perm filename E79.IN[LET,JMC]2 blob
sn#488062 filedate 1979-10-02 generic text, type C, neo UTF8
COMMENT ⊗ VALID 00214 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00021 00002 ∂02-Jul-79 1116 PAT
C00024 00003 a030 0134 02 Jul 79
C00028 00004 ∂01-Jul-79 1947 BIS Hand/Eye Personnel
C00030 00005 ∂02-Jul-79 1505 ARR Proposal for PASCAL*
C00031 00006 ∂02-Jul-79 1752 LGC PSOLFOL
C00032 00007 ∂02-Jul-79 1806 DEW thesis
C00033 00008 ∂03-Jul-79 1000 JMC*
C00034 00009 ∂03-Jul-79 1048 PJH via LONDON turner
C00037 00010 ∂03-Jul-79 1634 PAT concep
C00038 00011 ∂04-Jul-79 1258 WVC CAD research
C00041 00012 ∂05-Jul-79 0140 100 : mlb zork
C00042 00013 ∂05-Jul-79 1540 DEW Friday
C00043 00014 ∂05-Jul-79 2004 Mark Crispin <Admin.MRC at SU-SCORE> GSB status
C00044 00015 ∂06-Jul-79 0900 JMC*
C00045 00016 ∂06-Jul-79 0900 JMC*
C00046 00017 ∂06-Jul-79 1531 Pickering at SUMEX-AIM meeting
C00047 00018 ∂06-Jul-79 1532 ARK S-1 meeting notice
C00048 00019 ∂06-Jul-79 1656 HVA Bon Voyage
C00049 00020 ∂06-Jul-79 2123 HQM at MIT-ML (Henry Minsky)
C00053 00021 ∂08-Jul-79 0900 JMC*
C00054 00022 ∂09-Jul-79 2100 HPM ursula called, nothing urgent
C00055 00023 ∂09-Jul-79 2335 ME
C00056 00024 ∂10-Jul-79 2141 Vwiederhold at SUMEX-AIM Bastille day S-1 party.
C00057 00025 ∂11-Jul-79 1522 DBG
C00058 00026 ∂12-Jul-79 0904 CET
C00059 00027 ∂12-Jul-79 1119 LCA thanks
C00060 00028 ∂12-Jul-79 1338 GIO
C00061 00029 ∂14-Jul-79 0324 REM at MIT-MC (Robert Elton Maas) A.I.
C00063 00030 ∂15-Jul-79 1857 PWM via SUMEX-AIM termination of TAK
C00064 00031 ∂16-Jul-79 0915 Ijcai79 at SUMEX-AIM IJCAI-79
C00065 00032 ∂16-Jul-79 1051 DCO
C00066 00033 ∂17-Jul-79 1059 PN Version changer for Pascal programs
C00067 00034 ∂17-Jul-79 1243 BPM Net access from New York
C00068 00035 ∂17-Jul-79 1405 BH Dialnet
C00069 00036 ∂18-Jul-79 0157 MRC San Diego DECUS, December 10-13, 1979
C00071 00037 ∂19-Jul-79 0058 PMF
C00072 00038 ∂20-Jul-79 1051 100 : patte wexelblatt
C00073 00039 ∂20-Jul-79 1925 Rindfleisch at SUMEX-AIM PC BOARD LAYOUT SYSTEM DEMO
C00074 00040 ∂21-Jul-79 1705 Purger
C00075 00041 ∂22-Jul-79 1623 Purger
C00081 00042 ∂24-Jul-79 1151 ARK Reminder: S-1 Meeting
C00082 00043 ∂24-Jul-79 1521 CET David Yun
C00083 00044 ∂24-Jul-79 1639 RPG
C00084 00045 ∂25-Jul-79 0222 PTZ New PCPASC and SOPA (PCPZ09 , SOPZ17)
C00089 00046 ∂26-Jul-79 1216 BH DIALNET mail
C00092 00047 ∂26-Jul-79 1656 Ijcai79 at SUMEX-AIM IJCAI Proceedings
C00094 00048 ∂26-Jul-79 2120 BIS Robotics Expenditures
C00096 00049 ∂27-Jul-79 0722 Mark Crispin <Admin.MRC at SU-SCORE> Dialnet MAIL
C00099 00050 ∂27-Jul-79 0754 Mark Crispin <Admin.MRC at SU-SCORE> Sumex ''TTYFTP'' progress
C00101 00051 ∂27-Jul-79 0857 Mark Crispin <Admin.MRC at SU-SCORE> SUMEX TTYFTP implementation notes
C00121 00052 ∂27-Jul-79 2122 PEG New version of PCPASC
C00123 00053 ∂27-Jul-79 2217 ARR New VERCH up.
C00125 00054 ∂29-Jul-79 1443 ARR PEG ARR New VERCH up.
C00128 00055 ∂30-Jul-79 0000 Mark Crispin <Admin.MRC at SU-SCORE> having only Dialnet and local mail
C00129 00056 ∂30-Jul-79 1305 DCL Guest users at heavy load times
C00130 00057 ∂30-Jul-79 1358 BH Dialnet and no Arpanet
C00131 00058 ∂31-Jul-79 1021 PAT
C00132 00059 ∂01-Aug-79 1240 ALS Status of UCode SOPA
C00135 00060 ∂01-Aug-79 1519 CET Policy Information re Dissertations
C00136 00061 ∂01-Aug-79 2246 LES Books for Michaels
C00137 00062 ∂03-Aug-79 1039 MRC Sumex's TTYFTP program
C00143 00063 ∂03-Aug-79 1056 Yeager at SUMEX-AIM Re: Sumex's TTYFTP program
C00146 00064 ∂03-Aug-79 1115 DEW
C00148 00065 ∂03-Aug-79 1129 MRC TTYFTP
C00151 00066 ∂03-Aug-79 1956 DPB Zohar
C00152 00067 ∂04-Aug-79 0255 PTZ new SOPA - SOPZ18[DMP,S1]
C00153 00068 ∂04-Aug-79 1347 RWG via MIT-AI
C00154 00069 ∂04-Aug-79 1908 LGC 1. Important New Ideas 2. IJCAI
C00159 00070 ∂05-Aug-79 1029 PAT
C00160 00071 ∂05-Aug-79 1647 TM log area
C00162 00072 ∂06-Aug-79 1041 PAT Zohar
C00163 00073 ∂07-Aug-79 1115 ARK S-1 Meeting
C00164 00074 ∂07-Aug-79 2053 DCL AI Lab. cost breakdown
C00165 00075 ∂08-Aug-79 1119 CET Computers
C00166 00076 ∂08-Aug-79 1149 CET Classnotes for CS206
C00167 00077 ∂08-Aug-79 1321 HVA A.I. Lab Cost Breakdown
C00168 00078 ∂08-Aug-79 1920 REM via AMES-TIP
C00170 00079 ∂08-Aug-79 2035 DCL Comment on Jacks revision of August 1st.
C00172 00080 ∂09-Aug-79 0929 CET
C00173 00081 ∂09-Aug-79 1253 Evans@BBND(Art-Evans) Extra Sept meeting
C00176 00082 ∂10-Aug-79 0000 JMC*
C00177 00083 ∂10-Aug-79 1147 TW Access to the SAIL computer
C00182 00084 ∂10-Aug-79 1748 100 : REM via AMES-TIP Tech. Licensing / IRSM
C00184 00085 ∂14-Aug-79 1113 PN Introducing PAS, a pseudo-monitor (Away with all DO files!)
C00187 00086 ∂14-Aug-79 1512 Purger
C00188 00087 ∂14-Aug-79 1622 ARR Reminder: S-1 meeting
C00189 00088 ∂14-Aug-79 1936 BIS Hand/Eye Equipment
C00192 00089 ∂15-Aug-79 1500 LES AISA
C00194 00090 ∂16-Aug-79 1445 TOB
C00196 00091 ∂17-Aug-79 0043 BIS Languages
C00197 00092 ∂17-Aug-79 0922 DEW
C00198 00093 ∂17-Aug-79 1526 DPB Teaching assignments for 79-80
C00201 00094 ∂18-Aug-79 1419 TOB
C00202 00095 ∂20-Aug-79 1400 Bmoore at SRI-KL (Bob Moore) AI-Phil Workshop
C00204 00096 ∂20-Aug-79 1512 BOBROW at PARC-MAXC2 (Response to message)
C00205 00097 ∂21-Aug-79 2342 MARCF at MIT-MC (Marc K. Furon)
C00208 00098 ∂22-Aug-79 1244 LES
C00209 00099 ∂22-Aug-79 1302 ARR Test programs
C00210 00100 ∂23-Aug-79 1823 ''Richard Kovalcik, Jr.'' <RK at MIT-MC> forwarding this ...
C00211 00101 ∂23-Aug-79 1848 ME death of a record WAITS system
C00213 00102 ∂23-Aug-79 1907 Mark Crispin <Admin.MRC at SU-SCORE> All good things must come to an end...
C00215 00103 ∂24-Aug-79 1424 DPB office plans for Fall Quarter
C00216 00104 ∂24-Aug-79 2230 PEG Documents
C00217 00105 ∂25-Aug-79 1525 BIS DDs in Jacks Hall
C00218 00106 ∂25-Aug-79 1700 MARCF at MIT-MC (Marc K. Furon)
C00219 00107 ∂25-Aug-79 1703 MARCF at MIT-MC (Marc K. Furon)
C00220 00108 ∂25-Aug-79 1703 ARR File name extension for PASCAL* sources.
C00221 00109 ∂26-Aug-79 1353 Wiederhold at SUMEX-AIM Re: File name extension for PASCAL* sources.
C00222 00110 ∂26-Aug-79 1930 PEG Documents
C00224 00111 ∂27-Aug-79 1157 ARR Wiederhold Re: File name extension for PASCAL* sources.
C00226 00112 ∂28-Aug-79 0859 ARR S-1 meeting.
C00227 00113 ∂28-Aug-79 1813 ARR meeting postponed
C00228 00114 ∂29-Aug-79 0617 LLW No Work, No Eat!
C00229 00115 ∂29-Aug-79 1459 ARR LLW No Work, No Eat!
C00232 00116 ∂30-Aug-79 0906 PTZ SIGPLAN Compiler Conference Proceedings
C00233 00117 ∂30-Aug-79 1130 CET Student support for Autumn Quarter
C00234 00118 ∂31-Aug-79 1609 TOB
C00238 00119 ∂31-Aug-79 1755 CLT MTC Qual.
C00239 00120 ∂01-Sep-79 0759 PN Yet another S-1 area: [CSP,S1]
C00240 00121 ∂02-Sep-79 0000 CLT
C00241 00122 ∂02-Sep-79 1949 PN S-1 meeting
C00243 00123 ∂03-Sep-79 1810 HPM Space on $5 a day
C00244 00124 ∂04-Sep-79 0046 MRC via SU-SCORE HBTY
C00245 00125 ∂04-Sep-79 0942 MAX
C00246 00126 ∂04-Sep-79 1006 LJH Today!
C00247 00127 ∂04-Sep-79 1501 ROY on TTY107 (at TV-104) 1501
C00248 00128 ∂04-Sep-79 2326 Mark Crispin <Admin.MRC at SU-SCORE> news
C00251 00129 ∂05-Sep-79 0603 Mark Crispin <Admin.MRC at SU-SCORE> Proposal for HP
C00253 00130 ∂05-Sep-79 0859 g.eldre at SU-SCORE (Tim Eldredge) Re: Proposal for HP
C00254 00131 ∂06-Sep-79 1109 g.eldre at SU-SCORE (Tim Eldredge) Dialnet at HP
C00255 00132 ∂06-Sep-79 1232 100 : patte New graduate student
C00256 00133 ∂07-Sep-79 1128 PAT future plans
C00260 00134 ∂07-Sep-79 1301 Wiederhold at SUMEX-AIM Information on Stanfords S-1 efforts
C00262 00135 ∂07-Sep-79 1936 FB talk
C00264 00136 ∂08-Sep-79 2134 Mark Crispin <Admin.MRC at SU-SCORE> VADIC 1200 baud modems
C00267 00137 ∂09-Sep-79 0007 MFB
C00268 00138 ∂09-Sep-79 0104 Mark Crispin <Admin.MRC at SU-SCORE> Re: your message to Rindfleisch
C00270 00139 ∂09-Sep-79 0842 HPM SUCCESS (of sorts)
C00273 00140 ∂10-Sep-79 0121 CLT pub
C00274 00141 ∂10-Sep-79 0857 Rindfleisch at SUMEX-AIM Re: VADIC 1200 baud modems
C00277 00142 ∂10-Sep-79 1021 Wiederhold at SUMEX-AIM Destination for SMA-3 Manual and architecture problems
C00279 00143 ∂10-Sep-79 2045 SKU SARAH CALLED....
C00280 00144 ∂11-Sep-79 0905 LCA
C00281 00145 ∂11-Sep-79 0901 JK via SUMEX-AIM
C00284 00146 ∂11-Sep-79 1657 MFB seminar
C00285 00147 ∂13-Sep-79 1125 CET
C00286 00148 ∂13-Sep-79 1223 CLT cs206 ta
C00287 00149 ∂13-Sep-79 1254 DRA Article?
C00288 00150 ∂13-Sep-79 1541 Ijcai79 at SUMEX-AIM CS225 Projects
C00290 00151 ∂13-Sep-79 1543 DCO Yapp - Yet Another Pretty Printer
C00292 00152 ∂13-Sep-79 1609 MFB TALK
C00293 00153 ∂13-Sep-79 1607 CET
C00295 00154 ∂13-Sep-79 2007 FB via SU-SCORE Topics for faculty meeting
C00297 00155 ∂13-Sep-79 2036 WAXMAN at USC-ECL REQUEST FOR PAPER
C00298 00156 ∂13-Sep-79 2338 JMC
C00299 00157 ∂14-Sep-79 0232 Mark Crispin <Admin.MRC at SU-SCORE> Re: Re: VADIC 1200 baud modems
C00301 00158 ∂14-Sep-79 1203 RPG CASBS Seminars
C00302 00159 ∂14-Sep-79 1202 ARK S-1 Meeting
C00303 00160 ∂14-Sep-79 1230 LGC via SU-TIP Discussion of Plans
C00304 00161 ∂14-Sep-79 1508 MFB TALK
C00305 00162 ∂14-Sep-79 1649 100 : patte phone
C00306 00163 ∂14-Sep-79 2216 LGC Discussion Tomorrow?
C00307 00164 ∂14-Sep-79 2256 FB display terminals
C00308 00165 ∂17-Sep-79 1013 ROY on TTY15 1013
C00309 00166 ∂17-Sep-79 1243 PAT Lawrence Markosian
C00310 00167 ∂17-Sep-79 1312 CLT book
C00311 00168 ∂17-Sep-79 1555 MAX
C00312 00169 ∂17-Sep-79 1553 MAX
C00313 00170 ∂17-Sep-79 2000 JMC*
C00314 00171 ∂18-Sep-79 1546 DPB New student support
C00320 00172 ∂19-Sep-79 1004 MAX seminar at CASBS
C00322 00173 ∂19-Sep-79 1810 the tty of Geoffrey S. Goodfellow [Brian.Reid at CMU-10A: First mail system?]
C00324 00174 ∂20-Sep-79 1038 CET Lots
C00325 00175 ∂20-Sep-79 1201 LJH
C00326 00176 ∂20-Sep-79 1203 LES SU-TIP move
C00328 00177 ∂21-Sep-79 0000 JMC*
C00329 00178 ∂21-Sep-79 1124 LJH
C00330 00179 ∂21-Sep-79 1355 CET
C00333 00180 ∂21-Sep-79 1904 LES
C00336 00181 ∂21-Sep-79 1946 LES
C00338 00182 ∂22-Sep-79 1334 RMS at MIT-MC (Richard M. Stallman)
C00339 00183 ∂23-Sep-79 1740 TOB
C00340 00184 ∂24-Sep-79 0838 CET Reminder
C00341 00185 ∂24-Sep-79 1028 HEARN at UTAH-20 Herbert Stoyan
C00342 00186 ∂24-Sep-79 1058 LES Vadic modems
C00343 00187 ∂24-Sep-79 1225 PAT
C00345 00188 ∂24-Sep-79 1219 DCO
C00346 00189 ∂24-Sep-79 1357 MFB THANKS, AND SOME BUSINESS.
C00348 00190 ∂24-Sep-79 1609 CET Time of CS Faculty Meeting
C00349 00191 ∂24-Sep-79 1646 CSD.MILLER at SU-SCORE new PhD student
C00350 00192 ∂24-Sep-79 1736 LES Bit map displays
C00352 00193 ∂25-Sep-79 0230 Mark Crispin <Admin.MRC at SU-SCORE> Re: Vadic modems
C00355 00194 ∂25-Sep-79 1210 PAT
C00356 00195 ∂25-Sep-79 1210 DCO
C00360 00196 ∂25-Sep-79 1548 LES
C00361 00197 ∂26-Sep-79 0229 REM via AMES-TIP Multi-CPU computers
C00366 00198 ∂26-Sep-79 0248 REM via AMES-TIP
C00367 00199 ∂26-Sep-79 1242 100 : patte airfares
C00368 00200 ∂26-Sep-79 1248 LJH Caspian
C00369 00201 ∂26-Sep-79 1803 LES Display system discussion
C00370 00202 ∂26-Sep-79 2020 FB via SU-SCORE Display system discussion
C00371 00203 ∂26-Sep-79 2257 LES Display system
C00373 00204 ∂27-Sep-79 0836 FB displays
C00376 00205 ∂28-Sep-79 1241 LCA
C00377 00206 ∂28-Sep-79 1347 TOB Chinese visit
C00378 00207 ∂28-Sep-79 1607 John.Laird at CMU-10A Re: sex in wetsuit
C00379 00208 ∂28-Sep-79 1643 CET Colloq. Tuesday
C00380 00209 ∂28-Sep-79 1649 SEK Logic seminar.
C00381 00210 ∂28-Sep-79 2138 CLT
C00382 00211 ∂29-Sep-79 0904 JRA lisp conf
C00396 00212 ∂30-Sep-79 1000 JMC*
C00397 00213 ∂30-Sep-79 1508 EF at MIT-AI (Edward Fredkin)
C00398 00214 ∂01-Oct-79 1200 PAT Mr. Tang and Jacks Hall
C00400 ENDMK
C⊗;
∂02-Jul-79 1116 PAT
∂28-Jun-79 1731 JMC
Please check into the following: It occurs to me that it might be
advantageous for me to take a flight that stops in NY and leave some
baggage in NY, since I am not coming back to California till the end
of summer. Is there a reaonably good connection that will do it and
still get me to Edinburgh Monday evening? Does Hirsch know whether
I can check a suitcase or two at Kennedy for a week? Anyway, please
try to find out about it.
-----------
There is no problem in getting to Edinburgh Monday evening if you take
a SFO-NY overnight flight then catch a NY-London flight in the morning.
Concorde leaves at 10am and arrives at 6:45pm.
Unfortunately, there are NO storage facilities at Kennedy because of the
bombings a couple of years ago.
There are a couple of choices in getting luggage to NY. Each of them
except one involve additional cost.
1. Greyhound has a storage service. You can ship by them and they
will charge per day for holding it.
2. There is a storage company at Kennedy (Forman's) that picks up and
delivers. It is supposed to be fairly expensive.
3. You can air freight it. They will hold it till you or someone else
picks it up.
4. You can fly to Kennedy and have someone pick it up from you before
you take your flight to London.
∂02-Jul-79 0140 MRC Amuzin' story...
a030 0134 02 Jul 79
PM-Russian Carrier,400
By FRED S. HOFFMAN
AP Military Writer
WASHINGTON (AP) - The Soviet aircraft carrier Minsk, one of the
world's mightiest warships, had a confrontation the other day with
some pesky Chinese navy boats, government sources say.
The Russians won - not with their powerful missiles and bombers, but
with their firehoses.
So far as it is known, the only casualties were some thoroughly
drenched Chinese sailors.
According to U.S. government sources, who asked not to be
identified, the episode occurred in the East China Sea last week as
the 40,000-ton Minsk steamed northward through international waters
several hundred miles from the Chinese coast.
This is the first time a Soviet carrier has moved into the Pacific
and the Peking government obviously doesn't like it.
The sources said small Chinese vessels have been making repeated
passes around the Minsk since the carrier entered the South China Sea
June 17, accompanied by a Soviet cruiser, an amphibious ship and a
tanker.
At one point, several armed Chinese trawlers darted to within 50
yards of the Russian carrier, the sources said.
The tiny Chinese craft and their pop-gun weapons posed no threat to
the huge Soviet warship.
But their maneuvers caused the Russian captain to take hasty,
evasive action to avoid running them down.
Finally, the Minsk skipper apparently decided he had had enough.
Without warning, Soviet sailors unreeled hoses, aimed them at the
Chinese boats bobbing in the waves nearby and opened fire with
high-pressure streams of water.
U.S. observers report the tactic worked.
Earlier, in the Minsk's voyage, the sources said, two Chinese
hydrographic ships crisscrossed only about 500 yards ahead of the
Minsk, forcing the carrier into emergency measures.
In reply to signals from the Minsk, the sources said, the Chinese
captains ran up international code flags reading: ''Diver below'' and
other messages designed to confuse the Russian commander.
U.S. naval officers who served aboard American carriers during the
Vietnam War chortled when they heard of the Chinese capers.
They recalled that Soviet intelligence-gathering trawlers used to
cut across the bows of American carriers in the Gulf of Tonkin,
disrupting the launching and recovery of U.S. planes, and there wasn't
much they could do about it.
ap-ny-07-02 0434EDT
***************
∂01-Jul-79 1947 BIS Hand/Eye Personnel
To: DPB at SU-AI
CC: CET at SU-AI, TOB at SU-AI, LES at SU-AI, JMC at SU-AI
Denny,
Welcome back to the front!
After discussions with Carolyn and Les I have evolved the latest list of
Hand/Eye Personnel. This may be found as the last page of the file
PERSON.TXT[ADM,BIS].
