657*tony@pembina.alberta.cdn>
Subject: 6th World Computer Chess Championship
Just a test message to provide my preferred email address of
tony@pembina.alberta.cdn
Look forward to receiving your brief bio. for inclusion in our
advertizing literature. Thankyou for the abstract you sent earlier,
I don't know Kelly Aldrich, but assume that she is working for the
conference chairman Ted. Barnicoat with whom I am in contact.
Regards
Tony Marsland