Since (1) it seems probable that not all Hand/Eye people can be housed
in JAX Hall and (2) Data Disc terminals can not be available in whatever
annex is selected for housing JAX overflow, I have indicated on this
latest list of personnel those whose research in computer vision currently
requires virtually constant use of Data Discs. I hope that this will be
helpful in partitioning people among the available offices.
I am most interested in working with you as you evolve the floor plans
for housing our personnel. Call me when you will. I'd rather know what's
happening in advance than be surprised later.
Best wishes,
Barry
∂02-Jul-79 1505 ARR Proposal for PASCAL*
To: S1 at SU-AI
A preliminary copy of the final draft for the proposal for PASCAL* is available
in XGP form in COMPAS.XGP[1,JLH]. A copy exists in the S1 library (223 SAIL).
If you don't have access to SAIL, ask me to send you a copy.
Comments to JLH, please.
Armando.
∂02-Jul-79 1752 LGC PSOLFOL
To: JMC, RWW
I'll appreciate having your reactions to page 3 of PRBSLV.TXT[1,LGC].
∂02-Jul-79 1806 DEW thesis
To: JMC
CC: DEW
I want to get some signatures before you leave. Would you like a copy
of the final (or close to it) draft by Friday? This would not be much
of a problem. If so, would you like xgp output or two-sided stuff
from the xerox (which would make it half as thick)?
Dave
I prefer two-sided.
∂03-Jul-79 1000 JMC*
Call Scheerger
∂03-Jul-79 1048 PJH via LONDON turner
in answe to your queries on ray turner. He was a mathematicien
under Peter Landin, did phd on algebraic stuff. Now teaches computing
but is chiefly interested in philosophy and phil. logic, especially
counterfactual conditionals and semantics of tenselogic, and in
programming language semantics a la scott&strachey. He works with
Hans Kamp at London on the philosophy, and recently took the MA in
philosophy there, getting the first distinction for several years. He
is working on a doctoral thesis on counterfactuals: he wants the second
doctorate to get an eventual job teaching philosophy. He is very smart,
smarter than me anyway, and he reads a lot and does not bullshit in
discussions. He also has a fantastic library of phil. books: he is
trying to become independent of university libraries as insurance against
redundancy. He is a quiet guy and does not write a lot, but what does
apear is high quality: his stuff on counterfactuals is definitive, i
think.He is interested in "cognitinve science" in an essentially
amateur way, but his wide reading and technical competence make
him a very useful critic. He is genuinely interested in the philosophy
of mind but was i think scared off it by working in close association
with Wiggins. He was very interested in Dennetts talk here a while ago
and , it turned out, had actually read the relevant philosophy literature. unlike me.
I will send you a copy of his counterfactual paper.
hope this helps. Pat.
Please hold Turner counterfactuals paper for me at Edinburgh.
∂03-Jul-79 1634 PAT concep
The file with the differences is called CONCEP.COM[F78,JMC].
SRCCOM also will merge two files which might or might not be the right thing
to do in this case.
∂04-Jul-79 1258 WVC CAD research
To: JMC at SU-AI
CC: csl.wvc at SU-SCORE
Forrest Baskett and Ed Feigenbaum mentioned to me that you are
interested in CAD for VLSI.
Since I have been working on this for the last 10 years I would like to
get together and talk with you.
Are you free next week Moday or Wednesday ?
Bill vanCleemput.
∂05-Jul-79 0140 100 : mlb zork
You get to the Bank by going W from the Gallery
∂05-Jul-79 1540 DEW Friday
I want to find you Friday to get signatures and give you my thesis.
When and where will you be available for a few minutes?
∂05-Jul-79 2004 Mark Crispin <Admin.MRC at SU-SCORE> GSB status
Date: 5 Jul 1979 1853-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: GSB status
To: JMC at SU-AI, Les at SU-AI
GSB is down hard and has been for several days. So my vacation came
at an ideal time. They are hoping to be up at the beginning of next
week. SCORE is still having problems with the new memory, so moving
Dialnet to SCORE won't help much, if any.
-------
∂06-Jul-79 0900 JMC*
glasses
∂06-Jul-79 0900 JMC*
money, tickets, appraisal
∂06-Jul-79 1531 Pickering at SUMEX-AIM meeting
To: S1 at SU-AI
Date: 6 Jul 1979 1522-PDT
From: Pickering at SUMEX-AIM
Subject: meeting
To: s1 at SAIL
cc: wiederhold
s1 meeting Thursday July 12 at 1:30 in the sail conference room
jayne for ark
-------
∂06-Jul-79 1532 ARK S-1 meeting notice
To: S1 at SU-AI
CC: pickering at SUMEX-AIM
There will be an S-1 meeting on Thursday, 12 July at 1:30 pm in the Sail
Conference Room. Note change in date.
Arthur
∂06-Jul-79 1656 HVA Bon Voyage
If I don't talk to you before you leave, have a marvelous trip--and send us
a post card!
∂06-Jul-79 2123 HQM at MIT-ML (Henry Minsky)
Date: 6 JUL 1979 2008-EDT
From: HQM at MIT-ML (Henry Minsky)
To: jmc at SU-AI
i met margaret around 1955 when we were both undergraduates at M.I.T.
She had just decided to switch to a physics major from chemistry and we
began to be in many classes together. (Margaret graduated with a degree
in both physics and chemistry.)
Some of you may remember what society was like 25 years ago,and most
relevant ,what was expected of females. To be "feminine" was to be pretty,
less intellegent than any amle who happened to be in your presence,and above
all else to devote all your time to catching a husband. Even atM.I.T. most
women willingly accepted this role. margaret did not. She was an
unusually independent and creative person.
Margaret had many intersts. At the time she was naturally most involved with
the study of physics,but she loved the world and exploring itand often
often found time to go sailing,swimming,kayaking,hiking,flying and rock climbing.
Folk dancing was a great passion of hers. Margaret was a fine craftsman
and among other things was noted for being an especially skilled glassblower.
She had an outlook on life that I think is typical of M.I.T. She believed
in control of emotion,in the application of reason to all phases of existence,to the sophisticateduse of technology to solving problems,and she loved
science. She could always be relied on to be infullcontrol in dangerous
situations - like when we began to be surrounde by thunderheads when flying
in a small plane. I have a problem with vertigo and tend to avoid heights,but
at margaret7s urginhg I went rock climbingwith her. I had full confidence
that if Margaret chose the equipment,tied the knots,and held the rope
everything would be in order andsafe if she said it was.
margaret and I shared many things at a sensitive time inour lives.
Ilearned to rely on and trusther strenght andcontrol. I think in me and her
other friends at M.I.T. she found understanding.
∂08-Jul-79 0900 JMC*
Emma opus
∂09-Jul-79 2100 HPM ursula called, nothing urgent
∂09-Jul-79 2335 ME
To: S1 at SU-AI
You are exceeding your disk quota.
Files that occupy space beyond your quota are subject to purging.
If you don't delete some of your files, the purger will.
Your disk quota is: 9000
Your files occupy 10802
∂10-Jul-79 2141 Vwiederhold at SUMEX-AIM Bastille day S-1 party.
To: S1 at SU-AI
Date: 10 Jul 1979 2129-PDT
From: Vwiederhold at SUMEX-AIM
Subject: Bastille day S-1 party.
To: s1 at SAIL
We have not had responses from you all., including some that I believe
will come. Please let us know so a headcount can lead toa foodcount.
Thanks Voy and Gio.
-------
∂11-Jul-79 1522 DBG
It seems that the move to have Binford approved as my principal
dissertation advisor has failed, unless Binford gets promoted in time.
Therefore, Denny Brown says that I should make out a new G81 form with
you as my official advisor. Do you approve of this? If so, I will need
to have you sign the form some time.
Also, I want to schedule my orals for some time in October or
early November. What dates during that time will you be available?
Don Gennery
∂12-Jul-79 0904 CET
To: JMC, PAT
∂11-Jul-79 1452 CET MTC Qual Sign up
To: "@PHD.DIS[1,CET]" at SU-AI
The MTC Qual will be held Autumn Quarter. Information
and sign-up list can be found in
Qual.MTC[1,cet]
∂12-Jul-79 1119 LCA thanks
Les told me that the Lab. is going to support me next year. I am very pleased.
Thank you very much for your consideration.
I am going back to Italy from July 30th till August 27th. I wish you I nice stay
at Yorktown. See you and, once again, thanks a lot, Gigina
∂12-Jul-79 1338 GIO
To: S1 at SU-AI
s1 meeting is beginning now. gio
∂14-Jul-79 0324 REM at MIT-MC (Robert Elton Maas) A.I.
Date: 14 JUL 1979 0623-EDT
From: REM at MIT-MC (Robert Elton Maas)
Sent-by: ←←←015 at MIT-MC
Subject: A.I.
To: JMC at SU-AI
I think I've found a hill-climbing algorithm in a bug.
(Not a bug in a hill-climbing algorithm.)
For the past half hour or more a small brown two-segment bug about
1/4 inch long and slender mostly sort-of has been running around the top
edge of a plastic cover to a plastic macaroni-salid container that
is upside-down on sink and slightly tilted. It seems to be trying to
find the top of the world by running along the top edge for about
3 revolutions then giving up and running ther other direction for
a while. It almost never leaves the sharp edge, but occasionally
goes into the main part of the inside of the lid then shortly returns
to the sharp edge which is higher.
Well, it's still there, running CCW at the moment, as I finish this msg.
∂15-Jul-79 1857 PWM via SUMEX-AIM termination of TAK
John - apparently J Moore has been able to prove the termination of TAK.
He is sending me a memo with details - when i get it I will try it in the
Pascal Verifier.
Peter
Nozaki has proof for real numbers. RWW has copy of Nozaki memo.
∂16-Jul-79 0915 Ijcai79 at SUMEX-AIM IJCAI-79
Date: 16 Jul 1979 0914-PDT
From: Ijcai79 at SUMEX-AIM
Subject: IJCAI-79
To: jmc at SU-AI
Will you be able to chair a technical session at IJCAI-79??
PLEASE RESPOND IMMEDIATELY, as we are prining the schedule now.
Thank you, The IJCAI-79 Program Committee
-------
∂16-Jul-79 1051 DCO
What is the most uptodate reference to your paper where you discuss
the decidability of the first order theory of lisp functions and show that
adding subtree makes it undecidable? Thanks.
The McCarthy and Cartwright AI memo refers correctly to the printed
paper.
∂17-Jul-79 1059 PN Version changer for Pascal programs
To: S1 at SU-AI
... is up. See VERCH.DOC[DOC,S1] for more details.
∂17-Jul-79 1243 BPM Net access from New York
To: JMC
CC: PAT
There isn't a TIP in New York City--nor at Rutgers anymore. The closest
TIPs to Yorktown Heights are RADC-TIP, Rome Air Development Center,
Griffiss AFB, Rome, NY, (315) 339-1400; MIT-TIP, Cambridge, MA, (617)
491-1150; and CORADCOM-TIP, Fort Monmouth, NJ 07703, dialup number not
known. If you decide that CORADCOM-TIP is the one to use, MRC or GFF
might know a number. Otherwise, use the official channel and contact the
TIP Liaison, Joseph Arnold, ARPAnet address DRDCO-AM@OFFICE-2, (201)
544-2527. Good luck.
∂17-Jul-79 1405 BH Dialnet
The status of Dialnet mail is that I am in the middle of writing
the user program (as part of the existing MAIL program). MRC is
working on the server. Naturally we both have to be finished
before either can really debug. MRC has to write the routines
for the new multi-network host table, and probably won't have
that done for a while; in the meantime we plan to use the
SAIL-to-SAIL link for debugging and avoid the host table problem.
∂18-Jul-79 0157 MRC San Diego DECUS, December 10-13, 1979
To: JMC at SU-AI, LES at SU-AI, Admin.Gorin at SU-SCORE
I have made a commitment to present two formal papers at this DECUS. One
involves research work I did as an undergraduate before coming to Stanford,
the other, entitled DIALNET - A TELEPHONE NETWORK DATA COMMUNICATIONS PROTOCOL,
is (of course) about Dialnet. The latter paper will be a half-hour
presentation, and may be published in the official DECUS Proceedings.
I assume there will be no problem getting me to this DECUS, and I would like
some assistance in the preparation of the Dialnet formal paper. I intend to
discuss the following points:
. the overall Dialnet philosphy
. the initial design goals
. the actual design, including specifics of the protocols
. the implementations
. the future
Mark
∂19-Jul-79 0058 PMF
To: S1 at SU-AI
I copied all of the S1 areas to the new S1 udp. Files that are not
actively being used should be deleted. (We are three million words over
allocation.) If this doesn't work, we'll just get purged anyway.
∂20-Jul-79 1051 100 : patte wexelblatt
called. He never received the graphic examples from your LISP talk, to go
along with the manuscript. I made a search but found nothing. He said
you had two or three slides that illustrated points in your talk and wants
to include them. Any idea where they might be?
∂20-Jul-79 1925 Rindfleisch at SUMEX-AIM PC BOARD LAYOUT SYSTEM DEMO
Date: 20 Jul 1979 1923-PDT
From: Rindfleisch at SUMEX-AIM
Subject: PC BOARD LAYOUT SYSTEM DEMO
To: FB at SU-AI, ADMIN.GORIN at SU-SCORE, JMC at SU-AI,
To: LES at SU-AI
DEC tells me there will be demos of a commercially available PC board
layout system running on VAX next Monday and Tuesday at DEC Santa Clara.
This system is by Scientific Calculations, Inc. from <I forget where> in
the east. Contact your favorite PDP-10 sales person to make arrangements
to see it.
Tom R.
-------
∂21-Jul-79 1705 Purger
To: S1 at SU-AI
You are exceeding your disk quota.
Files that occupy space beyond your quota are subject to purging.
If you don't delete some of your files, the purger will.
Your disk quota is: 9000
Your files occupy 12593
∂22-Jul-79 1623 Purger
To: S1 at SU-AI
You have exceeded your disk quota.
The files listed below have been purged to reduce your disk
area to your quota of 9000
Before purging, your files occupied 12647
SOSG16.PAS[SOP, S1]
SOPZ11.PAS[SOP, S1]
OLDDIR[SOP, S1]
SOPW15.PAS[SOP, S1]
SOSG14.DIF[SOP, S1]
SOPS09.PAS[SOP, S1]
SOPG14.PAS[SOP, S1]
DEMO.FOR[SOP, S1]
SOPZ17.PAS[SOP, S1]
DEMO2.FOR[SOP, S1]
SOIG16.LDI[SOP, S1]
SOIG16.PS1[SOP, S1]
SOIG16.LST[SOP, S1]
SYMTBL[SOP, S1]
SOIG16.PCO[SOP, S1]
QRR[SOP, S1]
SOIG16.PAS[SOP, S1]
SOPG16.PAS[SOP, S1]
PCPG08.PAS[PCP, S1]
PCSG08.PAS[PCP, S1]
PCPW07.PAS[PCP, S1]
PCSW07.PAS[PCP, S1]
PCIG08.PAS[PCP, S1]
PCPG06.PAS[PCP, S1]
T5.UCO[FOR, S1]
NUM.PAS[FOR, S1]
UCFOR1.PAS[FOR, S1]
T6.PUT[FOR, S1]
SOPBUG.FOR[FOR, S1]
PCFOR1.PAS[FOR, S1]
TT1.FOR[FOR, S1]
R2.PAS[FOR, S1]
T4.FOR[FOR, S1]
T11.FOR[FOR, S1]
TRIAL.UCO[FOR, S1]
T12.FOR[FOR, S1]
T6.FOR[FOR, S1]
T7.FOR[FOR, S1]
CUTOFF.PAS[FOR, S1]
T8.FOR[FOR, S1]
T13.FOR[FOR, S1]
TST.LST[FOR, S1]
PFORR.DO[FOR, S1]
SOPA.DO[FOR, S1]
RSIM.DO[FOR, S1]
T10.FOR[FOR, S1]
TR.UCO[FOR, S1]
TST.PCO[FOR, S1]
T9.FOR[FOR, S1]
TST.UCO[FOR, S1]
R1.PAS[FOR, S1]
QQ1.PAS[FOR, S1]
T5.PCO[FOR, S1]
X.FOR[FOR, S1]
TRIAL.F4[FOR, S1]
QQ1.PCO[FOR, S1]
PAS6.DO[FOR, S1]
QQ1.UCO[FOR, S1]
TODO[FOR, S1]
T0.FOR[FOR, S1]
T1.FOR[FOR, S1]
T2.FOR[FOR, S1]
T3.FOR[FOR, S1]
TST.PAS[FOR, S1]
PCFOR0.PAS[FOR, S1]
TR.FOR[FOR, S1]
PTRAN.PAS[FOR, S1]
Q2.LST[FOR, S1]
Q2.PCO[FOR, S1]
Q1.LST[FOR, S1]
TT5.FOR[FOR, S1]
TT5.LST[FOR, S1]
FOR.DO[FOR, S1]
TT5.UCO[FOR, S1]
Q1.UCO[FOR, S1]
X3.FOR[FOR, S1]
INPUT[FOR, S1]
X1A.LST[FOR, S1]
OUTPUT[FOR, S1]
PFOR.DO[FOR, S1]
FUSE.DO[FOR, S1]
X4.FOR[FOR, S1]
SOPAF6.DMP[FOR, S1]
PASPCO.DO[FOR, S1]
Q1.PCO[FOR, S1]
TRIAL.FOR[FOR, S1]
TPU.DO[FOR, S1]
PTRANS.PAS[FOR, S1]
PRR[FOR, S1]
PAS.DO[FOR, S1]
PTRANS.DMP[FOR, S1]
FOR01.DAT[FOR, S1]
X1.LST[FOR, S1]
T5.FOR[FOR, S1]
UCFOR2.PAS[FOR, S1]
X1.FOR[FOR, S1]
PAS2.DO[FOR, S1]
X2.FOR[FOR, S1]
UAS.DO[FOR, S1]
Q1.PAS[FOR, S1]
Q2.UCO[FOR, S1]
Q2.PAS[FOR, S1]
PRD[FOR, S1]
TR1.FOR[FOR, S1]
SYMTBL[FOR, S1]
X1.PCO[FOR, S1]
X2.LST[FOR, S1]
QRD[FOR, S1]
FORR.DO[FOR, S1]
TR1.LST[FOR, S1]
X2.PCO[FOR, S1]
X3.LST[FOR, S1]
TR1.UCO[FOR, S1]
X3.PCO[FOR, S1]
SEVAL.FOR[FOR, S1]
X1A.FOR[FOR, S1]
QRR[FOR, S1]
X1.LDI[FOR, S1]
OUT[FOR, S1]
PPSAV.TMP[FOR, S1]
PCROSS.DOC[DOC, S1]
RUNFOR.DOC[DOC, S1]
EXS78.PUB[DOC, S1]
ALLDOC.PUB[DOC, S1]
LP2[DOC, S1]
TEMP[DOC, S1]
MULT1.PUB[DOC, S1]
PDS78.PUB[DOC, S1]
S1CALL[DOC, S1]
S1ADDR.COM[DOC, S1]
SCA78.PUB[DOC, S1]
DRW78.CMD[DOC, S1]
LIS78.TXT[DOC, S1]
PASDOC.MSS[DOC, S1]
SOPADO.ABS[DOC, S1]
TID78.TXT[DOC, S1]
MAIL.FOR[DOC, S1]
PASMAC[DOC, S1]
PAIL5.ALS[DOC, S1]
MACBOD.DRW[DRW, S1]
ECLF1.DRW[DRW, S1]
ECLG1.DRW[DRW, S1]
ECLG2.DRW[DRW, S1]
ECLM1.DRW[DRW, S1]
ECLM2.DRW[DRW, S1]
ECLM3.DRW[DRW, S1]
ECLMEM.DRW[DRW, S1]
FLOW.DRW[DRW, S1]
LIBDRW.CMD[DRW, S1]
QREPL.BLI[FIN, S1]
CURSOR.BLI[FIN, S1]
FINE.DOC[FIN, S1]
REST.MID[FIN, S1]
CALIAD.SAV[FIN, S1]
FILESC.BLI[FIN, S1]
RFINE.MID[FIN, S1]
RFINE.REL[FIN, S1]
RFINE.DMP[FIN, S1]
DELETE.ME[FIN, S1]
REDISP.BLI[FIN, S1]
WRKPPR.HWC[AMB, S1]
ADN1.TXT[AMB, S1]
PASCAL.RVW[AMB, S1]
ADN1.UND[AMB, S1]
ADN1.CMT[AMB, S1]
ADN1.SWL[AMB, S1]
MEMOS.LST[AMB, S1]
S1DESC.LSP[AMB, S1]
DESC.TXT[AMB, S1]
ASM.TXT[AMB, S1]
ADN4[AMB, S1]
SOPG14.DMP[DMP, S1]
PLOT.DMP[DMP, S1]
SOPAW5.DMP[DMP, S1]
RIX.DMP[DMP, S1]
PCPASX.DMP[DMP, S1]
SOPAW4.DMP[DMP, S1]
PCPW5.DMP[DMP, S1]
PCPW4.DMP[DMP, S1]
SOPAE1.DMP[DMP, S1]
SOPAW6.DMP[DMP, S1]
∂24-Jul-79 1151 ARK Reminder: S-1 Meeting
To: S1 at SU-AI, pickering at SUMEX-AIM
There will be an S-1 meeting tomorrow, Wednesday, 25 July at 1:30 pm at the
SAIL small conference room.
Arthur
∂24-Jul-79 1521 CET David Yun
To: JMC, LES
David Yun came to see me about his AI Lab account.
His account was closed and he was moved to heaven.
I put him back in People.dat, and told him I would
sndmsg to Lester relaying his request to have his account
remain open. He will talk to John at IBM about it.
Carolyn
∂24-Jul-79 1639 RPG
To: "@FAC.DIS[1,DPB]" at SU-AI
∂23-Jul-79 CSL.OWICKI at SU-SCORE (Susan Owicki) 24-Jul-79 BTH Apartment for rent, Sept - Dec 1979
Date: 23 Jul 1979 1537-PDT
From: CSL.OWICKI at SU-SCORE (Susan Owicki)
Subject: Apartment for rent, Sept - Dec 1979
To: BBoard at SU-SCORE, BBoard at SU-AI, BBoard at SUMEX-AIM, BBoard at SRI-KL,
To: BBoard at PARC-MAXC
2BR/ 2BA condominium on Stanford campus. Furnished, all-electric kitchen, washer
/drier, pool, jacuzzi. $650/mo. No children or students allowed. Exact dates
are negotiable. Mail CSL.Owicki @ Score or call 497-1075 (day) 329-1867 (eve).
∂25-Jul-79 0222 PTZ New PCPASC and SOPA (PCPZ09 , SOPZ17)
To: S1 at SU-AI
This message is intended to reach S-1 Pascal or Fortran users or implementors;
if you are neither then you needn't read further. Implementors: more info
exists in SOPA[LOG,S1]/2P.
Brief overview of the differences over PCPG08 and SOPG16:
I got tired of not being able to tell what version of PCPASC or SOPA wrote
what .PCO, .PS1, or .LDI file, particularly after copying the files around or
restoring them from a purge (the directory WRITER info gets clobbered). So
PCPZ09.DMP[DMP,S1] writes its name and creation date on a "comment" line
(beginning with ' LOC 0 ') which is the first line in the output .PCO file.
SOPZ17.DMP[DMP,S1] writes its name and creation date immediately after the
START OF STATISTICS line in the .PS1 file. If the assembly code is being
printed [(*$A+*) in a Pascal source], it is also printed as a comment on the
2nd line of output for each procedure. The same information is written as
"comments or insults" on each EOM (End Of Module) record in the .LDI file.
This doesn't help with the problem of knowing which versions of SOPA are
compatible with which versions of PCPASC, but at least you can tell who wrote
which file.
SOPZ17 also has new peephole optimization: 9 new forms of MOV-collapsing
were added; and INC, SKP are collapsed to ISKP (INC, JMPA to IJMPA; DEC, JMPA
to DJMPA; etc.). Yet another flag parameter has been added (the 7th) to the
BGN P-instruction to control INC-SKP collapsing. If the parameter is 0,
collapsing is suppressed; if it is 1 or absent, collapsing will be attempted.
Peephole optimization statistics printed in the .PS1 file are now more detailed.
PCPZ09 compatibility: SOPZ17, SOPG16, or SOPW15
SOPZ17 compatibility: PCPZ09, PCPG08, or PCPW07
PCFOR1 (rewrite and reset done in Fortran runtimes)
PASRUN runtimes
[ for those of you who keep .PCO files around...
PCPG06 if no files are rewritten or reset
PCPE04 or PCPE03 if above and no sets are used ]
Do files on [DO,S1] have been updated to reference PCPZ09[DMP,S1] and
SOPZ17[DMP,S1].
STATISTICS FROM TESTING (STATIC):
---------------------------------
Programs tested: S1PZ16.PAS, SIMPLE.FOR, HYDRO.PAS, FFT.PAS, T5.FOR, SINE.PAS
Size in instrs: 28523 4251 3655 412 286 146
Avg percentage of total instrs removed: 11.57%
Avg increase in instr removal over SOPZ12: 2.59%
Avg increase in compilation time over SOPZ12 is close to 2.5% (some data missing)
STATISTICS FROM TESTING (DYNAMIC):
----------------------------------
Program SINE.PAS computes and prints 2 cycles of a sine curve on a 51x51 grid.
improvement
dynamic (FSIM ICOUNT) static (SOPA instr count)
SOPZ17 over no optimization 11.20% 18.49%
SOPZ17 over SOPZ12 5.81% 4.11%
SOPZ12 over no optimization 5.72% 14.38%
∂26-Jul-79 1216 BH DIALNET mail
To: JMC at SU-AI
CC: LES at SU-AI, ME at SU-AI, MRC at SU-AI, BH at SU-AI
Here is the state of things: I have written code in MAIL, and
MRC has written a server, so that we can now send dialnet mail
over the debugging link. The following things stand in the way
of a real dialnet mail service:
1. At present, MAIL is structured so that each message requires
a separate phone call, even if several are for the same host,
or even the same message to several users at one host. Fixing
that will be a significant task for me.
2. MRC has to write the NETWRK code for the multi-net host table.
3. I haven't addressed the issue of night mail to take advantage
of cheaper phone rates.
4. MRC and I disagree about how the server should handle incoming
mail to a nonexistent person. The Arpanet server checks the
validity of the recipient before accepting the message. I want
the Dialnet server to do the same; MRC wants to accept mail for
Anything that would optimize the probability and speed of
getting mail between two different computers would be of
greater import.
Right now we have the pessimal compromise: the server accepts the
mail, and the mail program deletes it without a trace!
anyone and have invalid mail dealt with later by human intervention.
The problems you list as unsolved sem to me to be minor.
∂26-Jul-79 1656 Ijcai79 at SUMEX-AIM IJCAI Proceedings
Date: 26 Jul 1979 1656-PDT
From: Ijcai79 at SUMEX-AIM
Subject: IJCAI Proceedings
To: IJCAI-79 Attendees:
INTERNATIONAL JOINT CONFERENCES ON ARTIFICIAL INTELLIGENCE
PUBLICATIONS
IJCAI-79 PROCEEDINGS WILL BE AVAILABLE SEPTEMBER 1979.
IJCAI-79, Tokyo, Japan, 20-24 August 1979. 2 vols.
Copies available for $25.00, payment in advance, book rate;
add: $5 for 1st class - US, Canada, Mexico; $20 air parcel
post elsewhere. Calif. residents add sales tax. Make check
payable to
Order from: IJCAI-79
Computer Science Dept.
Stanford University
Stanford, CA 94305 USA
PRE-PAID ORDERS ONLY. THANK YOU.
-------
∂26-Jul-79 2120 BIS Robotics Expenditures
I just sent this paragraph out. Is it substantially correct?
"Regarding plugging into human sensory and motor neurons, this could be
done reasonably efficiently within five years given a dedicated research
program. After all, we took atomic energy from pure theory to a bomb
in only six years! At the current rate of spending on robotics research,
we will reach the Manhattan Project total expenditure in about a century
if we continue our current pace."
I don't agree with the statement. Besides money it took the conviction
of the American physics community that this was the most important thing
to work on. If the country's smartest people were dedicated
personally to connecting to human sensory and motor organs, it
might be done in a few years with only modest increases in funding.
I guess connecting to human sensory and motor neurons in a simple
way is a straightforward problem that could be solved in a fixed
time by dedicated effort. Other robotics problems, such as the
ability to skin a cat, require conceptual advances that may be
as hard as the theory of the nucleus as developed between
Rutherford's work early in the century and the discovery of
fission in 1939.
∂27-Jul-79 0722 Mark Crispin <Admin.MRC at SU-SCORE> Dialnet MAIL
Date: 27 Jul 1979 0722-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Dialnet MAIL
To: JMC at SU-AI, BH at SU-AI, Les at SU-AI
As I have repeatedly stated, my opinion is that all mail should be
accepted, and that unknown mail should be kept in a dead letter box
for manual forwarding by an "operator" (or system programmer, or
whatever). Many sites do this: MIT, ISI. Especially with Dialnet
mail, I consider it to be highly objectionable to place a cross-
country call and have the message rejected because an account ID has
a minor misspelling, when a reasonable human being could cause the
message to get through!
There is another matter of greater concern to me. I consider it to
be bad design and extremely fragile to have several subsystems which
know how to do a task that belongs to one. In particular, I feel that
only one subsystem should know how to validate mail, only one should
know how to deliver it, etc. Everybody else should only know how to
put requests in. Every time something new like forwarding is put in,
every program has to be changed to know about it otherwise. And they
all are slightly different in some way; I can imagine the Arpanet
mail server rejecting an address which remind would have accepted.
The current implementation of mail on Tops-20 and Tenex is an outstanding
case in point of how fragile it can get. I have been busy in a rewrite
of Tops-20 mail, and the main intent is to make it a lot more robust. In
the standard configuration, no fewer than 4 processes know how to deliver
mail, and 5 know how to validate it!
-------
Remember that sites sholuld be able to use dialnet that have only
dialnet and local mail.
∂27-Jul-79 0754 Mark Crispin <Admin.MRC at SU-SCORE> Sumex ''TTYFTP'' progress
Date: 27 Jul 1979 0753-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Sumex "TTYFTP" progress
To: JMC at SU-AI, Les at SU-AI
They have TTYFTP up between Sumex and Rutgers, and are trying to get it
up on RT11 and RSX-11 for the PDP-11. They're having lots of fun and
laughter there because those operating systems are the pits.
On a 1200 baud link, they get about 800 baud with 54 byte buffers, but
start losing when they increase to 80 byte buffers. We're still going
to try a Dialnet/TTYFTP link, but my 256 byte buffers would really zap
it.
-------
∂27-Jul-79 0857 Mark Crispin <Admin.MRC at SU-SCORE> SUMEX TTYFTP implementation notes
Date: 27 Jul 1979 0854-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: SUMEX TTYFTP implementation notes
To: JMC at SU-AI, Les at SU-AI, IAZ at SU-AI
TTYFTP - a brief description
------------------------------
INTRODUCTION:
TTYFTP is a MAINSAIL program designed to transfer data between
computers over ordinary phone lines. All transmitted data is packeted,
checksummed, and errored packets are retransmitted. Currently, TTYFTP runs
as a job at both the HOST and USER sites, and is consequently subject to
whatever load constraints these sites choose to impose. It takes MONITOR
input and output buffer size into consideration and will not allow these
buffers to be overrun. The program at the HOST site is called SERFTP.
After initialization TTYFTP(SERFTP) is composed of three modules:
the scheduler, SCHED, the file transfer protocol handler, FTP, and the packet
handler, PH.
SCHED is the heart of the the program and controls all of the
data flow between the FTP and PH, and guarantees frequent entries to the PH
module so that the USER/SERVER PHs can communicate with each other.
At the expiration of a given demand response interval, DRI, if the
PH has NOT SENT a packet, it will send a NOP packet to alert the listening
PH that it is alive, and waiting to send/receive data. This communication
is continuous, and appropriate messages are relayed to the USER at 1 minute
intervals if the HOST has failed to send a packet. The DRI is currently
set to 30 seconds. If the HOST has not been heard from for 15 minutes, the
USER program will abort. The HOST has the same time out.
The FTP manages all high level protocols. It is the only module that
understands the notion of a file, or command. The user FTP opens a
FTP-channel between itself and the HOST FTP, and sends data
over this channel. Neither SCHED, nor PH is aware of the kind of data that
is being sent over the FTP channel; just that there is a channel open, and
data can be sent on this channel if requested. There can be up to eight
FTP-channels per line, and each FTP can have a maximum of eight sub-channels
per FTP-channel. Currently, only 3 sub-channels are used:
1. Channel 0 is used for FTP commands,
2. Channel 1 is used for RECEIVING data from SERFTP,
3. And, channel 2 is used for SENDING data to SERFTP.
These conventions were adopted to maintain compatibility with
DIALNET a la Mark Crispin.
In the simplest context we have something like the following:
USER - site HOST - site
consule
FTP <--> SCHED <---TTY line---> User FTP <--> SCHED
↑\ ↑
| \ |
| \ |
v \ data v
PH <-modem---- TTY line ------modem-> PH
!OPERATING PROCEDURE:
TTYFTP starts SERFTP at the HOST site as a job. If login sequences
are required they may either be read from a file, or requested from the
user at runtime. Currently, on SUMEX, the remote username and password are
prompted for by the program, a login sequence is created, and sent to the
host site, and TTYFTP can connect to the following sites via TYMNET:
1. Sumex,
2. SRI-KL and,
3. RU-AI.
After the login sequence is sent, all characters received from
the HOST are echoed at the users console so that he/she has the option
of aborting the job if, for example, the host site has no available ports.
During this process ↑X will abort the job. If on the other hand, SERftp
gets started, then it will send the string
SERftp started @ <host site name>,
to which the user responds by typing an S. TTYftp will then send an open
connection packet to SERftp, and if all goes well, then
(OK @<host site name>)
>>
will be typed at the user's consule, and TTYFTP is in command mode.
The following commands are available:
NOTE: Command recognition only requires the letters in uppercase;
BUT "$" (escape) does NOT work as in the TENEX Exec. In particular,
File names must be completely typed.
1. REQuestConnection Requests connection to HOST,
2. CloseConnection Closes open connection at HOST site,
3. Store <fd> Send <file name> to HOST site. A NEW file,
is always created at the destination.
<fd> = <file name>sp<type>,
<type> = See DATA command,
(type is optional and causes
an implicit DATA command),
sp = space(a blank),
4. RETrieve <fd> Retrieve <file name> from the HOST site,
5. Append <fd> Append TO <file name> at HOST site,
6. User <user-id> USER file access protocol (CONNECT to <user-id>)
<user-id> = valid HOST identification,
A password is prompted for when required.
7. DAta <type> Sets the data format in which to read/write,
a file for both USER and HOST.,
<type> = ASCII,
Standard 8 bit ASCII,
*** ASCII is DEFAULT type ***,
NOTE: The following are NOT implimented:
EBCDIC,
8 bit EBCDIC characters,
IMAGE,
contiguous stream of 36 bit
binary data,
! The DATA command remains in,
effect until <type> is changed,
by either a DATA, STORE or RETRIEVE,
command,
8. DElete <filename> Deletes the HOST file <filename>,
9. DIr <filid> Searches connected directory for files ,
satisfying the <filid>, where,
<filid> = <filename> OR appropriate,
WILDcard specifications,
NOTE: <fileid> = NULL gives a complete
directory listing.
10. Exit Exit program,
SIGNAL RESPONSES can be given while a TRANSACTION is in progress.
These responses are recognized by the scheduler, and are NOT commands.
Consequently, they will be taken as command characters if typed on a
command line.
↑S Gives transmission status for packet transmission,
↑F File transfer status,
↑Y Will ABORT a FILE transfer. This response causes the FTP to send
an INTERRUPT packet to the HOST, which, upon receiving this
packet, sends an INTERRUPT packet back to the USER.
When the USER FTP receives the INTERRUPT from the HOST, it sends
an ABORT message on channel 0, and, the HOST responds with
(OK (RETRIEVE ABORTED)), or (OK (STORE ABORTED)).
The USER will see one of the above at the conclusion of the
abort sequence. If the aborted command was a STORE, or APPEND,
the partial file is deleted at the HOST site, but NOT expunged.
If the command was a RETRIEVE, the partial file is deleted at the
USER site, but not expunged (If the consensus is to leave
partially transferred files hanging around, this can be easily
changed. So, if you find deleting particularly distressing,
please let me know. (Bill <Yeager@sumex>).
↑X Will ABORT the program. This response sends a CLOSE connection
packet to the HOST, closes all files, releases the line,
and exits.
!BATCH MODE:
TTYFTP will run in a batch mode. It looks for a file, TTYFTP.CMD,
on the USER's directory, and reads all commands from this file. If TTYFTP
batch is used in conjunction with SUBMIT on Sumex, then files can be queued
for transfer when the load is low, and the NET BAUD rate is high. 1200
BAUD is the optimal BAUD rate on TYMNET when both the NET load, and USER
and HOST loads are minimal. Similarly, if one is using an ordinary phone
line from an 11 to the 10 for example, then if the load is high on the
10, the job can get SCHEDP'd and data will trickle along the phone line
if you're lucky.
Even under the worst of load conditions the file transfers
will complete as long as someone sends a packet before the 15 minute timer
expires. But, anyhow, batch is a good idea when the load is high and the
file sizes are large.
BATCH example:
sample TTYFTP.CMD file (comments - NOT in file please)
RU-AI for connection
YEAGER mypassword "
Store test.file local file to be sent.
test.fileRU-AI host file name to be created.
Close CLose connection
Exit Exit
sample .CTL file for SUBMIT at SUMEX
@TTYFTP
*b
If the above .ctl file is submitted, then you will login @ ru-ai
under the account YEAGER; Store test.file, Close the connection and Exit.
See the sample session below for details:
!SAMPLE SESSION:
@ttyftp
TTYFTP Jogging at 143 LA = 2.01, used 0:00:01.8 in 0:00:07
Please enter NETWORK name: ?
TYMNET
bTYMNET TYMNET batch mode,
Quit Exit program.
Please enter NETWORK name: t
- - Sumex TTYFTP - version 2.0 - -
(? for aid)
>>Connect TO: ?
Select from following sites:
Sumex
SRI
RU-AI
>>Connect TO: r
(WAITING for connection @RU-AI)
Enter login Directory: yeager
Enter password:
making AUXILIARY connection...
Starting the server, SERFTP, @ RU-AI
[Type S when SERFTP is Started, ↑X to Abort]
p 03
RUTG is online
For trouble with Tymnet access, please report to (201)-932-2443.
Tymnet users, type HELP TYMNET.
RUTGERS/LCSR, TOPS-20 Monitor 3A(23013)
@
@
@
@
@LOGIN YEAGER
Job 18 on TTY61 20-Jul-79 16:13:50
Last login on 20-Jul-79 at 15:55:00
@
@
@
@
@
@
@
@ Load 1.61 Used 0:00:00.2 in 0:00:01
@
@REFUSE LINKS
@
@RU <YEAGER>SERFTP
-- TTYFTP SERVER is running @ RU-AI
s
<< TYMNET up >>
(OK @RU-AI)
>>dir
(OK (>> Here comes list ...))
File Name Pages
20-ARCHIVE.DIR.1 1
BAR..1 0
FTPHDR..1 1
FTPLIB..1 4
FTPPAK..1 1
FTPSER..1 28
FTPSER.RIM.7 10
JSYHDR..1 2
JSYLIB..2 10
LIBHDR..1 1
MAIL.TXT.1 3
PAKHDR..1 2
PAKSER..20 29
PAKSER.RIM.2 9
SCDHDR..1 1
SERFTP..1 2
SERFTP.EXE.2 1
SERFTP.RIM.2 1
SERSCH.RIM.3 4
SERSCH.RIM↑V(OLD↑V).1 4
TTYSCH.NEW.1 17
TV-STRINGS.PMAP.1 1
(DONE (>> list terminated))
>>del bar..1
(OK (>> bar..1 Deleted))
>>ret libhdr
>>>Write To LOCAL File: zapme.now
(OK (>> here comes libhdr [1145(7) bytes]))
↑F
For File zapme.now:
BYTES rcvd/sent BYTES left Elap. time BAUD rate
160(7) 985 00:00:05 320
↑F
For File zapme.now:
BYTES rcvd/sent BYTES left Elap. time BAUD rate
960(7) 185 00:00:32 300
1145(7) bytes transmitted in 00:00:38. NET BAUD RATE = 301
>>st zapme.now
>>>Write To HOST File: ?
Enter the HOST file name to which the
LOCAL file is to be written
>>>Write To HOST File: zapme.fromSUMEX
(OK (>> send zapme.fromSUMEX!))
(>> Storing 1145(7) bytes.)
↑F
For File zapme.now:
BYTES rcvd/sent BYTES left Elap. time BAUD rate <THIS IS UPDATED every
0(7) 1145 00:00:18 0 1280 bytes for STORE>
(DONE)
1145(7) bytes transmitted in 00:00:43. NET BAUD RATE = 266
>>dir zapme.*.*
(OK (>> Here comes list ...))
File Name Pages
ZAPME.FROMSUMEX.1 1
(DONE (>> list terminated))
>>del zapme.fromsumex.1
(OK (>> zapme.fromsumex.1 Deleted))
>>user sumex
(FAILED (>> "sumex" Illegal USER))
>>user <sumex>
(OK (PASSWORD?))
>>PASSWORD =
(OK (>> [<sumex>] connected))
>>diir
(ERROR: "diir" INVALID command!)
>>dir
(OK (>> Here comes list ...))
File Name Pages
FOO.MAC.1 1
FOO.REL.1 1
LOGIN.CMD.1 1
MAIL.TXT.1 2
MAKTAB.TNX.2 5
TV-STRINGS.PMAP.1 1
(DONE (>> list terminated))
>>ret foo.mac
>>>Write To LOCAL File: zapme.soon
(OK (>> here comes foo.mac [25(7) bytes]))
25(7) bytes transmitted in 00:00:04. NET BAUD RATE = 62
>>close
(OK RU-AI is DISCONNECTED)
>>exit
.DONE ttyFTP.
@
!
-------
∂27-Jul-79 2122 PEG New version of PCPASC
To: S1 at SU-AI
A new version of PCPASC is up with a new feature and a small, but important,
bugfix. PCPASC now supports the %<comment>\ convention, making it compatible
with the system PASCAL and allowing VERionCHange nested comments. Note: the
double-quote ("<comment>") convention is still supported. The former is the
new feature; the bugfix is that CHK instructions for characters check the range
0..127 (previously: 0..63), allowing UASMINT to use PTRANSed code without
failing a CHKH instruction. Note: using the B+ switch instead of the D+ switch
to turn on debugging emits CHK instructions for characters in the range 0..255,
as previously.
New files:
PCPG10.PAS[PCP,S1] New source.
PCPG10.LST[PCP,S1] List produced by DEC-10 PASCAL.
PCPG10.DMP[DMP,S1] New executable file.
Note: DO files on [DO,S1] have been updated to use PCPG10.
∂27-Jul-79 2217 ARR New VERCH up.
To: S1 at SU-AI
The version changer for PASCAL sources, VERCH, has been modified to use
% and \ instead of (* and *), to prevent the compiler from getting confused
by nested comments with the same symbology. The old one, which uses (*, *)
is in VERCH.OLD[DMP,S1].
If you have used VERCH since it was released, you need to get your sources
fixed for the new convention. You need to do it in two steps:
1. Set to false all the flags that are currently true.
2. Use VERCH.OLD, to suppress all the (*...*) marks.
3. Set the flags as you want them.
4. Pass it through VERCH, as it is now.
If you have (*%ELSE *) parts, you will need to fix those by hand.
From now on, I propose the next policy for the use of comment symbology:
A. Use (*...*) for your comments.
B. Use %...\ ONLY for version changes.
C. DO NOT use "...", please.
Armando.
∂29-Jul-79 1443 ARR PEG ARR New VERCH up.
To: S1 at SU-AI
∂27-Jul-79 2323 PEG new VERCH
∂27-Jul-79 2217 ARR New VERCH up.
To: S1 at SU-AI
The version changer for PASCAL sources, VERCH, has been modified to use
% and \ instead of (* and *), to prevent the compiler from getting confused
by nested comments with the same symbology. The old one, which uses (*, *)
is in VERCH.OLD[DMP,S1].
If you have used VERCH since it was released, you need to get your sources
fixed for the new convention. You need to do it in two steps:
1. Set to false all the flags that are currently true.
2. Use VERCH.OLD, to suppress all the (*...*) marks.
3. Set the flags as you want them.
4. Pass it through VERCH, as it is now.
If you have (*%ELSE *) parts, you will need to fix those by hand.
From now on, I propose the next policy for the use of comment symbology:
A. Use (*...*) for your comments.
B. Use %...\ ONLY for version changes.
C. DO NOT use "...", please.
Armando.
[Could you clarify what you mean by (B) above? i.e., do I say
1. %%ift ....\
<code> (*<comment>*)
%%endc\
or
2. (*%ift ...*)
<code> %<comment>\
(*%endc*)
or
3. (*%ift ...*)
<code> (*<comment>*)
(*%endc*)
or what?]
[[ Use (3) above: in other words, the only commenting simbology that you have
really is (*...*)
Armando.]]
∂30-Jul-79 0000 Mark Crispin <Admin.MRC at SU-SCORE> having only Dialnet and local mail
Date: 30 Jul 1979 0000-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: having only Dialnet and local mail
To: JMC at SU-AI, BH at SU-AI, Les at SU-AI
My Tops-20 implementation has always considered this from its
very beginnings. I don't know if Brian is worrying about the
WAITS running on the musician's Foonly or not, but none of my
WAITS Dialnet software requires an Arpanet.
-------
∂30-Jul-79 1305 DCL Guest users at heavy load times
To: LES
CC: JMC
Les, the afternoon loads have been creeping up; queuing time for a channel
have been up to 1/2 hr. At these times there have been quite a few non
project users and even some "unknown"s.
Could you reiterate your policy of such people getting off the system at
these times.
-David
∂30-Jul-79 1358 BH Dialnet and no Arpanet
To: JMC at SU-AI, MRC at SU-AI, LES at SU-AI
No problem, assuming the host table reader in NETWRK doesn't mind.
∂31-Jul-79 1021 PAT
∂30-Jul-79 2047 JMC
Any sign of Tang or news of visa?
--------
Nothing yet
∂01-Aug-79 1240 ALS Status of UCode SOPA
To: S1 at SU-AI
As per a request by LCW, I am mailing a note on the current status of
the UCode version of SOPAIPILLA, hereinafter called SOPU.
SOPU is descended primarily from SOPG16, with minor sections from the
preliminary UCode SOPA, SOX. Therefore it is completely up-to-date as
regards the four-word set change, and only slightly behind SOPZ17 as
regards optimization improvements. ALS and I anticipate little difficulty
in merging the latter into SOPU, since PTZ's optimization changes are
pretty much localized.
As of this writing, SOPU will compile the test program SINE (the UCode
being provided by PCPG10 and PTRANS), and the result runs correctly to
completion on FSIM. Due to a slightly different stackframe format, we
found it necessary to make very minor changes to the S-1 assembly-coded
runtimes; SOPU-produced LDI files run with the PASUCO runtime files.
SINE is a main-block-only program, with calls only to standard procedures.
Debugging of the user-procedure calling sequence code is currently proceeding
rapidly on FFT, a fairly call-rich program. SOPU does not yet run to
completion on FFT. All of the errors found so far have been minor ones
(e.g., typographical errors or forgetting to increment a counter) rather
than conceptual ones.
Phil Gerring
∂01-Aug-79 1519 CET Policy Information re Dissertations
To: GHG at SU-AI, RWF at SU-AI, FB at SU-AI, DEK at SU-AI,
JMC at SU-AI, TW at SU-AI, Buchanan at SUMEX-AIM
CC: DPB at SU-AI
A student needs a decision by Friday as to whether he can produce his
thesis printed on two sides (there are 400 pages). The Grad. Program
Office said it is up to the Dept. Denny is not sure if there is a
Dept. Policy guideline, but we need one. Please send your opinion as
quickly as possible.
I am dubious about 400 page dissertations, but allowing that, it
seems that double sided would be better.
Thanks,
Carolyn
∂01-Aug-79 2246 LES Books for Michaels
We can do it on 8 channel, but as you know that is a fair hassle.
It's Michaelson, and I know it's a hassle, but the bread will return.
∂03-Aug-79 1039 MRC Sumex's TTYFTP program
To: LES at SU-AI, JMC at SU-AI, Yeager at SUMEX-AIM
I just TTYFTPed a big file (183243 bytes) to Rutgers from Sumex. It
took over an hour and 40 minutes, with an overall average baud rate
of 304 baud on a 1200 baud Tymnet link. In case you forgot, TTYFTP
is essentially Sumex's implementation of Dialnet, with a few minor
incompatbilities because TTYFTP is in user mode.
To be fair, I must mention that this is with an increasing load
average on Sumex, having started at 8:30 in the morning, a noon-
time load at Rutgers, and a morning daytime load on Tymnet. The
last is a major consideration. I don't know of any way to meter the
performance over Tymnet; but we cannot assume a dedicated 1200 baud
link to Rutgers, which Dialnet does have. Towards the end of my
transmission, Sumex's load average was hitting over 8, although it
made little effect on my cumulative average baud rate. Sumex didn't
get too bad, because I didn't get SCHEDP'd.
Still, part of the problem is due to the TTYFTP implementation. It
had to be this way considering Sumex's needs - the Sumex management
is unwilling to have operating system changes made to support
Dialnet. These implementation factors are:
. TTYFTP is written in MAINSAIL, not assembly language. Even
though MAINSAIL is a optimizing compiler it still cannot hope
to match a well-written assembly language program in terms of
performance, especially given its large object time system.
Writing in MAINSAIL did speed up TTYFTP's implementation time,
of course, and at no time was the operating system's integrity
jeapordized during TTYFTP debugging.
. TTYFTP is user-mode, so it can't get its data at interrupt
level, nor can it output at interrupt level. Thus TTYFTP is
greatly affected by variations of system load and runs at the
whim of the operating system schedular, while Dialnet gets its
cycles whether or not the schedular likes it, assuming there
are any cycles to take.
. This is made worse by the fact that Tenex TTY buffers are only
about 90 bytes long, making the maximum packet size that TTYFTP
can use effectively (even on a lightly loaded SUMEX) about 50
bytes. This compares with Dialnet's 256 data byte packets.
With 10 protocol overhead bytes this means that 1/5 of TTYFTP's
bandwidth is wasted in protocol, compared to 1/25 in Dialnet.
When all these factors are considered, the performance difference
is completely understandable. Even given Dialnet's "optimal"
implementation, you only get about 2/3 - 3/4 of the media baud
rate with Dialnet, compared to the 1/4 (under non-optimal conditions)
with TTYFTP. This indicates to me that the TTYFTP is in fact a
reasonable approach when "network" performance is not as important
as the purity of the operating system.
I don't know if Bill Yeager (TTYFTP's author) has any comments to
add to these. I'm sure he has a better idea than I do about which
factors were more important than others in the performance in my
TTYFTP session this morning.
I should close with the note that I am NOT criticizing TTYFTP - it
has a VERY tasteful user interface and its performance was in fact
better than I had expected when all the problems I mention above are
allowed for. I am a lot less happy with Dialnet's performance at
SU-AI, even though I know where the performance problem is and it's
not my fault.
- Mark
∂03-Aug-79 1056 Yeager at SUMEX-AIM Re: Sumex's TTYFTP program
Date: 3 Aug 1979 1055-PDT
From: Yeager at SUMEX-AIM
Subject: Re: Sumex's TTYFTP program
To: MRC at SU-AI, LES at SU-AI, JMC at SU-AI
cc: RINDFLEISCH
In response to the message sent 03 Aug 1979 1039-PDT from MRC@SU-AI
The NET BAUD RATE of 304 computes to about a GROSS BAUD RATE of
400. THE NET calculation is for file bytes only. This time of
day the TYMNET load is quite high, and given the SUMEX load that
Mark ran under, I am suprised he did so well...
TTYFTP sends 54 byte packets, with a packet window of 2. I have
been able to obtain NET BAUD RATES of 600 which computed to
GROSS BAUD RATES of 800 during very early am. This is quite good
considering the packet size. The major limitations don't have much
to do with MAINSAIL since the program is has no trouble keeping up
with packet IO. It in fact dismisses frequently so as not to use
up too much CPU time and get SCHEDPd while waiting for packet input.
It actually has a scheduler that controls just who is called when, and
the packet handler is called the most frequently.
Again, TTYFTP is 90% DIALNET protocols, and the differences are at the
FTP level, and not at the packet level, and I think that Mark's efforts
have a great deal to do with TTYFTP's performance and reliability...
Bill
-------
∂03-Aug-79 1115 DEW
To: REF, JMC
I couldn't lend the expertise to REF's paper that JMC could but would be
happy to deliver it. Only 2 possible problems:
1) it must be scheduled so as not to conflict with my talk or Lew Creary's,
and 2) I have to find a way to Japan. I called the travel agent the same
day I found I was getting money, but all flights to Tokyo for Aug 17-20 from
LA, SF, or Seattle are completely booked with their waiting lists closed.
So it's not obvious I can get there. (It would have been nice to have more
than 3 weeks notice to arrange everything.) John, if you have a flight to
Tokyo booked and aren't going to use it, perhaps I could get it somehow.
Dave
I haven't booked a flight. If I go, it would be from New York.
∂03-Aug-79 1129 MRC TTYFTP
To: Yeager at SUMEX-AIM
CC: JMC at SU-AI, LES at SU-AI, Rindfleisch at SUMEX-AIM
In response to Bill Yeager's message at 8/3/79 1055:
I was talking about net baud rates as well. While media affects the gross
baud rate, I don't think it's right to judge a protocol by anything other
than the net baud rate, which is what the end user sees. Talking about
gross baud rates simply makes inefficient protocols look good. We can say
that the Arpanet is a 50kbd network, but when a user sees 5kbd in his or
her FTP, s/he doesn't see it as a 50kbd network! I don't think any of us
have ever bothered to measure gross baud rates.
The packet size IS quite significant in determining the difference.
Dialnet/TTYFTP require 4 framing, 4 packet header, and 2 checksum bytes in
addition to the packet data bytes. Assuming that Bill's 54 byte packet
size includes the packet header and checksum, but not the framing bytes
(48 data bytes sounds logical), we are talking about 10/48 vs. 10/256
overhead/data ratios or 10/58 vs. 10/266 overhead/total data. This is
assuming that there aren't any '220 bytes in the data steam which need
quoting.
I didn't get SCHEDP'd this TTYFTP run, and when I started off Sumex was
fluttering at a load between 1 and 3. I was pushing to Rutgers, which is
probably less time-consuming at Sumex than pulling to Sumex or pushing
from Rutgers to Sumex. Yesterday, I did get SCHEDP'd, and eventually was
forced to give up.
Mark
∂03-Aug-79 1956 DPB Zohar
To: JMC
CC: PAT, CET
∂03-Aug-79 1952 JMC
To: PAT, DPB
Any sign of Zohar?
DPB- I heard something from Carolyn like Sept. 1 for his arrival date.
∂04-Aug-79 0255 PTZ new SOPA - SOPZ18[DMP,S1]
To: S1 at SU-AI
I recoded SOPA routines GETFIELD, PUTFIELD, and GETSIGNEDFIELD in
FAIL (for the PDP-10 version) and S-1 assembler.
Testing indicates that SOPZ18 runs about 40% faster than SOPZ17.
More details in SOPA.[LOG,S1]/2P.
Polle
∂04-Aug-79 1347 RWG via MIT-AI
tnx for mandelbrot msg. it turns out i am looking for him!
∂04-Aug-79 1908 LGC 1. Important New Ideas; 2. IJCAI
1. New Ideas: My reading, early in July, of the Ph.D. theses of
Mark Stefik and Peter Friedland brought the set of ideas about the
Advice Taker that I have been developing over the past year up beyond
the critical mass, and provided a strong stimulus to my thinking.
Since that time, a powerful and creative thought-reaction has been
taking place, and the results so far include many valuable insights,
clustered around 2 good scientific ideas and 2 good new practical ideas
for implementing the scientific ones. I am very excited by these ideas,
and I have reason to believe that you will like them too (they go well
beyond a minimal Advice Taker, and are not the same as those you read just
before you left). They more than fulfill the high hopes that I had at
this time last year, after you first made it possible for me to work in
this particular corner of AI. Thus, I look forward very much to
discussing the new ideas with you in detail when you return in September.
I will be on vacation during much of August.
2. IJCAI: As you probably know, there wasn't enough ARPA travel money
to send everyone at the lab who wanted to read a paper at IJCAI, and the
result of LES's democratic decision procedure was that Rod Brooks and Dave
Wilkins will attend, with the understanding that they are willing to read
papers for those of us who can't attend due to lack of funds. However,
the decision came about 6 weeks too late, with the result that Dave
Wilkins may not be able to get a plane to Japan at IJCAI time. All of
this raises the question as to who will present my IJCAI paper for me. At
first I had thought that having Dave do it would be the simplest satisficing
solution, but now there is a significant chance (at least 1 in 4) that he
will not get to IJCAI. Two possibilities that should be considered at
this point are that either you or RWW might read my paper for me, either
unconditionally, or on the condition that DEW doesn't make it to IJCAI.
According to the IJCAI schedule, you are to chair session 9C (8/23 @ 9:00
a.m.), while Richard is to chair sessions 1E (8/20 @ 14:00) and 3C (8/21 @
9:00). My paper (20 + 8 minutes) is scheduled for session 7E (8/22 @
14:00), as is Bob Filman's. Jon Doyle and Drew McDermott will also have
papers in that session. Dave Wilkins is scheduled for session 9D (8/23 @
9:00), in conflict with your 9C. Before I attempt any further
arrangements, I would appreciate having your advice on this matter.
Thanks -- Lew
∂05-Aug-79 1029 PAT
no, haven't seen him yet......
∂05-Aug-79 1647 TM log area
To: S1 at SU-AI
I have cleaned up the area [LOG,S1]. There were a number of distribution
lists with two logs (one with a null extension, and one with a MSG extension).
Where there were two log files, I merged them. Now all of the log files
on [LOG,S1] have a null extension, and the distribution lists have been
updated accordingly. The larger log files have been split into two files,
one for messages before May 1, 1979 (.OM2), with the messages after that date
in the log file. All of the OMS files have been renamed to OM1. Both
the OM1 and OM2 files have been moved to the S1 UDP. The idea is that whenever
the LOG area gets too large, somebody will go through moving the old messages
to files with the extension OM#, and then those logs of old messages will be
moved to the UDP. It has been suggested that FICHE copys of the old logs be
made. Anybody that feels so inclined is welcomed to do so.
Tom
∂06-Aug-79 1041 PAT Zohar
To: JMC
CC: DPB
I had a letter from Zohar saying he will be back at the end of the month.
∂07-Aug-79 1115 ARK S-1 Meeting
To: S1 at SU-AI
CC: pickering at SUMEX-AIM
There will be an S-1 meeting NEXT WEEK, Wednesday, 15 August at 1:30 pm at the
SAIL small conference room. There will be no S-1 meeting this week due to
Brian McCune's Ph.D. orals.
Arthur
∂07-Aug-79 2053 DCL AI Lab. cost breakdown
To: JMC at SU-AI, HVA at SU-AI
CC: LES at SU-AI, FB at SU-AI, SAL at SU-AI
John,
My ARPA proposal has now reached the government preaudit stage.
The auditors are requesting a breakdown of the computation of AI LAb.
computing costs. Unfortunately LES is away bicycling until Aug. 13th.
Did he leave details with you? (or Hersche?)
Considering the hassles that went on over the SCORE costs with FBs VLSI
proposal, I expect that LES has something prepared and left it with you
or Hersche.
- David
∂08-Aug-79 1119 CET Computers
To: FB
CC: JMC
John McCarthy is trying to reach you re possibility of getting 801 computer
instead of 4331. Please call
914/945-2893
∂08-Aug-79 1149 CET Classnotes for CS206
To: CLT
CC: JMC, CJS
Carolyn,
Connie Stanley will be handling the printing of classnotes.
This is a reminder that she will need some lead time in getting them
printed. When will they be ready? They should reach the bookstore
by Thursday, Sept. 20.
Carolyn
∂08-Aug-79 1321 HVA A.I. Lab Cost Breakdown
To: DCL
CC: LES, JMC, HVA
David--Rec'd. your message this morning and shortly thereafter I received a tel-
phone call from Fred Gustafson, S.P.O., who said he had asked Sally Burns to
send me the exact questions DCAA had raised. Fred Gustafson will be out of town
until next week (and so Les), but I'll take a look at questions when rec'd. Les
did not leave any formal statements on this with me.
∂08-Aug-79 1920 REM via AMES-TIP
Technology Licensing wants to drop my invention. Unless you or RWW can
suggest to them a good course of action on my invention, I'll be forced to
withdraw it from Stanford and persue it on my own. Lack of help from
Stanford will mean (1) loss of revenue to Stanford in the case I succeed
in making lots of money (2) loss of time for me as I spend my time doing
everything myself (3) increased chance of total failure due to lack of
good contacts with businesses via Stanford expertise.
Inventions, in my experience, are always a long shot to make money.
My advice is to listen carefully to the reasons technology
licensing gives for not being hopeful about this one.
∂08-Aug-79 2035 DCL Comment on Jacks revision of August 1st.
To: DPB
CC: FB, JMC
Denny,
I see in the latest revision that the Program Verification group has no
group office at all, and that the Texas Instruments computer donated to the
group, is not allocated space.
IN THE PREVIOUS PLAN THE GROUP HAD 2 1/2 SMALL SHARED OFFICES.
Perhaps we should make an appoinment to talk all this over.
PV is currently budgeted by ARPA to use 15% of the AI machine over the next
two years. This represents $80k. We can only do this work if the group has
access to data disc terminals on an exclusive basis - there are 5 fulltiime
postdoctoral workers (DCL,DCO,FWH,WP,SMG) plus students. I think the minimum
would be one large office with three data disc terminals or two small offices
with two terminals each.
In addition, we are budgeted by ARPA to develop a concurrent systems programming
language for the TI system. So one office should be close to that facility, if
there is room for it (if not please let me know now).
I am at a complete loss to see how we can carry out our contracts under the
present scheme.
-David
∂09-Aug-79 0929 CET
To: TW, CLT, JMC
∂09-Aug-79 0923 CJS classnotes
I'm going away on vacation from Sept. 6-15, and the printers are going
to hold my stuff for me. If I don't have the master for printing before
I leave (and I'm not talking about the day before, give me a break),
classnotes are going to be late in being printed.
Connie
[Carolyn and Terry, can you get material to Connie by 8/31?]
∂09-Aug-79 1253 Evans@BBND(Art-Evans) Extra Sept meeting
Date: 8 Aug 1979 2229-EDT
Sender: EVANS at BBN-TENEXD
Subject: Extra Sept meeting
From: Evans@BBND(Art-Evans)
To: BOSTON-SIGPLAN:
Message-ID: <[BBN-TENEXD] 8-Aug-79 22:29:28.EVANS>
Fcc: <EVANS>MAIL.DIV6
Redistributed-To: JMC at SU-AI
Redistributed-By: EVANS at BBN-TENEXD
Redistributed-Date: 9 Aug 1979
SIGPLAN Meeting, Thursday August 16, 8:00 PM
Intermetrics, 737 Concord Ave, Cambridge
Programs as First Order Theorems
John McCarthy, Stanford University
The talk will investigate verifying programs by computer-checked
logic proofs of their properties. What are the specifications of
a program? The talk will consider recursive and sequential
formalisms and show why Perlis' pessimism is unjustified.
Professor McCarthy is widely known as the inventer of LISP, and
not so widely known as one of the first (late fifties) proponents
of computer-checked proofs of correctness. He is one of the
grand old men of computing (forgive me, John), and is currently
director of the Stanford Artificial Intelligence Laboratory.
Those interested may join the speaker for dinner at Joyce Chen's
at 6:00 -- make reservations before August 16 with Lynn Bates at
491-1850, ext 4634, or by ARPANET mail to Bates@BBND.
Upcoming speakers:
Sep 6 Bob Frieburghouse, "Realtime Extensions to PL/I"
Sep 20 Nico Haberman, a topic in Software Engineering
Oct 4 John Reynolds, to be determined
Nov 1 Brian Kernighan, "Elements of Style"
Dec 6 Paul Abrahams, "The PL/I Standard, an Apostate's View"
Jan 10 Fran Allen, "Optimization"
Art
∂10-Aug-79 0000 JMC*
GAUSS letcure
∂10-Aug-79 1147 TW Access to the SAIL computer
To: LES at SU-AI, JMC at SU-AI, EAF at SU-AI, DPB at SU-AI,
FB at SU-AI
CC: TW at SU-AI, PAM at SU-AI, RPG at SU-AI, DES at SU-AI,
JLS at SU-AI, CF at SU-AI, AH at SU-AI, DEA at SU-AI, JED at SU-AI,
DLO at SU-AI, chapiro at SRI-KL, gardner at SUMEX-AIM,
levy at PARC-MAXC2, masinter at PARC-MAXC2
I have been following the various statements about changes to the policies
for the AI lab computer after its move to campus, and have had discussions
with several students who are worried about the effects it will have on
their work. I understand that the policy will be to provide accounts only
for people (I am not sure whether this includes faculty as well as
students) covered on sponsored projects which pay for a pie slice. Since
I have no grant now which includes money for that, and since there is no
possibility of having one in less than 6-9 months, this will affect a
number of people. My current understanding of the list is:
1) People who have a serious investment in research programs which run on
the SAIL machine and would take a lot of work to move:
Paul Martin, Dick Gabriel, David Shaw, Juan Ludlow
2) People who are well along in doing dissertation writing using the text
preparation facilities at the lab and who would be seriously
inconvenienced by having to transfer:
Annette Herskovits, Fernando Flores
3) People who have access to other facilities, but make extensive use of
the SAIL information and communication environment:
Terry Winograd, Doug Appelt
4) Students who I am an academic advisor for, who make use of the AI lab
facility and do not yet have a dissertation advisor:
Jim Davidson, David Lowe, Daniel Chapiro
5) Other CS students I am supervising who do not make significant use of
SAIL facilities:
Larry Masinter, David Levy, Anne Gardner
It is important that people of groups 1 and 2 be able to continue their
use of the SAIL facility. It would also be quite convenient for groups 3
and 4. Presumably people in group 4 will eventually end up in a
particular project. I am happy for people in Group 5 and other new
students not to have accounts, as long as there are satisfactory text
editing and communication facilities on SCORE.
If it is not possible to provide accounts without a paid-for pie slice, I
propose that some kind of exchange arrangement be worked out where the
SAIL sponsored projects get some kind of access to SCORE in exchange for a
slice of SAIL for CS students and faculty not under sponsored projects. I
will leave the details of this up to you to work out (or any other
arrangement which makes it possible to have accounts). If it is not
possible with the current plans, I would like to further discuss those
plans.
--terry
Thanks for outlining the dimensions of the problem. Feigenbaum and
I discussed the problem, and we agreed that there would be a
departmental pie slice that the department would pay for. It
was not discussed how big this slice would be.
∂10-Aug-79 1748 100 : REM via AMES-TIP Tech. Licensing / IRSM
Mostly they contacted a few companies and got a few not-interested
responses, and gave up. -- I plan maybe to tackle it ass-backwards, first
lining up potential customers who do lots of data communication and/or
storage, getting letters of intent to maybe-buy, then taking these letters
to manufacturers trying to get someone to build a demo prototype.
I wish you luck, but you will need to be very persistent and still
may not succeed.
Actually, a more promising route to fame and fortune may be
to publish a paper in a journal of reasonably wide circulation
such as one of the IEEE publications. Your services as a consultant
or employee will then become valuable. If you are interested, I
will inquire whether IBM might be interested in employing you.
i
∂14-Aug-79 1113 PN Introducing PAS, a pseudo-monitor (Away with all DO files!)
To: S1 at SU-AI
I was typing [DO,S1] for the 2,436th time two days ago, and decided I
couldn't take it any more. The result is PAS, a combination of SNAIL and
DO that can be used to compile and run Pascal and Fortran programs at SAIL
using the S-1 Pascal system.
Some examples:
R PAS;FNAME (compiles FNAME.PAS or FNAME.FOR and runs it on the simulator)
EX (usually re-executes last PAS command (also control-X G from
the editor))
R PAS (ALWAYS re-executes the last PAS command)
R PAS;FNAME.PCO (put FNAME.PCO through SOPA and run it on the simulator)
R PAS;FNAME/C100000 (when running on simulator, increase core to 100000)
R PAS;FNAME/PS1/E (translate FNAME.PAS to FNAME.PS1 and edit the result)
R PAS;FNAME/OUT (compile and run FNAME.PAS, feeding FNAME.OUT to FSIM
for the Pascal output file)
R PAS;FNAME.LDI/TTYOUT (run FNAME.LDI on simulator, use "TTY:" for output)
R PAS;FNAME/U (use U-code system instead of P-code system)
R PAS;FNAME/O (run U-code through the optimizer)
R PAS;FNAME/INT/E (compile FNAME.PAS into U-code, run it on the interpreter,
using FNAME.OUT for OUTPUT, then look at the result)
R PAS;FNAME/LDI/D (don't say "END" to debugger when running SOPA (D stands
for "DEBUG")
R PAS;FNAME.LDI/FOR/RIM (load FNAME.LDI using Fortran run-times and write
out the RIM file)
R PAS;? (display HELP file)
For more details, type READ PAS.
∂14-Aug-79 1512 Purger
To: S1 at SU-AI
You are exceeding your disk quota.
Files that occupy space beyond your quota are subject to purging.
If you don't delete some of your files, the purger will.
Your disk quota is: 9000
Your files occupy 10318
∂14-Aug-79 1622 ARR Reminder: S-1 meeting
To: S1 at SU-AI
CC: pickering at SUMEX-AIM
As announced last week by Arthur, there will be an S-1 meeting tomorrow, wednesday,
15-aug-79, at 13:30, in the small conference room of SAIL.
Armando.
∂14-Aug-79 1936 BIS Hand/Eye Equipment
To: DPB at SU-AI, ME at SU-AI, REG at SU-AI, LES at SU-AI,
JMC at SU-AI
CC: DCL at SU-AI, MSM at SU-AI, JKS at SU-AI, TOB at SU-AI,
ARG at SU-AI, JM at SU-AI, HPM at SU-AI
I'm glad you asked that question! We were just discussing it this afternoon,
with the following results.
First order constraints. (Equipment itself)
PDP11 - 2 racks + VT05 Terminal
4 arms + hand/eye table
2 solid state cameras
2 vidicon cameras (may be retiring)
ceiling lights for illumination
TIs - 2 racks + 2 disks + line printer + 2 terminals
Grinnell display memory (1 rack)
2 data disk terminals
bookshelf
filing cabinet
table space for papers while working
storage cabinet for jigs, assembly fixtures, calibration devices
2-3 chairs
Second order constraints. (Equipment interactions)
Solid state cameras must be very near the Grinnell →
Grinnell in the Hand/Eye room.
TIs will process vision (in their Hand/Eye capacity) →
TIs should be in or very near the Hand/Eye room,
remembering that they will be used also by the
Verification Group.
Grinnell will supply data to monitor(s) at the Stereo Station → ?
Further considerations:
Where will the CART sleep between runs?
Where should the TIs be located so as to facilitate their use
by both Hand/Eye and Verification? (A single larger room
might be a good solution; we could knock out the wall
between the two small rooms.)
Hope that this information aids in planning.
∂15-Aug-79 1500 LES AISA
Raj Reddy came by for lunch yesterday and was talking about fairly concrete
plans for an "Artificial Intelligence Society of America". He wishes to
talk to you about this scheme sometime soon.
The current plan apparently involves two main functions: publishing a
newsletter and putting on an annual conference in the U.S. They plan
to have these functions performed by almost-independent committees running
under a board of directors (or trustees or whatever).
One early goal will be to either absorb or kill off the IEEE SIGART
newsletter. A possible future goal will be to do the same thing to the AI
Journal.
Another plan is to offer "Kramer Prizes" for various AI accomplishments,
such as first program to receive a Chess Grand Master rating or first
program to take dictation with certain prescribed error rates, or whatever.
He asked me to help with certain parts of the newsletter project.
From what I know, this looks like a reasonable undertaking.
Seems reasonable. I will be back about the 30th of this month.
∂16-Aug-79 1445 TOB
To: JMC, FB
Visitor - Aug 16, 79
I have a request for one year visit at the robotics group
from Prof P.J.MacVicar-Whelan, Grand Valley State Univ, Mich.
Dr. George Giralt, leader of the largest robotics group in France,
who knows him well, gives him a high recommendation as a
scientist and serious researcher who works hard, is highly
motivated, has a strong interest in robotics and has a good analytic background.
On this basis I would like to invite him as visitor without
support from AI, and with no commitment about office space.
I will ask for office space from Mechanical Eng or Aero and
Astro if none is available here. Failing that, he would have to
do without office space. Are you willing to approve this on
those terms?
∂17-Aug-79 0043 BIS Languages
John,
I know French and German, and I'm planning to learn another language soon.
Reasonable alternatives seem to be Chinese, Japanese, and Russian. Considering
that I work in Robotics, which would you recommend?
Thanx,
Barry
If you can stick it out, Japanese.
∂17-Aug-79 0922 DEW
YES, I'll give Filman's paper
∂17-Aug-79 1526 DPB Teaching assignments for 79-80
To: "@FAC.DIS[1,DPB]" at SU-AI
The following message will reach (most of) you in hard copy in a few days.
For those of you who prefer the online version, check COURSE[CSD,DPB].
If you want a hard-copy, and don't get one by next Friday, remind me
to send you one.
Enclosed are two lists of course assignments for the coming academic
year. One is organized by course; the other is organized by individual.
Included are all of the CSL courses. The assignments for the CSL
courses may be out of date. When Sally gave me the information she warned
me that the information might not be the latest. Please send corrections
to me and I'll keep the list up to the minute. The lists can be found in
COURSE[CSD,DPB] at SAIL. If I get a request to put up a copy of the file
at SCORE, I'll do so.
There are still a few un-staffed courses. Most of the openings are in
the introductory programming courses and in CS111/EE181. Suggestions
and volunteers are solicited.
It is time to begin recruiting TAs. If you have already chosen a TA
for your course, please notify me and supply details (per cent time etc.).
If you want me to help find a TA, send me your hiring criteria and I'll
try to send some live bodies for your inspection.
∂18-Aug-79 1419 TOB
To: JMC, FB, LES
∂16-Aug-79 2025 JMC
To: TOB
CC: FB
No problem Giralt visit under stated conditions.
John
Whelan to visit. Giralt was reference.
Tom
ok, Macvicar visit then.
∂20-Aug-79 1400 Bmoore at SRI-KL (Bob Moore) AI-Phil Workshop
Date: 20 Aug 1979 1402-PDT
From: Bmoore at SRI-KL (Bob Moore)
Subject: AI-Phil Workshop
To: pjh at SAIL
cc: jmc at SAIL
Ostensibly, I am responsible for ogranizing the workshop, but
(1) I have been too busy finishing up various projects at SRI to
devote much time to it.
(2) I feel rather hesitant about quite how it should be organized,
and I have been reluctant to plunge ahead on my own on the basis
of the relatively small amount of input I have received so far.
In view of this situation and the fact that we will all be at the Center
fairly soon, I intend to put off doing anything more (which is to say
anything at all, really--i.e. no one has been invited yet) until the
group assembles in September. This may require postponing the meeting,
say until January, but I don't think there was ever anything sacred about
having it in the Fall. If this workshop is going to come off, it will have
to be a group effort, and I will look forward to working with everyone on it
as soon as we can all get together. Let me know when you arrive, an we can
get together to discuss it.
Bob
-------
∂20-Aug-79 1512 BOBROW at PARC-MAXC2 (Response to message)
Date: 20 AUG 1979 1513-PDT
From: BOBROW at PARC-MAXC2
Subject: (Response to message)
To: JMC at SU-AI
In response to your message sent 17 Aug 1979 1910-PDT
I just received your message about the paper. Fine about finishing it up when you get back. Hope it can be done soon.
danny
-------
∂21-Aug-79 2342 MARCF at MIT-MC (Marc K. Furon)
Date: 22 AUG 1979 0244-EDT
From: MARCF at MIT-MC (Marc K. Furon)
To: JMC at SU-AI
THE COMPUTER SCIENCE ASSOCIATION AT CSUN SAW THE LAST ISSUE OF BYTE
MAGAZINE (AUGUST 1979) -- DEDICATED TO LISP. SOME USERS SEEMED TO
THINK ABOUT PUTTING IT ON THEIR MICRO. THIS BROUGHT TO MIND A QUOTE
YOU MADE WHEN YOU SPOKE HERE SOME TIME AGO (CORRECT ME IF I'M WRONG):
"Q. WHAT DO YOU THINK THE POSSIBILTITIES ARE FOR LISP ON A MICRO?"
"A. WELL, LISP TENDS TO USE A GREAT DEAL OF MEMORY. I DON'T
KNOW IF IT WOULD BE PRACTICAL. I WONDER WHY ANYONE WOULD WANT
LISP ON A MICRO ... BUT THEN AGAIN, I DON'T KNOW WHY ANYONE WOULD
WANT BASIC ON A MICRO EITHER."
WE WERE THINKING ABOUT WRITING A 'LETTER TO THE EDITOR' ABOUT LISP
SYSTEMS FOR MICROS, AND WE WOULD LIKE TO KNOW IF YOU WOULD MIND US
QUOTING THE ABOVE. WE REALIZE THAT IT WAS MEANT IN A MILDLY
HUMOROUS VEIN, BUT THERE SEEMS TO BE A GRAIN OF TRUTH IN IT...WOULD
BE INTERESTING TO LET OTHER PEOPLE IN ON IT.
IF YOU HAVE ANY COMMENTS, WE WOULD APPRECIATE THEM - YOU MAY SEND
THEM TO THIS NET ADDRESS (MARCF AT MIT-MC).
THANK YOU.
MARC FURON
VP, COMP. SCI. ASSOC.
PS. HAPPY BIRTHDAY!
I certainly did say that Lisp requires considerable memory and would
be inclined to update my remark to say that Lisp has good prospects
on the new 16 bit micros with large address spaces. I would prefer
not to joust with Basic.
Therefore, a suitably adjusted comment would be best, but
I suppose I can't object to taking responsibility for the remark I
made at the time. Best regards. John McCarthy
∂22-Aug-79 1244 LES
∂22-Aug-79 0938 the tty of Geoffrey S. Goodfellow FYI -
Date: 22 Aug 1979 0933-PDT
Sender: GEOFF at SRI-KA
Subject: FYI -
From: the tty of Geoffrey S. Goodfellow
To: BPM at SAIL, LES at SAIL
Message-ID: <[SRI-KA]22-Aug-79 09:33:59.GEOFF>
Reply-to: Geoff @ SRI-KA
As of September 1st, Dr. Robert E. Kahn will become Acting
Director of ARPA-IPTO.
Col. Dave Russell will retire from the Army the end of August
and will join Satellite Business Systems.
∂22-Aug-79 1302 ARR Test programs
To: S1 at SU-AI
I have just created the PPN [TST,S1], and put there the test PASCAL programs
that I am using to debug UPAS. I think it is a good idea if we all have our
test files together, so that when someone needs a test file, he/she does not
have to go around searching all over the place. It should also prevent
duplicated effort. Please move your test files to [TST,S1]
Armando.
∂23-Aug-79 1823 ''Richard Kovalcik, Jr.'' <RK at MIT-MC> forwarding this ...
Date: 23 August 1979 21:21-EDT
From: "Richard Kovalcik, Jr." <RK at MIT-MC>
Subject: forwarding this ...
To: s1 at SU-AI
MSG: WHAT 1
DISTRIB: *CMU, *DM, *MC, *ML, *AI
EXPIRES: 09/13/79 10:21:54
LARSWE@MIT-AI 08/23/79 10:21:54 Re: what is the s-1?
Does anyone know of a journal article, accessible proposal, or file
that says what the S-1 machine project at Stanford is all about? Pls
reply to ERICSON@ECL.
∂23-Aug-79 1848 ME death of a record WAITS system
To: BH at SU-AI, JBR at SU-AI, MRC at SU-AI, ROB at SU-AI,
ROY at SU-AI, DON at SU-AI, LES at SU-AI, JMC at SU-AI
This afternoon a record-uptime WAITS system finally came to the end of its
road when someone blew a breaker by plugging in two TI computer disk drives
on the same circuit as various other equipment, including the BBN pager,
which went down and hung the memory, which hung the KL-10 microcode. The
new record is 377:02, only one day short of 400 hours. The record uptime
exceeded two weeks (336 hours) for the first time. The old record was
302:xx. The system is dead -- long live the system.
∂23-Aug-79 1907 Mark Crispin <Admin.MRC at SU-SCORE> All good things must come to an end...
Date: 23 Aug 1979 1902-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: All good things must come to an end...
To: ME at SU-AI
cc: BH at SU-AI, JBR at SU-AI, Rob at SU-AI, Roy at SU-AI, Don at SU-AI,
Les at SU-AI, JMC at SU-AI
Marty -
That is indeed an outstanding accomplishment. You might well hold a
Stanford record for PDP-10 uptime. I know the business school has
cracked 300 hours on their 20, but how much further I don't know. I
doubt SUMEX or IMSSS have had uptimes of that order, and I'm not sure
about LOTS. My record is around 225 hours, but that system was taken
down. I've only had one crash this month, and that was hardware.
When we're both in Jacks it'll be neat to have an uptime competition!
XEROX PARC must have the record on their MAXC - over nine months.
Are you sure about nine months. The bulletin board notice said
2100 hours which is about three months.
Mark
-------
∂24-Aug-79 1424 DPB office plans for Fall Quarter
Do you have definite plans for your Jacks office for Fall Quarter? Do
you intend to use it? Do you have a visitor or such who you want to put
there in your absence? If you have three "no's" to these questions, I will
try to find an appropriate visitor to put there, subject to your approval.
I expect that you will want to "move in", but you may want to sublet the
space to the right person. -Denny
Since Carolyn will be teaching CS206 in the Fall, I thought she might use it.
∂24-Aug-79 2230 PEG Documents
To: S1 at SU-AI
1. The file ACKNOL.TXT[DOC,S1] contains a standard acknowledgement paragraph
(with contract numbers, etc.) for S-1 documents.
2. The file S1DOC.TEX[DOC,S1] contains a set of macros for the TEX typesetting
program designed for use for S-1 documents. The file SOPADO.TEX[DOC,S1]
is intended to be an example of the use of these macros; PAIL6.TEX[DOC,S1]
will also serve as such.
Phil Gerring
∂25-Aug-79 1525 BIS DDs in Jacks Hall
To: LES
CC: ARG, JKS, JMC, MSM, TOB
The troops and I would like to know what's becoming of the 64 (?) DD terminals
currently in Jacks. There is considerable feeling that the Arm Room should
have a second DD terminal so that two groups (e.g. CS227 studs and full-time
research workers) can work at the same time. After all, we're going to have
four arms in that room shortly, and that will give considerable impetus to
designing systems for controlling two separate assembly tasks. In any case,
we're curious about who's getting the DD terminals in Jacks.
∂25-Aug-79 1700 MARCF at MIT-MC (Marc K. Furon)
Date: 25 AUG 1979 1959-EDT
From: MARCF at MIT-MC (Marc K. Furon)
To: JMC at MIT-MC
~
∂25-Aug-79 1703 MARCF at MIT-MC (Marc K. Furon)
Date: 25 AUG 1979 2001-EDT
From: MARCF at MIT-MC (Marc K. Furon)
To: JMC at SU-AI
THANKS FOR YOUR COMMENTS. I DOUBT WE'LL GO AHEAD WITH THE LETTER,
BUT NONETHELESS, IT WAS NICE TO HAVE YOUR IDEAS.
ENJOY -- MARC
∂25-Aug-79 1703 ARR File name extension for PASCAL* sources.
To: S1 at SU-AI
The extension .PST will be used for source programs that are written in
PASCAL* and are not valid standard PASCAL programs.
Armando.
∂26-Aug-79 1353 Wiederhold at SUMEX-AIM Re: File name extension for PASCAL* sources.
Date: 26 Aug 1979 1344-PDT
From: Wiederhold at SUMEX-AIM
Subject: Re: File name extension for PASCAL* sources.
To: ARR at SU-AI, S1 at SU-AI
In response to the message sent 25 Aug 1979 1703-PDT from ARR at SU-AI
PSTn looks like P-standard - yach. Why not P* - yes , I know
why, but that shows a problem with the source name.
prehaps PSU for stanford. On machine other than sail the extension
could indeed be PSTAR Gio
-------
∂26-Aug-79 1930 PEG Documents
To: S1 at SU-AI
∂24-Aug-79 2240 ARR re: PEG Documents
To: PEG
CC: ARR
I have no intention to learn how to use TEX, at least not yet. Anyway,
I have a question: are you going to write a DOC-style file, describing
the use of the macros? or, do the macros themselves contain any comments
on their usage? I understand that the idea is to get all the S-1 documents
to look more or less the same, so some kind of a standard on their use
should be defined.
Also: is there anything in INDEX that points to these files? I remember
there is a page called 'help with documents'.
Armando.
[ 1. I may or may not write a DOC-style file, depending on time constraints.
2. The macros are commented, though rather tersely at this point. The
comments may well improve at some point.
3. S1DOC produces an SMA-3 style format; this is, at best, an informal
standard (i.e., I agree, something needs to be defined).
4. Yes, the file INDEX[DOC,S1] contains pointers to all files involved.
Phil. ]
∂27-Aug-79 1157 ARR Wiederhold Re: File name extension for PASCAL* sources.
To: S1 at SU-AI
∂26-Aug-79 1353 Wiederhold at SUMEX-AIM Re: File name extension for PASCAL* sources.
Date: 26 Aug 1979 1344-PDT
From: Wiederhold at SUMEX-AIM
Subject: Re: File name extension for PASCAL* sources.
To: ARR at SU-AI, S1 at SU-AI
In response to the message sent 25 Aug 1979 1703-PDT from ARR at SU-AI
PSTn looks like P-standard - yach. Why not P* - yes , I know
why, but that shows a problem with the source name.
prehaps PSU for stanford. On machine other than sail the extension
could indeed be PSTAR Gio
-------
[[OK: so, let's change it: it will be PSU at SAIL. Armando.]]
∂28-Aug-79 0859 ARR S-1 meeting.
To: S1 at SU-AI
Chronologically, we should have am S-1 meeting tomorrow, but apparently there
aren't that many of us around. If you are going to be here tomorrow, wednesday
and/or you have a topic to discuss, please let me know. I'll send a message this
afternoon confirming whether we have the meeting tomorrow or postpone it for
next wednesday.
Armando.
∂28-Aug-79 1813 ARR meeting postponed
To: S1 at SU-AI
Due to the overwhelming response to my last message, the S-1 meeting that would
be held this week is postponed until next week, wednesday, 5-sep-79, at 1:30 PM
in the small conference room, at SAIL.
Armando.
∂29-Aug-79 0617 LLW No Work, No Eat!
To: S1 at SU-AI
CC: LLW at SU-AI
∂28-Aug-79 1813 ARR meeting postponed
To: S1 at SU-AI
Due to the overwhelming response to my last message, the S-1 meeting that would
be held this week is postponed until next week, wednesday, 5-sep-79, at 1:30 PM
in the small conference room, at SAIL.
Armando.
[Pay for this week likewise, I presume? Lowell]
∂29-Aug-79 1459 ARR LLW No Work, No Eat!
To: S1 at SU-AI
∂29-Aug-79 0617 LLW No Work, No Eat!
To: S1 at SU-AI
CC: LLW at SU-AI
∂28-Aug-79 1813 ARR meeting postponed
To: S1 at SU-AI
Due to the overwhelming response to my last message, the S-1 meeting that would
be held this week is postponed until next week, wednesday, 5-sep-79, at 1:30 PM
in the small conference room, at SAIL.
Armando.
[Pay for this week likewise, I presume? Lowell]
[[You mean, money? I don't know. I am just following instructions: I was told
by Gio that he could not make it, that a large, undetermined (but determinable)
number of people were out on well-deserved vacation, in other places of the
planet, and that I SHOULD DO EXACTLY AS I DID, namely, check to see if anyone
had anything important or urgent to discuss, and if not, postpone the meeting.
What is the point on having a meeting with four people, when the group is a
lot larger? One of the things I like about the S-1 project is precisely that
it is a flexible, people-oriented system. I am not going to be dragged to a
meeting just for the sake of punctuality, when I can use that time producing
something else that I can report in the meeting next week. This is not the first
time this happens: exactly three weeks ago, the meeting was also postponed
for one week, and then nobody, not even you, complained. And if you were so
interested on our having the meeting, why didn't you reply to my first message,
in the first place?
I hope your message was a joke. If, on the contrary, you do have specific
complains about the people in the project, you should use the established means
of communication to express them specifically. This is too vague a statement to
produce anything constructive.
No bad feelings. Just making the point.
Armando.]]
∂30-Aug-79 0906 PTZ SIGPLAN Compiler Conference Proceedings
To: S1 at SU-AI
...are in the S-1 library at SAIL. Papers whose talks I thought were more
interesting (or more relevant) than average are marked with * in the table
of contents.
Polle
∂30-Aug-79 1130 CET Student support for Autumn Quarter
To: "@FAC.DIS[1,DPB]" at SU-AI
CC: BS at SU-SCORE, CMA at SU-SCORE
Please send information on your plans for student RA and/or TA
support for Autumn Quarter to me as soon aspossible. Student
appointments must be prepared immediately.
Information on future quarters would be helpful also.
Thanks,
Carolyn
∂31-Aug-79 1609 TOB
Visitors
John Scull, JPL, has been invited for a year by the Aero and Astro
department. He has managed JPL's robotics program and has a strong
interest in robotics. He expressed interest in working with us
part of the time during his visit.
He would visit 1979-80.
Bob Cannon, head of Aero and Astro, suggested that Computer Science
might write a letter inviting him as a joint invitation. Aero and
Astro will provide space; support is not required.
We would offer use of the robotics laboratory and computers (including KL10).
I would like to make the invitation. I am asking:
1. for authorization
2. Who should write the letter?
I will provide a copy of the letter from Aero and Astro
Dr. Yoram Koren, Technion Israel, has written to request a year's visit
with the robotics group. I spoke to Bob Cannon who said he was interested
and would provide space, provided Koren's plans fit in some way with
his experimental program.
He would visit Sept 1980.
Dr. Koren has 8 years beyond the PhD. He has a background in control,
mechanical engineering, and computers. He has worked extensively with
NC and computer controlled fabrication, also on a prosthetic knee.
Bernie Chern, NSF, visited there and talked with him extensively.
Bernie thinks highly of Koren, who he thinks is bright, serious,
hard-working, and has good judgment. Koren wants to break into robotics
and Chern suggested strongly that he come to Stanford if possible.
I am interested in his working with us. I think that his background
would be useful as background for our work in industrial robotics,
and that he could contribute to the robotics experimental system.
Koren seems perfect for the things we are beginning with Aero and
Astro.
I have written to know in detail what he plans. I plan to evaluate
carefully whether he will contribute to the research program.
I am asking for authorization to invite him, probably jointly with
Aero and Astro, provided:
1. the evaluation is that he would contribute
2. that no space and no support are required.
Use of robotics laboratory facilities and computers (including KL10)
would be provided.
∂31-Aug-79 1755 CLT MTC Qual.
To: CET, JMC, ZM
I will sign up for the MTC Qual the fall when I see some evidence of
organization on the part of the examining committee. In particular
I would like to see a carefully thought out reading list.
∂01-Sep-79 0759 PN Yet another S-1 area: [CSP,S1]
To: S1 at SU-AI
[CSP,S1] is designed for the miscellaneous utility programs that are
beginning to collect, such as VERCH, PTRANS, and PAS. (For more info
about any of these see page 5 of INDEX[DOC,S1].) CSP stands for Current
Source Programs.
A plea: if I can remain the chief maintainer of all of these (as well
as the interpreter), rather than everyone modifying them at random, life
will be ultimately easier for everyone, since a lot of people use them
and depend on their reliability.
∂02-Sep-79 0000 CLT
∂01-Sep-79 2353 JMC
See EQUALI[E79,JMC].
fine. You decide where
∂02-Sep-79 1949 PN S-1 meeting
To: S1 at SU-AI
There will be an S-1 meeting this wednesday, 5-sep-79, at 1:30 PM in
the small conference room, at SAIL.
I am interested in discussing some moving questions that have been
bothering me. They run in two general lines. They are: how much worth it
is it to put up what software at Score? The other one is, what
connections to LLL are needed? To make such decisions, the most important
thing to know is what people are trying to do, in general, and which of
these can be done before the move and which during the move. If everyone
can think a little about this before the meeting, we might get somewhere.
∂03-Sep-79 1810 HPM Space on $5 a day
The intrepid asteroid colonists are from O'Neill's book (as Dyson mentions).
O'Neill presented them as a family that found life in the big space
colonies too confining. One of the massiest items in their low budget
trek to the belt was a small radiation shelter to hunker down in
during solar flares. They made it (and the other pressure vessels)
by blowing molten metal bubbles. They take along a machine shop,
computers and asteroid farming equipment with a total cost of about
$100,000. Of course they have an advantage we don't: They are already
in space from the start.
∂04-Sep-79 0046 MRC via SU-SCORE HBTY
Happy Birthday to you
Happy Birthday to you
Happy Birthday dear JMC
Happy Birthday to you!
Mark
∂04-Sep-79 0942 MAX
Happy Birthday to You!
∂04-Sep-79 1006 LJH Today!
I sincerely hope that you have a very pleasant day.
∂04-Sep-79 1501 ROY on TTY107 (at TV-104) 1501
Your Imlac channel has been shifted over to the dca scanner and I have
verified thatthere is a carrier present from your modem. When you have
time, call me and we can check out its operation further. Roy
∂04-Sep-79 2326 Mark Crispin <Admin.MRC at SU-SCORE> news
Date: 4 Sep 1979 2218-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: news
To: JMC at SU-AI
1. Your COMAND.CMD loses because it has the command SET EDITOR EMACS in
it. This command is obsolete, and has been replaced by the command
DEFINE EDITOR: SYS:EMACS
You can thank Digital Equipment Corporation for that great idea.
2. About Dialnet. Well, the business school is now going through its
twelveth (or is it sixteenth?) week of downtime, with sporadic periods
of being up. DEC finally gave in and gave GSB their old processor
back, but things still aren't settled down. Needless to say, GSB is
regretting the day they first did business with DEC.
Moving to SCORE is out of the question right now. SCORE is out of
TTY ports, moving in less than two weeks, and most importantly is
running an experimental monitor (Tops-20 release 4) which is itself
is a source of problems. In addition, SCORE still has serious memory
problems and its magtape drive is down, meaning no system backups.
That leaves LOTS. LOTS too is out of TTY ports and is also suffering
with memory problems, but still is in a better shape right now than
either GSB or SCORE. Maybe you want to persuade Ralph into freeing
up a port at LOTS for Dialnet. There is an advantage to LOTS in that
there are a lot of kiddies who would be interested in writing user
software for Dialnet; but the cost of a TTY port may be too great.
Mark
-------
Thanks for the news. Is it possible to phone you at MJH or for you to
phone me.
∂05-Sep-79 0603 Mark Crispin <Admin.MRC at SU-SCORE> Proposal for HP
Date: 5 Sep 1979 0559-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Proposal for HP
To: G.Eldre at SU-SCORE
cc: JMC at SU-AI
Tim -
As you may or may not know, GSB has been down for almost 12 weeks with
DEC's bitage of the bag. This has really thrown a wrench in Dialnet
development. SCORE and LOTS both aren't suited for Dialnet hacking;
SCORE because it has experimental release 4, and LOTS because the kiddies
are coming (not to mention the fact that neither have any ports).
I remember you telling me you were ordering a VADIC. What do you think
of Stanford lending you the VADIC which is currently at GSB? I don't
see any downtime necessary at HP, modulo getting it into your monitor
and finding the channel-6 looping bug.
Are you interested?
Mark
-------
∂05-Sep-79 0859 g.eldre at SU-SCORE (Tim Eldredge) Re: Proposal for HP
Date: 5 Sep 1979 0853-PDT
From: g.eldre at SU-SCORE (Tim Eldredge)
Subject: Re: Proposal for HP
To: Admin.MRC at SU-SCORE, jmc at SU-AI
cc: g.eldre at SU-SCORE
In-Reply-To: Your message of 5-Sep-79 0559-PDT
I am interested, but I will have to discuss it with my management
before I can give you an OK. I'll get back to you as soon as I
can. Sounds like a winning idea to me.
Tim
-------
∂06-Sep-79 1109 g.eldre at SU-SCORE (Tim Eldredge) Dialnet at HP
Date: 6 Sep 1979 1044-PDT
From: g.eldre at SU-SCORE (Tim Eldredge)
Subject: Dialnet at HP
To: admin.mrc at SU-SCORE
cc: jmc at SU-AI
Has been approved. Call me and we will work out the details.
Tim
-------
∂06-Sep-79 1232 100 : patte New graduate student
Bill Griffiths, a new graduate student in AI, would like to talk to
you. Please call him at 851-1847.
∂07-Sep-79 1128 PAT future plans
I would like to confirm what Les told you regarding my staying here with
the Music group. Since this opportunity has come up, I would like to
take it. It will mean, both a promotion and substantial raise in pay and
would give me an opportunity to pursue my music interests. I have enjoyed
working at the AI Lab and with you. Since you will be on leave for most
of this year, now seems a good time to make a change. We have not set a
firm date, but October 1 seems likely as they will need someone here to
take care of things. We will need to decide what you want to do about a
secretary while you are on leave as there is some training that should be
done. I will be available to help if there is a need. Hope you enjoy
your study at CASBS and many good wishes for the future. Patte
∂07-Sep-79 1301 Wiederhold at SUMEX-AIM Information on Stanfords S-1 efforts
Date: 7 Sep 1979 1248-PDT
From: Wiederhold at SUMEX-AIM
Subject: Information on Stanfords S-1 efforts
To: larswe at MIT-AI, ericson at ECL
cc: s1 at SAIL
In response to the message sent 06 Sep 1979 1316-PDT from PEG at SU-AI
The only articles I know of are about the SCALD design system -
contact TM@sail.
Our proposals are at SAIL, mine is S1PROP.7980 under [1,GIO] if it hasn't been
archived, earlier ones are archived. Contact JLH for his proposals.
There are a fair number of CSL reports out, and available on request, sometimes
with request for copy reimbursement.
This includes PCFORT,PCODE,UCODE,Interpreter,loader-linker,old overview, and
detailed architecture,macro preprocessor, ...
Any report that's oyt isapt to be obsolescent.
Contact PICKERING@SUMEX for requests of reports, she'll bill you if neccessary.
Gio
-------
∂07-Sep-79 1936 FB talk
I need to talk to you about Hersche Allen.
Forest
∂08-Sep-79 2134 Mark Crispin <Admin.MRC at SU-SCORE> VADIC 1200 baud modems
Date: 8 Sep 1979 1848-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: VADIC 1200 baud modems
To: Rindfleisch at SUMEX-AIM
cc: JMC at SU-AI, Les at SU-AI
Tom -
I understand you have some VADIC 3400 series 1200 baud full duplex modems
which aren't being used currently. If this is the case, the Dialnet
project can put a few to good use, at least until you want them back for
something else. Even if they don't have an autodialer on them we can
still use them as Dialnet servers.
The problem is, the Dialnet project only has two modems. A third is
supposed to be on order but nobody seems to know when it will be delivered.
One modem is more or less permanently installed at SAIL, and the other sort
of floats around to whatever Tops-20 system Dialnet is being debugged on at
the time. I would like to have two Tops-20 systems running with Dialnet so
I can determine if a communication problem in dialing from Tops-20 to SAIL
is a SAIL or Tops-20 bug or phone line problem or what (I think it's a SAIL
bug).
In addition, there are a bunch of people who desperately need a communication
link to LOTS from SAIL, and even a receive-only modem would serve that
purpose.
All told, two modems would suit my needs just fine, although I could find
good homes for as many as you're willing to give me.
Regards,
Mark
-------
∂09-Sep-79 0007 MFB
ZOHAR SUGGESTS THAT I GIVE A SEMINAR BEFORE SCHOOL STARTS.
SEEMS LIKE A REASONABLE IDEA TO DO IT AT JACKS. WHAT DO YOU THINK?
WHEN IS A GOOD TIME DURING THE WEEK AFTER NEXT?
∂09-Sep-79 0104 Mark Crispin <Admin.MRC at SU-SCORE> Re: your message to Rindfleisch
Date: 8 Sep 1979 2258-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Re: your message to Rindfleisch
To: JMC at SU-AI
In-Reply-To: Your message of 8-Sep-79 2246-PDT
Well, I guess you didn't know that last Wednesday I brought up the
Stanford Tops-20 monitor up on Sumex's 2020 in one night, without
any problems on the way; and that I'm preparing a care package of
all the good versions of Tops-20 software we have here at SCORE and
LOTS. The 2020 hasn't crashed yet, and they were suitable impressed
by my bringing our stuff up there single-handedly in one night.
So, since he owes me a favor, I thought I'd be in a good position to
ask in a friendly manner about the rumored modems he has (I really
don't know for sure that they exist, or if they have already been
claimed, or if there isn't some restriction imposed by their source
on their use). I didn't think I was being demanding, but I didn't
see a reason to beg and put him on the spot.
Needless to say, part of the reason I was so helpful to them on their
2020 system was the modems! I plan such things more than you think!
Mark
-------
∂09-Sep-79 0842 HPM SUCCESS (of sorts)
To: TOB at SU-AI, JMC at SU-AI, LEVINTHAL at SUMEX-AIM, BIS at SU-AI,
SL at SU-AI, DBG at SU-AI, ROD at SU-AI
The cart avoided its first obstacle today. In fifteen lurches of about
a meter each, it rounded a chair in the lounge. It decided to back up
twice during this process to get enough maneuvering room.
The first ten moves were just about perfect. It saw all the important
obstacles and generated good path plans, and executed its moves correctly,
though a little sloppily. In the 11'th move I began to worry, as the
diagram of the planned path took it very near the second obstacle
(another chair behind the first). By the twelfth move it was obvious
that the program was planning to run down the second chair. I aborted
it before the collision.
I understand the problem; I'd left part of the obstacle path planning code
unfinished. The planned paths consist of tangents between circular
obstacles, and arcs around the obstacles between the tangents: the
part of the code that tests for obstructions on the arcs was left
uncompleted, and the lack bit me on the second obstacle today.
The vision worked very well. The program found about a dozen seperate
features on each chair, and developed a good idea of the size and
extent of their obstructiveness. It picked up the features
most effectively when it was about 3 meters in front of each chair.
I will work on the planning problem, and a few less serious foibles
and try again during the next week. If it performs adequately well,
I'll take it outside a subsequent morning and try to get film
of an interesting outdoor run. At that point I'll declare my thesis
research done.
Sounds good to me.
∂10-Sep-79 0121 CLT pub
.pub provin
while aliased to [E79,JMC] should pub the PROVIN file on [E79,JMC]
∂10-Sep-79 0857 Rindfleisch at SUMEX-AIM Re: VADIC 1200 baud modems
Date: 10 Sep 1979 0851-PDT
From: Rindfleisch at SUMEX-AIM
Subject: Re: VADIC 1200 baud modems
To: Admin.MRC at SU-SCORE
cc: JMC at SU-AI, Les at SU-AI, ADMIN.GORIN at SU-SCORE, CMILLER
In response to your message sent 8 Sep 1979 1848-PDT
Mark,
The Vadics we are holding have some administrative constraints on their
use. They are surplus from NIH-DCRT and are intended for medical
applications. We are acting as the west coast depot for NIH to
distribute modems as authorized. I have discussed previously with NIH
the possibility of using some of these modems with SCORE, based on the
connection between HPP work on SUMEX and SCORE. The answer was that
we could if there were no other direct medical needs for the hardware.
I think we have waited long enough for immediate NIH needs to be
satisfied and we can let you borrow several modems -- these would
remain NIH property and be subject to recall if needed for other medical
projects. In order to meet our needs in accounting for the hardware,
these modems should be physically installed only on SCORE. If LOTS,
GSB, or other sites need modems, they should borrow from available
SCORE modems and SCORE should replace these with SUMEX modems. I can't
justify NIH equipment being used at sites other than SCORE.
Let me know if you are still interested with these strings ...
Tom R.
-------
∂10-Sep-79 1021 Wiederhold at SUMEX-AIM Destination for SMA-3 Manual and architecture problems
Date: 10 Sep 1979 1017-PDT
From: Wiederhold at SUMEX-AIM
Subject: Destination for SMA-3 Manual and architecture problems
To: s1 at SAIL
cc: sjc at SAIL
Mail from SU-AI rcvd at 9-Sep-79 0945-PDT
Date: 09 Sep 1979 0948-PDT
From: Steven Correll <SJC at SU-AI>
Subject: SMA3.CHG
To: GIO at SU-AI
I'm editing SMA3 to produce SMA4 for the next S1 annual report, and I
will update SMA3.CHG whenever I hear of problems in it. Since I'm not
by any means expert on the S1 architecture, I definitely invite comments
and corrections. Send them to S1ARCH if they're of general interest, or
directly to me if they deal purely with grammar and syntax.
←←←←←←←←←←←←←
please make sure all found problems go then to Steven / SJC
Gio
-------
∂10-Sep-79 2045 SKU SARAH CALLED....
BRING HOME A 9 VOLT BATTERY...
HAVE FUN SENDS STEVE
∂11-Sep-79 0905 LCA
I am applying for a position at the (italian) University as a teacher
of Computer Science. My application will be mailed in a week or so.
I'll include a letter of presentation-reccomendations written by the
Director of my Institute in Pisa, one by RWW and one by Robin Milner.
A letter from you ("to whom it may concern", or something like) could
be very helpful. Do you mind writing one?
I apologize for bothering you. Thanks a lot, Gigina
∂11-Sep-79 0901 JK via SUMEX-AIM
Two things: (1) I took the liberty of mentioning to Lee Blaine (who,as you
may know, is the head of the set-theory project at IMSSS) about the
seminar at the Behavioral Sciences Center.He would like to come.Can he?
(2) aA thought concerning our discussion on functions in the elephant
language: It seems to me that the mu-operator is in some sense already in
the language(given conditional terms); namely consider the following
suggestion: First , add a new clause for building terms: If t and u are
terms so is t(u).(This is trivial but necessary.) Now one can introduce
functions using the following schema: Assume P(x,y) is a predicate such
that for all x there is a unique y with P(x y). Then one may introduce the
corresponding function f as a term thru the axiom: f(x)=If P(x,y) then
y else f(x). The varaible y is a kind of a temporary variable in this
statement.
Many thanks for your suggestion. As is, it isn't what I want, because
introducing the f is conditional on first proving that there is a unique
y. However, tinkering with it led to an even simpler solution to
my problem. Given any wff P(x,y) with x and y free and a previously
unused function symbol f, I can write (∀x)((∃y)P(x,y) ⊃ P(x,f(x))) as
a kind of partial definition of f and be guaranteed against contradiction.
For any values of x for which I can prove (∃!y)P(x,y), f(x) will be
determined.
In case the terminal on which you read this doesn't have logical symbols,
the formulas are (for all x)((there exists y)P(x,y) implies P(x,f(x)))
and (there is a unique y)P(x,y).
You are probably saying, "had I known that was all he wanted ...".
∂11-Sep-79 1657 MFB seminar
ill give a talk describing my work next thursday, sept 20,
at 2:00 in rm 402 a jacks hall.
if that time is not good with you ill change it.
please let me know right away.
martin
Monday, Wednesday and Thursday afternoons are out for me.
∂13-Sep-79 1125 CET
To: "@FAC.DIS[1,DPB]" at SU-AI
9/22 "See you in September Party" at Carolyn and Joe Tajnai's
8:00 p.m. -->
1532 Emperor Way
Sunnyvale
738-2376
We will provide beer and sangria.
If you want anything else to drink (i.e. wine or soft drinks,please bring them).
Also, please bring munchies. It is helpful if you bring food already in a
serving dish.
∂13-Sep-79 1223 CLT cs206 ta
To: DPB, CET, JMC
Martin Brooks will be my TA for cs206 this fall.
∂13-Sep-79 1254 DRA Article?
Would an upgraded and expanded version of AIM 321 be suitable
for publication in Computer Magazine? I'd like to publish an
overview paper of the field. Let me know if you're interested.
dra
I will see if someone is interested in doing the upgrade.
∂13-Sep-79 1541 Ijcai79 at SUMEX-AIM CS225 Projects
Date: 13 Sep 1979 1538-PDT
From: Ijcai79 at SUMEX-AIM
Subject: CS225 Projects
To: AI group:
cc: ijcai79
AI students taking CS225 this fall will again be asked to
design and implement a program. I want to give them a choice of
persons to work with and projects to work on. Please respond
telling me how many students you would like to interest in
working with you and briefly describe what they might do.
The 225 project sometimes is a forerunner of a significant thesis,
so this is an opportunity for you as well as for the students.
Thanks. Bruce Buchanan
-------
I am on leave this fall, but might be willing to supervise one very
competent student interested in epistemological problems.
∂13-Sep-79 1543 DCO Yapp - Yet Another Pretty Printer
To: "@LISP.DIS[LSP,DCO]" at SU-AI
I have written a general purpose pretty printer. Primitives are
provided for pretty printing; the primitives are driven by an unparser
which you must write yourself. The pretty printer is suitable for pretty
printing programs (Lisp, Pascal, whatever), formulas, etc. It does not
wait for the entire expression or program to be input, but begins printing
as soon as it has received a linefull of text. It requires time linear in
the length of the input and space linear in the linewidth of the output
device.
The lisp version of the algorithm is PP.FAS[LSP,DCO]. Documentation
is only available in hardcopy; send me a note if you want documentation.
∂13-Sep-79 1609 MFB TALK
HOW ABOUT FRIDAY AT 1 OR 2 (NEXT WEEK)?
ok, please confirm it
∂13-Sep-79 1607 CET
To: "@FAC.DIS[1,DPB]" at SU-AI, jrg at SU-SCORE, don at SU-SCORE
∂13-Sep-79 1601 Mcgoveran at SUMEX-AIM Faculty Meeting
Date: 13 Sep 1979 1557-PDT
From: Mcgoveran at SUMEX-AIM
Subject: Faculty Meeting
To: CET at SAIL
cc: McGoveran
The first faculty meeting of the 1979-80 school year will be on September 25
at 2:30pm, Jacks Hall 402. The first item on the agenda is the approval of
degree candidates.
If there are other items you would like to have on the agenda, please contact
either me or Denny Brown.
I hope that you will be able to attend this meeting. Please let me know
whether you plan to be there.
Mary McGoveran
McGoveran@SUMEX
7-3264
-------
∂13-Sep-79 2007 FB via SU-SCORE Topics for faculty meeting
To: "@FAC.DIS[1,DPB]" at SU-AI, jrg at SU-SCORE, don at SU-SCORE
I would like to have two related items added to the agenda of our first
fall faculty meeting. At our last "black Friday" meeting we voted to
put a hold on the winter registration of some students who had apparently
refused to file their G81 forms indicating their topic and reading committee.
The stink this caused among the students made me reflect on this action and
regret it as inappropriate. Thus I would like to propose that we remove the
hold on their winter registration. Related to that I would like us to
consider a revision of our G81 filing requirement policy in such a way that
clearly indicates that a student's thesis advisor has the authority (and
obligation) to indicate that a reading committee and topic are being
adequately persued by a student.
I would have had this information put in the first announcement if I had
known about the first announcement of this meeting.
Forest
∂13-Sep-79 2036 WAXMAN at USC-ECL REQUEST FOR PAPER
Date: 13 SEP 1979 2032-PDT
From: WAXMAN at USC-ECL
Subject: REQUEST FOR PAPER
To: MCCARTHY at SAIL
cc: WAXMAN
PROF. MCCARTHY,
CAN I GET A COPY OF YOUR PAPER ON "CIRCUMSCRIPTION INDUCTION- A
WAY OF JUMPING TO CONCLUSIONS"?
IF SO CAN I GET IT VIA A FILE TRANSFER OVER THE NET?
THANK YOU,
MILT WAXMAN,HUGHES
WAXMAN@USC-ECL
-------
There will be a new version in about a week, and I'll send you a message
then.
∂13-Sep-79 2338 JMC
To: "@SEMLIS[E79,JMC]"
The CASBS AI and Philosophy seminar will meet Tuesday, Wednesday
and Thursday at 1pm at CASBS next week. It will start with John McCarthy
talking about belief, concepts and ascribing mental qualities to machines.
The references are McCarthy's Ascribing Mental Qualities to Machines,
First Order Theories of Individual Concepts and Propositions, Epistemological
Problems of AI (IJCAI 1977), and Formalization of the wise man and
Mr. S and Mr. P puzzles. Copies available.
∂14-Sep-79 0232 Mark Crispin <Admin.MRC at SU-SCORE> Re: Re: VADIC 1200 baud modems
Date: 14 Sep 1979 0223-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Re: Re: VADIC 1200 baud modems
To: Rindfleisch at SUMEX-AIM
cc: JMC at SU-AI, Les at SU-AI, ADMIN.GORIN at SU-SCORE, CMILLER at SUMEX-AIM
In-Reply-To: Your message of 10-Sep-79 0851-PDT
Tom -
Sorry for taking so long to get back to you. We have a problem in that
SCORE doesn't have any VADICs at all. Since SCORE is running an experimental
system that doesn't (yet) support Dialnet, it's unlikely that Dialnet would
get much benefit from installing a modem at SCORE.
If one of your people wants a VADIC installed at SCORE for medical uses,
we can probably find a port for that. I was afraid there would be this
kind of restriction on their use; thanks anyway.
Mark
PS - I just thought of something. We could install a VADIC on your 2020 and
bring up Dialnet there. Ed Feigenbaum was interested in having a 2020
implementation of Dialnet for SDRC on the east coast, so it could be useful
to have a modem on your 2020.
-------
∂14-Sep-79 1203 RPG CASBS Seminars
I hear that they are next week? Could I attend? Where is it?
-rpg-
∂14-Sep-79 1202 ARK S-1 Meeting
To: S1 at SU-AI
CC: pickering at SUMEX-AIM
The next S-1 meeting will be held on Wednesday, 19 September at 1:30pm at SAIL
Small conference room.
Arthur
∂14-Sep-79 1230 LGC via SU-TIP Discussion of Plans
Will you have some time today or tomorrow (preferably today) to discuss
some ideas and plans that I have for work this fall? It won't take more
than an hour, and may take less than that.
∂14-Sep-79 1508 MFB TALK
ALL SET FOR FRIDAY AT 2:00, RM 402, JACKS.
∂14-Sep-79 1649 100 : patte phone
Integrated Circuits called and your superphone is fixed. $25 service charge.
∂14-Sep-79 2216 LGC Discussion Tomorrow?
Would it be convenient for you if we met sometime tomorrow (Saturday) afternoon?
If you'll be at CASBH on your stated schedule even on weekends, we could meet
at 5:30 or so; otherwise, it could be somewhat earlier.
∂14-Sep-79 2256 FB display terminals
To: JMC, LES
I am nearly convinced. I discovered that Wang is selling (!) display editing
stations for $800 each. People say it is one of the best display editors you
can buy (shudder).
∂17-Sep-79 1013 ROY on TTY15 1013
The Imlac seems to be in good shape. Works well. We are presently fixing some
mechanical problems, air filters and cover fasteners etc. This message is via
the Imlac. ROY.
∂17-Sep-79 1243 PAT Lawrence Markosian
of IMSSS called. He would like to set up an appointment to talk with
you for about half an hour. x70411 afternoons, 854-4245 home.
∂17-Sep-79 1312 CLT book
To: JMC, CLT
where are the reviews we got in ny? they were with the old version
that you were reading. I would like to make copies to file.
∂17-Sep-79 1555 MAX
A copy of the axioms and the details of the proof is enclosed and put in your
mailbox at SAIL.
Thanks for the copy of the axioms and proof. Are you interested in the seminar at CASBS?
∂17-Sep-79 1553 MAX
S and P problem.
∂17-Sep-79 2000 JMC*
docdil and christian pysicists
∂18-Sep-79 1546 DPB New student support
To: "@FAC.DIS[1,DPB]" at SU-AI
This file lists the current state of support assignments for new PhD
students. Interests given were expressed on the original admission
application; they may have changed. Contact Denny Brown to express
interest in a student; students will be sent to you for interview.
When a project and a student have come to an agreement, the student
should inform Denny Brown. Appointment processing will then be done.
Legend: FB=Forest Baskett, JLH=John Hennessy,TOB=Tom Binford,HPP=Heuristic
Programming Project,DCL=David Luckham.
NSERC is a Canadian Fellowship, NSF is the National Science Foundation
Fellowship. Both of these require additional RA support at the indicated
levels.
--------------------------------------------------------------------------------
Name Interests Possible Assigned
Support Support
--------------------------------------------------------------------------------
ADAMS, LIA PSL FB/VLSI FB/VLSI
CL JLH
ASENTE, PAUL JOHN AI SRI*
PSL
BAXTER, DOUGLAS JAM NA NA NA
OR
BERLIN, DANIEL L. AI SRI NSERC Fellow
40% RA ?
BRODER, ANDREW ZARY AA HPP/VLSI
PSL
CLARKSON, KENNETH L Pattern Recognition TOB NSF
AI FB/SLAC 25% RA ?
SRI
DIETTERICH, THOMAS AI HPP/VLSI
PSL SRI
GISCHER, JAY L. NSF
25% RA ?
HAIMAN, MARK Unknown TOB NSF
HPP/VLSI 25% RA ?
SRI*
HARKNESS, JUDY PSL FB NSF
Unknown 25% RA ?
MILLER, ALLAN ADAM AI TOB
DCS SRI
NOWICKI, WILLIAM IA Computer Music Synt FB NSF
PSL 25% RA ?
OSSHER, HAROLD LEON PSL DCL
AI FB
SRI
ROSENSCHEIN, JEFFRE PSL FB
AI SRI*
SPENCER, THOMAS H. AA
CM
STOLFI, JORGE MTC
CM
THEIMER, MARVIN MIC NA NA NA
Modeling of physica
TREITEL, RICHARD JA AI DCL
PSL SRI*
∂19-Sep-79 1004 MAX seminar at CASBS
Yes, I am very interested in it. And, I think, Prof. Tang is too. But I don't
no where is CASBS.
CASBS stands for Center for Advanced Study in the Behavioral Science.
Its location may be on your Stanford map. Anyway you can easily find
it as follows: Go out of the campus on Mayfield to the north
and continue north on Junipero Serra. You will see a road on the
left with a sign saying Stanford Golf Course, CASBS and National
Bureau of Economic Research. Turn left there and immediately
left again and up to the top of the hill. If you go on Junipero
Serra farther north, you will immediately see a bridge across
the road. This means you have gone too far. The
distance is probably not too far for Prof. Tang to walk
if he doesn't have a ride, but if he phones me at
321-2052 around 1:30, I'll get him a ride or comae after him.
The meetings, starting tomorrow, are now on Tuesdays and
Thursdays at 2:30.
∂19-Sep-79 1810 the tty of Geoffrey S. Goodfellow [Brian.Reid at CMU-10A: First mail system?]
Date: 19 Sep 1979 1803-PDT
Sender: GEOFF at SRI-KA
Subject: [Brian.Reid at CMU-10A: First mail system?]
From: the tty of Geoffrey S. Goodfellow
To: jmc at SAIL
Message-ID: <[SRI-KA]19-Sep-79 18:03:57.GEOFF>
Reply-to: Geoff @ SRI-KA
Perhaps you should shed some light on this? If so, could you
reply to MSGGROUP@MIT-ML?
Begin forwarded message
===========================
Mail from MIT-ML rcvd at 19-Sep-79 1755-PDT
Date: Wednesday, 19 September 1979 2037-EDT
From: Brian.Reid at CMU-10A
Subject: First mail system?
To: MsgGroup at ML
Message-ID: <19Sep79 203703 BR10@CMU-10A>
Does anybody know authoritatively which was the very first computer
mail system? Which computer? When? Use your own intuition for
when to call something a "mail system".
Brian
===========================
End forwarded message
∂20-Sep-79 1038 CET Lots
To: "@FAC.DIS[1,DPB]" at SU-AI
LOTS has requested a list of classes that will use LOTS.
If you are planning to use LOTS Autumn Quarter, please let me know.
Thanks,
Carolyn
∂20-Sep-79 1201 LJH
You received a phone call wherein the caller left the following message:
Gold Cryptonite
∂20-Sep-79 1203 LES SU-TIP move
To: Carlson at USC-ISI, DCACode535 at USC-ISI
CC: JMC at SU-AI, FB at SU-AI, DPB at SU-AI, REG at SU-AI,
ROY at SU-AI, ROB at SU-AI, HVA at SU-AI, BS at SU-AI
The move of the Stanford Artificial Intelligence Laboratory, including the
SU-TIP, to Jacks Hall at Stanford cannot be accomplished on October 1 as
planned because the new computer room electrical power will not be ready.
The move has been rescheduled for October 29. Since it will take a few
days to dismantle and reassemble the computer facility, the TIP can be
moved almost any time during the week of October 29, at your convenience.
We regret any inconvenience caused by this rescheduling, but were unable
to get the required power installation done sooner.
Sometime later, probably in December, the SCORE computer will be
disonnected from the SUMEX IMP and moved to Jacks Hall, where we would
like to connect it to the SU-TIP also. The exact dates of this move are
not set yet -- we will communicate on this plan later.
Les Earnest
∂21-Sep-79 0000 JMC*
Circumscription to waxman%usc-ecl.
CIRCUM.NEW[S79,JMC]. If you would like a paper copy mail request
to PAT@SU-AI.
∂21-Sep-79 1124 LJH
Don Schick of Internal Audit telephoned regarding the audit report for LOTS.
He wanted to know if you had had a chance to prepare the response yet. His
extension is 7-4195.
∂21-Sep-79 1355 CET
To: "@FAC.DIS[1,DPB]" at SU-AI
∂21-Sep-79 1052 Mcgoveran at SUMEX-AIM Preliminary Agenda
Date: 21 Sep 1979 1049-PDT
From: Mcgoveran at SUMEX-AIM
Subject: Preliminary Agenda
To: CET at SAIL
cc: McGoveran
Carolyn, would you please send this to the list you have? I shall send
hardcopy
Faculty Meeting September 25, 1979
Preliminary Agenda
1. Degree candidates
2. Discussion of G81 policy
At our last "Black Friday" meeting we voted to put a hold on the
winter registration of some students who had apparently refused
to file their G81 forms indicating their topic and reading com-
mittee. The stink this caused among the students made me
reflect on this action and regret it as inappropriate. Thus I
would like to propose that we remove the hold on their winter
registration. (Baskett)
3. Hold on winter registration
I would also like us to consider a revision of our G81 filing
requirement policy in such a way that clearly indicates that a
student's thesis advisor has the authority (and obligation) to
indicate that a reading committee and topic are being adequately
pursued by a student. (Baskett)
4. Advising/144 A,B,C
-------
You may count me as voting for your proposed revised G81 policy
and removing the holds.
∂21-Sep-79 1904 LES
∂21-Sep-79 1324 ENGELMORE at USC-ISI Bob Filman's thesis
Date: 21 Sep 1979 1322-PDT
Sender: ENGELMORE at USC-ISI
Subject: Bob Filman's thesis
From: ENGELMORE at USC-ISI
To: LES at SU-AI
Cc: Carlson
Message-ID: <[USC-ISI]21-Sep-79 13:22:12.ENGELMORE>
Hi Les,
I was just perusing Bob Filman's thesis on oservation and
inference, and thought that the chess problem he picked as a case
study was a fascinating one. However, you should probably be
aware that Bill Carlson has had to fend off some reporters who
are trying to make a case that the DoD is spending (wasting)
taxpayers' money on computer chess programs. Just to be on the
safe side, it would be appreciated here if published papers that
look like they are about chess omit the ARPA acknowledgement.
Thanks,
Bob
∂21-Sep-79 1901 LES Theses relating to chess
To: Engelmore at USC-ISI
Bob,
Well, we just published another one that acknowledges ARPA support:
Dave Wilkins' thesis on "Using Patterns and Plans to Solve Problems
and Control Search" (AIM-329, August 1979). It came back from the
printers yesterday and was about to be distributed. It is a good
piece of work and well worth the support, but if it is likely to be
embarassing in the current climate, we could censor or redo the
cover and title page. This would be a messy project, of course.
Les
∂21-Sep-79 1946 LES
∂21-Sep-79 1938 ENGELMORE at USC-ISI Re: Theses relating to chess
Date: 21 Sep 1979 1936-PDT
Sender: ENGELMORE at USC-ISI
Subject: Re: Theses relating to chess
From: ENGELMORE at USC-ISI
To: LES at SU-AI
Message-ID: <[USC-ISI]21-Sep-79 19:36:02.ENGELMORE>
In-Reply-To: Your message of 21 Sep 1979 1901-PDT
Les,
As for Wilkins' thesis, I would leave it as it is. We're more
concerned about papers that would have a lot of exposure, such as
one that appears in a journal, where someone might get beyond
reading just the title. In the future, however, it would be
simpler to just apply the rule I suggested in the last message to
all publications. Thanks,
Bob
∂22-Sep-79 1334 RMS at MIT-MC (Richard M. Stallman)
Date: 22 SEP 1979 1633-EDT
From: RMS at MIT-MC (Richard M. Stallman)
To: jmc at SU-AI
Good bye. It was fun talking with you.
I still have Sato's thesis, which I will leave with Diffie
before I return to Boston.
∂23-Sep-79 1740 TOB
To: FB, JMC
John Scull, Sept 23, 1979
John Scull, JPL, is being invited by Aero and Astro department
as a Consulting Prof, beginning this Sept. JPL will support
him and A & A will provide space. Prof McCarthy approved a joint
invitation offering use of robotics laboratory facilities and
use of the computer system. It now is clear that what was requested
was a joint appointment. I have sent copies of
the letter of invitation from Aero and Astro
John Scull's resume
to Prof Baskett and Prof McCarthy for your decision. I will be
away Mon Sept 24 thru Wed Sept 26.
∂24-Sep-79 0838 CET Reminder
To: "@FAC.DIS[1,DPB]" at SU-AI
CS Faculty Meeting
Tuesday, September 25, 2:15 p.m.
Margaret Jacks Hall - room 402
∂24-Sep-79 1028 HEARN at UTAH-20 Herbert Stoyan
Date: 24 Sep 1979 1126-MDT
From: HEARN at UTAH-20
Subject: Herbert Stoyan
To: mccarthy at SAIL
I saw this gentleman recently in Dresden. He sends his regards. I also
have a copy of his LISP book (in German). Would you like a copy too?
Regards,
Tony
-------
Thanks for message. I have a copy of Stoyan's book.
∂24-Sep-79 1058 LES Vadic modems
To: MRC at SU-AI
CC: JMC at SU-AI
The new Vadic modems have arrived.
∂24-Sep-79 1225 PAT
∂22-Jun-79 1628 MDD SAIL MEMO
To: PAT
CC: JMC
John has suggested that two short memos I've written here be combined to form
(a) a SAIL MEMO and (b) a paper to be submitted to a special forthcoming issue
of Artificial Intelligence. He said you could take care of merging them.
They are stored in my file [1/MDD] as REP and CRIT. REP "requires"
PUBMAC.PUB from my file which in turn calls on some fonts still on tape.
I'll use "PUMPKIN" (see LES if you don't know about PUMPKIN) to try to get
them restored (probably will happen next week). Thanks and have a good summer.
----
I think I gave you these to look at and approve before having them printed.
Do you remember the status of this?
I have given them to Pat Hayes, Associate Editor AI Journal - belatedly.
∂24-Sep-79 1219 DCO
To: "@GROUP[1,DCO]" at SU-AI
PROGRAM VERIFICATION AND ANALYSIS SEMINAR
PLACE: ERL 237
TIME: 2:15 p.m., Tuesday, October 2
Our program verification and analysis seminar will begin again
next week --- same room, but note that we will begin 15 minutes earlier
this year (to allow another course to use the room later). If you do
not want to receive weekly announcements, please send me a message.
Also, please send me a message if you know of someone who does want to
receive them.
Our first speaker will be Gregor Bochmann. I will send an
abstract later this week.
∂24-Sep-79 1357 MFB THANKS, AND SOME BUSINESS.
THANKS FOR COMING TO MY TALK. I DONT THINK I GAVE IT THAT WELL, BUT THE BASIC
MESSAGE CAME ACROSS, I HOPE.
I WAS VERY GRATIFIED TO GET A POSITIVE RESPONSE, THANKS AGAIN.
I WOULD LIKE TO SCEDULE MY ORALS FOR THE END OF THE QUARTER.
I PRESENTLY HAVE 30 OR SO PAGES OF THE BASIC TECHNICAL CHAPTER WITTEN. PLUS
A BIT OF OTHER STUFF. I THINK I CAN GET THE WHOLE THING REAALLY DONE
THIS QUARTER. IS THERE ANY BAD TIMES FOR YOU AT THE END OF NOVEMBER
OR BEGINNING OF DECEMBER?
ALSO, I AM SAVING OUTPUT OF MY PROGRAM FOR USE IN THE THESIS. I AM WONDERING
IF I AM SAVING THE RIGHT INFORMATION. WHAT KIND OF STUFF IS IMPORTANT - JUST THE OUTPUT,
DETAILS ON TIME AND SPACE USE?, OTHER THINGS?
∂24-Sep-79 1609 CET Time of CS Faculty Meeting
To: "@FAC.DIS[1,DPB]" at SU-AI
The meeting is being called at 2:15 p.m. EE faculty meeting starts at
4 p.m.
Tuesday, Sept. 25, Jax 402, 2:15 p.m.
∂24-Sep-79 1646 CSD.MILLER at SU-SCORE new PhD student
Date: 24 Sep 1979 1632-PDT
From: CSD.MILLER at SU-SCORE
Subject: new PhD student
To: jmc at SU-AI
I am a first-year PhD graduate student, and I would like to talk to
you sometime about possible research (and how things work around here). Can
you please get in touch with me by mailing to csd.miller on the Score machine
or calling 493-9377 if it is not too late at night (i.e. before 11 PM).
Thank you,
Allan Miller
-------
∂24-Sep-79 1736 LES Bit map displays
To: FB at SU-AI, JMC at SU-AI, REG at SU-AI
I just remembered that I forgot to pass along some data that I collected
on bit-map displays, from Jim McKnight of Grinnell.
The video synthesizers cost $5,500 for an interface to the mini of your
choice (e.g. Unibus) plus $500 per 512x512 bit map. Thus, a 16 channel
system with a PDP-11/34 would cost $1500 per channel, not counting the
monitor or keyboard (add about $700 each for them).
With 32 channels on an 11/34 (assuming it can cope), the cost per channel
comes down to $1,000 per channel. By using a video switch, the effective
price can be reduced to about 2/3 of this value.
Chuck Reiger's company is rumored to be offering 512x512 bit maps
interfaced to the S-100 buss for $800 each. It appears to be attractive
for stand-alones if not for general use. I'll try to get more
information.
∂25-Sep-79 0230 Mark Crispin <Admin.MRC at SU-SCORE> Re: Vadic modems
Date: 25 Sep 1979 0226-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Re: Vadic modems
To: LES at SU-AI
cc: JMC at SU-AI
In-Reply-To: Your message of 24-Sep-79 1058-PDT
How many VADICs arrived? I was aware of only one. If there is one,
then I guess the thing to do is to take it over to HP and install it
there and possibly move the GSB one to LOTS. If there are two, then
we can install one at HP, one at LOTS, and leave the one at GSB for
now. I'm afraid that SCORE with its release 4 monitor is out of the
picture for a while, at least until Christmas. SCORE doesn't really
need Dialnet urgently anyway as long as the AI Lab has it.
About installing a VADIC at LOTS...you should know that Ralph has
expressed opposition to installing Dialnet at LOTS. LOTS doesn't
have any free terminal ports right now, and as I understand it Ralph
does not feel that Dialnet is worth the cost of liberating a terminal
port. You'll have to work this impasse out with Ralph - I'm stuck in
the middle and refuse to take sides.
Even given that it's worked out with Ralph, those modems sure picked
a lousy time to arrive. It's gonna be a pain to get downtime at LOTS
until Thanksgiving, maybe Christmas. At least HP downtime is still
available.
Mark
-------
∂25-Sep-79 1210 PAT
∂24-Sep-79 1706 JMC
I have given them to Pat Hayes, Associate Editor AI Journal - belatedly.
-------
Do you still want them to be an AIMemo??
Yes, but let's wait till Pat Hayes tinkers with them. I'll tell you.
∂25-Sep-79 1210 DCO
To: "@GROUP[1,DCO]" at SU-AI
PROGRAM VERIFICATION AND ANALYSIS SEMINAR
and
CONCURRENCY SEMINAR
PLACE: ERL 237
TIME: 2:15 p.m., Tuesday, October 2
TITLE: Specification, Verification and Implementation of
Data Communication Protocols
SPEAKER: Gregor Bochmann, Universite de Montreal
Gregor will be speaking for several weeks...
ABSTRACT:
Data communications and distributed data processing involve a
variety of communication protocols. For the design and implementation of
these protocols, specification and verification methods are needed. We
explain the issues involved in the specification of protocols and their
logical verification. The main approaches to formal protocol descriptions
are reviewed, and examples given to demonstrate the different
specification methods. We explain in particular an approach using a
general transition model which combines aspects of finite state transition
diagrams and programming languages. Different ways of structuring a
protocol into separate modules or functions are also discussed.
We also describe the issues involved in the specification of the
communication services provided by a protocol. Two aspects of a service
specification are distinguished: (a) the local properties which
characterize the interface through which the service may be accessed, and
(b) the global properties which describe the "end-to-end" communication
characteristics of the service. Examples based on the HDLC and X.25 data
communication protocols are given.
We finally describe an experience with an implementation of the
X.25 communication protocols for accessing public data networks. The
implementation effort is characterized by (a) the development of a
formalized protocol specification on which all further implementation work
is based, and (b) by the use of Concurrent Pascal as the implementation
language. The main features of the formalized protocol specification are
given, and a method for deriving a protocol implementation based on
parallel processes, monitors and classes is explained. The overall
structure of the system, and the step-wise refinements leading to the
commplete implementation, are discussed. Some comments on the possible
implementation on multiple microprocessors are also given.
∂25-Sep-79 1548 LES
To: REF, DEW
CC: JMC
ARPA acknowledgement
We received a note from Bob Engelmore at ARPA regarding research that
involves chess problems. They would prefer that any journal articles
based on such work not acknowledge ARPA support.
Les
∂26-Sep-79 0229 REM via AMES-TIP Multi-CPU computers
To: HPM at SU-AI, S1 at SU-AI
(1) Is there any distribution list for people randomly interested in
methods of hooking up many CPUs (perhaps a thousand or million; or at
least sixteen for a starter)? Some people I should distribute my ideas
to and get into discussions with?
(2) I think the right way to handle bootstrapping and system-crash
recovery as well as random patching is via a slave facility, namely that
any CPU can enslave any topologically-adjacent CPU at any time merely by
issuing the appropriate protocol (password) through the normal message link.
There are two ways to do this:
(a) A CPU is always running a pROM program which emulates some data-compressed
intermediate code such as P-CODE. This emulator includes interrupt handler
for getting messages from UART or other link from adjacent node, and in its
canned message-handler is the secret slave-password and a command handler
for slave-commands. Since this emulator is fully-debugged and never changed,
and can't be clobbered by code it is emulating, the CPU never gets hung
(due to invalid opcode or NXM) and thus can't ever be in a state where
slave-commands are ignored.
(b) Attached to each CPU is an IO-handler which intercepts all incoming
messages and upon receipt of a slave-password will gronk the main CPU
to interrupt the program inprogress and halt operations while the IO-h.
performs slave-commands (like examine and deposit memory).
In either case enslaving adjacent CPUs won't go astray because only the
human-interface CPU can accept slave-commands directly from non-computer
control (namely from the master human-to-computer keyboard) hence chains
of enslavings which propagate thru the multi-CPU system can start only
at the human-interface CPU. Other CPUs will only enslave adjacent CPUs
if directed by program, which means as part of some enslave-chain-letter
started by the operator. Some enslave-chain-letters may exist indefinitely,
i.e. the operator may load a program into each CPU which enslaves each
adjacent CPU from time to time to check its status (functioning or broken),
and of course the operator/programmer has to be careful when invoking such
a program (it might have bugs), but of course it is easy for the operator/
programmer to establish heirarchial slaving so that a buggy phantom can be
flushed by a manual enslave command. Anyway I think this is the way to
build a multi-computer system so that once it is built you can do all the
bootstrapping you want whenever you want without having to electrically
gronk any CPU other than one human-interface CPU.
Anybody want to discuss this further with me?
Anybody know of a multi-cell computer that is available now for me
toplay with? (Arpanet et al with their many computers don't count!)
∂26-Sep-79 0248 REM via AMES-TIP
To: S1 at SU-AI
Oh my goodness, I didn't know "S1" was a giant mailing list, sorry for the
long message.
∂26-Sep-79 1242 100 : patte airfares
Round trip SF → Dublin $968.00
Round trip SF → Detroit $251.00
∂26-Sep-79 1248 LJH Caspian
Boris Fine phoned and would like you to call him at 524-9987. There is no
problem with Caspian he just wants to give you some info.
∂26-Sep-79 1803 LES Display system discussion
To: FB at SU-AI, REG at SU-AI
CC: JMC at SU-AI
JMC proposes that we get together as soon as possible to discuss display
system alternatives. He prefers late morning (11am) anytime next week
(Oct. 1+). Second choice is MWF afternoon or Tu-Thur after 5pm.
∂26-Sep-79 2020 FB via SU-SCORE Display system discussion
To: LES at SU-AI, JMC at SU-AI, REG at SU-AI
I prefer to wait a week or two. John has never seen Alto's and we will have
our twice resolution Alto display running and Alto's will be available also.
I think these systems are going to change our expectations and I want people
to at least see them before we get too serious.
∂26-Sep-79 2257 LES Display system
To: FB at SU-AI
CC: JMC at SU-AI, REG at SU-AI
We can wait if you wish, but I doubt that anyone's expectations are going
to be raised in the next two weeks.
I'm surprised that John hasn't seen an Alto. Perhaps he never bothered
because he was aware that it was (and is and likley always will be) a
senseless machine configuration.
Of course, the quality of the Alto display is very good. It would
be nice if you could buy terminals like that for reasonable prices,
but you can't and won't for quite a while. So the best thing to do
is get a limited number of high quality displays for those tasks that
require them and get some reasonably priced terminals for the bulk of
the users.
∂27-Sep-79 0836 FB displays
To: LES at SU-AI, JMC at SU-AI, REG at SU-AI
The important things about the Alto display is its visual quality, it's size
and orientation (606 x 808, 15 inch tube oriented in portrait mode), the fact
that it is a minor modification of the standard Ball TTL-150 monitor with a
P-40 phospher, and that Xerox buys them from Ball for $200. The size (606 x
808) is not 512 x 512 (thank goodness) and it is nearly twice as many bits
but they all still fit in the 16 bit address space of most micros. Because
of these facts, its worth looking at.
(And I can't resist adding the irrelavent remark that the automobile is a
senseless machine configuration from certain points of view.)
I have seen Altos and would prefer high quality displays. I have
unrestricted money that will buy a high quality Baskett display for
JMC as soon as they are available. However, I think the Department
can and should cover the Department with displays of at least
Datadisc quality. If the Department can afford better quality, that's
fine, but I doubt that we should settle for less than full coverage
in order to get high quality. My guess is that if a person had a choice
between a Datadisc on his desk and an Alto in a terminal room, he would
choose the former. Should a poll of faculty and students show that
a reasonable number of people would prefer high quality in terminal rooms,
I would certainly go along. What I fear is that the Department will
divide into 1st class citizens with high quality personal displays,
1.3 class citizens with Datadiscs 2.8 class citizens with access to
terminal rooms.
I am willing to wait a "week or two", pointing out that we have been
waiting a year or two. I will be leaving for two weeks on October 14.
Could we set a definite date to happen before then?
∂28-Sep-79 1241 LCA
John, I'd like to remind you about my letter of reccomendations. Even a
very short one would be very useful to me. I need it by monday or tuesday.
If you need any information, please call me either at the lab or at home (856-0706).
I have put a short resume' of mine in your mail box at the lab. Thanks a lot!
Gigina
Richard:
Could you write up something I could reasonably sign? I haven't been
following Gigina's work carefully, and I am jealous of my time on leave.
∂28-Sep-79 1347 TOB Chinese visit
To: JMC, LES
I have received a request for an invitation to
visit from the chief of development, China
Precision Machinery Corp. A team of three specialists
in his corporation and two scientists from Academy of
Sciences will visit the US in December to January
(dates unspecific). I propose to respond with an
invitation to visit for one day and a request for
specifics of their plan, offering to consider a longer
visit given details of their proposal.
Tom
∂28-Sep-79 1607 John.Laird at CMU-10A Re: sex in wetsuit
Date: Friday, 28 September 1979 1904-EDT
From: John.Laird at CMU-10A
Subject: Re: sex in wetsuit
To: John McCarthy <JMC at SU-AI>
Message-ID: <28Sep79 190405 JL30@CMU-10A>
In-Reply-To: John McCarthy's message of 22 Sep 79 03:20-EST
It should be impossible. The current version here should already have that
restriction. If not I'll definitely fix it. I'm afraid there are versions
of HAUNT around that I don't have control of.
∂28-Sep-79 1643 CET Colloq. Tuesday
John, we acted on your recommendation and invited Dave Wilkins to present
the Colloq. on Tuesday, Oct. 2. We would be pleased if you would intro-
duce him (even though you are on leave) since he was your advisee.
Carolyn
No, I won't jeopardize my leave for ceremonial reasons.
∂28-Sep-79 1649 SEK Logic seminar.
To: "@FOL.DIS[FOL,RWW]" at SU-AI
Jon Barwise is starting up a logic seminar. First meeting Wednesday,
Oct. 3, 7pm in the Math Dept. lounge (second floor). Jon is a new
professor in Philosophy with a background in math, and is quite
interested in talking with AI people. In particular there is interest
in hearing about what has been done in FOL. (I'll be giving a presentation
on what I've been doing.) There will also be an informal dinner before the
talk, time/place as yet not set. --Scott.
∂28-Sep-79 2138 CLT
Put the following concerts on your calendar
∂29-Sep-79 0904 JRA lisp conf
John:
progress is being made; below are 1) an intial outline of the conference
and 2) proposed schedule to ready our goal. sugggestions are WELCOME.
in particular an estimate of addendance wouuld be most helpful.
letters went out last monday air mail special delivery to
Eiichi Goto, Patrick Greussay, Erik Sandewaall,and Joachim Labusch.
Danny Bobrow is booked up. I'm working on Warren; he's busy, but I think
convinceable. Carl Hewitt said yes, and suggested Rod Burstall. Gianfranco
Prini is also interested. I'm considering Dan Friedman; what about Carolyn
Talcott? I want to satisfy several simultaneous constraints: representative
geographical areas; representative topical areas; and maintain a quorum
which is reasonably accessible in a short notice.
the conference office says they have some scheduling problems for the
aug24-27 period; that's why U Santa Clara came up (Ruth Davis is on the faculty
there and will be handling local arrangements). My first preference is stanford
of course, but overflow from the AI meeting is attractive to our conference.
on sponsorship: is it possible to get some seed money through the lab? if we can
string out the debts until advanced reg. fees start to flow, then we can maintain
control (=profits) and be in a position to do this on a regular basis. so far
i've been paying postage and such myself, but sooner or later someone is
going to ask for lots of money; if outside sponsorship is necessary, then
it has to be requested immediately.
below are the blurbs:
-------------------------------------------------------
Sept 23,1979
.select 2
.once center
The LISP Conference
.select 1
%2PURPOSE.%1 Many areas of contemporary computer science have their spiritual
roots in developments related to LISP. The impact of LISP on AI
is well known. The thrust of this conference is the emphasis of
the non-AI contributions of LISP. These areas span architecture,
a unique view of programming, and a perception of the importance of
theory. The call for papers will reflect this breadth.
%2PROGRAM COMMITTEE.%1 In keeping with the international character
of LISP-related activities, I have invited
participants from several countries. Considering the geographical
dispersion and the time constraints, it is of critical importance that
our program committee be formed at once, and begin work.
%2THE CALL FOR PAPERS.%1 The following selection of topics is meant only as a guide.
The major concern is to attract participants outside of the traditional AI
areas.
.begin indent 5,10
%2Architecture.%1 Several groups are actively pursuing
the design and implementation of LISP hardware. These projects
range from re-microcoded
commercially available hardware to specially designed LISP chips.
%2Languages and Theory.%1 There are several research areas related to
LISP-like languages,
though the ancestry may be spiritual rather than physical.
For example,
papers on applicative languages, object-oriented languages and
formal semantics of LISP-like languages should be
requested.
%2Programming and Environments.%1 One important facet of LISP as
a programming language
is its flexible programming behavior including the support systems
which surround the language. These aspects of LISP should be represented
at our conference.
%2Applications.%1 LISP is finding great appeal outside the traditional
Artificial Intelligence area. We should offer our conference as a forum for
these new applications.
%2Note.%1 These areas are not exhaustive. All suggestions for revision
will be considered.
.end
%2PANEL DISCUSSION.%1 There should be a panel discussion involving the
topic "What is LISP?". Even informal conversations will
elicit several divergent if not contradictory views of LISP;
a organized effort should prove even more illuminating.
%2DATE.%1 Currently, the dates of August 24-27 1980 (with a party
on the evening of August 24), is most attractive. The preceding
week Stanford will host an AI meeting. Our conference should attract
many of those participants.
%2LOCATION.%1 The possibilities I am pursuing are the campus of
Stanford University and that of the University of Santa Clara (about
ten miles from Stanford). Both facilities have conference facilities,
lodging, and food services. Of course, a commercial facility could also be
used, but a campus atmosphere is advantageous.
%2SPONSORSHIP.%1 As yet unsponsored. I have heard conflicting reports about
the benefits of sponsorship. I will continue to investigate; perhaps
sponsorship is a question to be decided by the Program Committee.
%2QUESTIONS.%1 Several questions have occurred in this initial phase.
I would like your opinions, please.
.begin indent 5,10
Should the conference be sponsored, or should we try to maintain
our own financial control. It is at least conceivable that we
run the budget with a small capital outlay with the major expenses
to be covered by advance registration. On the other hand, if it
is to be sponsored, what agencies would you suggest?
A major concern is how to estimate attendance. Could you please
estimate how many people you would expect to attend from your
area or country? Could you also estimate the total conference
attendance.
.end
%2FINAL NOTE.%1 Time is a %2major%1 concern. Please respond as soon
as posssble. If you do %2not%1 feel that you can contribute the needed
effort I would appreciate it if you could direct me to alternate
candidates.
.group skip 6
Thank you for your time and consideration.
.group skip 4
John Allen
--------------------------------------------
Tentative deadlines:
October 30: Call for papers announced through Net and underground (in
various countries), with an official notification ASAP. The delay here
is the necessary naming of committee members.
(thirteen weeks)
January 28: Papers due;
I would favor
a detailed abstract (size specified in number of words, not pages).
blind refereeing, even though the authorship coan often be deduced.
each committee member reads all papers, possibly farming them out
to conscientious colleagues; of course the membership is ultimately
responsible for the papers. The logistics of acceptance/rejection
are to be worked out by the committee (before January 28);
therefore each author submits no. copies = size of committee.
(ten weeks)
April 7: Accept/Reject;
Ten weeks for the review process is within reason. It doesn't allow
a lot of room for error, but I intend to be donating almost full time
to this conference effort, persistently nagging if necessary.
(ten weeks)
June 16: Final papers due;
Ten weeks for revision is generous. This period
can be reduced if some of the prior periods look too short.
(ten weeks)
August 24: Conference begins;
This leaves ten weeks for printing and binding. This too seems
generous. Perhaps we can do with less.
1. Why not have the conference in 1981? This would give plenty of
time to prepare. 1980 is ok, but delaying might get better papers.
2. I suppose the AI Lab could pay postage, etc. What larger sums of
money do you have in mind.
3. Assuming equal physical facilities, having the conference at Stanford
presents no large advantages over Santa Clara, unless the fact that
conferences are an industry at Stanford helps. It might also generate
excessive greed on Stanford's part. Assuming an adequate conference
committee, there won't be a prestige difference.
4. I bet he won't take an active part, but you should solicit Bruce
Buchanan's advice. He was IJCAI program chairman.
5. Certainly Friedman should be invited. He is very energetic. The
conference could even be held at Indiana, although there is less incentive
to go there.
6. Proving the correctness of LISP programs should be a major topic.
You could invite either Carolyn Talcott, Bob Boyer, J. Moore or Corky
Cartwright to be on the program committee. This topic should include
both automatic proof as done by Boyer-Moore and the mathematics of LISP
as done by Cartwright and me. Inclusion of this topic will put LISP
one up on APL and other languages where this is not yet a consideration.
7. It occurs to me that IJCAI could afford to lend the conference money,
but I don't know who could make such a decision. Bruce would know. For
that it would be best to leave out the words about non-AI contributions.
8. If you take non-AI contributions seriously, then the biggest are to
REDUCE and MACSYMA. Hearn and Moses (Fateman at UCB locally) are obvious.
9. You should also talk to Les about practical matters.
∂30-Sep-79 1000 JMC*
Bring in mcc and hayes and programs with common sense
∂30-Sep-79 1508 EF at MIT-AI (Edward Fredkin)
Date: 30 SEP 1979 1806-EDT
From: EF at MIT-AI (Edward Fredkin)
To: JMC at MIT-AI
I won't try, The posters went up at Tech square
while I was on the phone talking to you, literally!!!!!...oh well,
Think of something or someone interesting that you might enjoy while in
cambridge and let me try to arrange it. Also , Joyce and I would love
to have you stay here with us, we live in an apartment with a great
view which is located on Beacon St. near Coolidge Corner. Regards, Ed.
Unfortunately, I must go on to Detroit that very evening. But thanks.
∂01-Oct-79 1200 PAT Mr. Tang and Jacks Hall
To: LES, JMC
There is a problem regarding Mr. Tang's accommodations at Jacks Hall.
Apparently, the Tang on the CS map is another Tang from China. The
department's solution to this problem was to go ahead and put our Mr. Tang
in that office also. There
are already 4 people in that office and Mr. Tang would make 5. As he
was very unhappy with the arrangements, he complained to Feigenbaum's
secretary. As far as I know nothing has been done to try and solve
the situation.
In the meantime, Mr. Ma has lent his office at Jacks Hall to
Mr. Tang until the lab moves down there.
I told Mr. Tang that we would try and solve the situation.
∂01-Oct-79 1521 PAT
To: JMC
CC: LES, HVA
itinerary
13 Oct LV San Francisco 1100A TWA 806Y
14 Oct AR Tel Aviv 400P
26 Oct LV Tel Aviv 850A BA 577Y
AR London-LHR 1255P
LV Tel Aviv 800A El Al 15Y Wait listed
AR London-LHR 1200N
LV London-LHR 150P Aer Lingus 161Y
AR Dublin 255P
30 Oct LV Dublin 730A Aer Lingus 150
AR London-LHR 835A
LV London-LHR 1115A BA 171 Concorde (this is the only flight that will
AR JFK 1014A make it on time, you should also
carry on your luggage and let
the stewardess know that you
have a tight connection and
would like to get off first to
get through customs and immigration
as fast as possible)
(you need to take a cab from JFK to La Guardia)
LV La Guardia open Eastern Shuttle
AR Boston
LV Boston 912P AA 513
AR Detroit 1059P
01 Nov LV Detroit 810A United 89Y
AR SFO 1100A
<
∂01-Oct-79 2134 BIS Pony
Will the XEROX machine be tied into the Pony in the new building?
∂02-Oct-79 1007 ARK S1 Meeting
To: S1 at SU-AI
There will be an S-1 meeting tomorrow, Wednesday, 3 October at 2:15 pm in
Margaret Jacks Hall room 301. Schedule conflicts to ARK.
Arthur
∂02-Oct-79 1940 DCO
To: "@GROUP[1,DCO]" at SU-AI
PROGRAM VERIFICATION AND ANALYSIS SEMINAR
and
CONCURRENCY SEMINAR
PLACE: ERL 237
TIME: 2:15 p.m., Tuesday, October 9
TITLE: Specification, Verification and Implementation of
Data Communication Protocols
SPEAKER: Gregor Bochmann, Universite de Montreal
Gregor will be continue speaking...
A limited number of copies of two papers by Gregor, "A General Transition
Model for Protocols and Communication Services" and "Specification and
Verification of Computer Communication Protocols", are available. If
you want a copy, send me a note.