∂26-MAR-75  1903		A,JJ
 I am working for Winograd.  I'll be at the lab most of tonight if you
 want to talk about it.  ...Jerry

∂26-MAR-75  1436		1,JB AT TTY110   1436
 I am enjoying working on Ramsey's theorem; I am keeping in touch with RWW
 about it. I wish you good luck in Japan!

∂26-MAR-75  1123		CYB,DBL
 please try to spare a few minutes and look over the proposed syllabus for this
 year's AI qual. You have a copy; also found on file SYL.PUB[AI,TW]. Thanks. -- Doug

∂26-MAR-75  1027		DOC,TOB
 KRK wants to work here this summer.
 He has made an intensity map for the video synthesizer as free
 student labor.  He impresses me as knowing what he wants and
 as pretty sharp, aggressive, and competent.
 WHAT DOES QUAM THINK OF HIM?
 WHAT DOES HE PLAN TO DO THIS SUMMER?
 He wants to work with digital hardware and hopefully servos.  He
 is interested in working on things which are of high priority to
 us also.
 WHAT ARE HIS LONG RANGE PLANS
 He will go off for a year or longer to get some experience with
 digital design.  He may return for a PhD, or may decide that he
 will gain more by working in industry without a PhD.
 WHAT DOES HE GAIN OUT OF BEING HERE THIS SUMMER?
 He is seeking a job which does not open up until September or so.
 He wants to be in the bay area somewhat longer.  He enjoys working
 at the lab.
 WHAT DOES HE WANT IN PAY?
 He wants to work fulltime.  He needs at least $350 a month.  If
 it were impossible to be paid full time, he would consider halftime
 pay.

∂26-MAR-75  0843		1,HVA
 CONgratulations and Bon Voyage. May I ask if decision has been made
 on Spring Quarter salary? HVA

∂26-MAR-75  0119		2,NXL
 yes, but at home.  The grade sheet you have (had?) should also tell.

∂26-MAR-75  0109		2,NXL
 ok.  Have you turned in the grades for the incompletes in 206 - someone
 was asking me again a couple of weeks ago - he said he'd talk to you.
 I'll see you in 3 months if not tomorrow.    Nick

∂26-MAR-75  0003		PNT,DCL (reminder)
 We have a date for a quick chat late this afternoon-David.


∂24-MAR-75  2336		network site SRI
 Date: 24 MAR 1975 2334-PDT
 From: BOYER at SRI-AI
 Subject: Nuclear Energy
 To:   JMC at SU-AI
 
   I have decided that because I am so ill-informed about the 
 case for or against nuclear energy I would not make a 
 useful contribution to the deliberations of the "steering" committee
 for SENSE; so I won't come to Connolly's Tues. evening.
 
   I am sorry that it never came to pass that I gave a lecture to
 your class on the theory of computation; perhaps I should have
 reminded you.  In any case, I would still like to give such a
 presentation.  Incidentally, J  Moore just tonight discovered
 that our theorem prover was in fact capable of proving
 something, at least, about the Fibonacci function, namely
 that Fib(n) is equal to cdr(f(n)) where Fib has the usual defintion
 (fib (n) (cond ((zerop n) 0)((equal n 1) 1)(t (plus (fib (sub1 n))(fib (sub1
 (sub1 n)))))))
 and where f is the following "optimization" of fib, namely
 (f (n) (cond ((zerop n) (cons 0 0))((equal n 1) (cons 0 1))(t
 (cons (cdr (f (sub1 n)))(plus (car (f (sub1 n)))(cdr (f (sub1 n].
 
 -------

∂24-MAR-75  1224		1,PAW
 your flight to japan has been changed as the flight connecting to the PANAM
 flight from Tokyo to Osaka sold out...you are confirmed on Northwest 9
 leaving SFO at 9:30 arriving Tokyo at 4:55 on the 28th connecting in Tokyo
 to Northwest 7 leaving Tokyo at 6:15pm and arriving Osaka at 7:20 pm.  Will
 send telegram to Takasu with change....patte

∂24-MAR-75  1032		RAP,MG
 Subject: MEETING TODAY OF DISCUSSION GROUP. 
 
 At today's meeting Jim  Morris will talk about joint work  he and Ben
 Wegbreit have been doing. 
 
 The  time and place are  3:30 today in  the ISHIZAKA room  (S 117) of
 SRI's INTERNATIONAL BUILDING. 
 
 If you  can't find  room  to park  in the  visitors lot  outside  the
 'I-building`  go along  Ravenswood  untill Middlefield.  Then take  a
 right  along Middlefield and  then take the first  entrance into SRI.
 Apparently there's a big car-park there. 
 
 Does anyone want to speak at the next meeting ? 

∂24-MAR-75  0947		P,JRA
 would you like more details?


∂24-MAR-75  0939		1,OH
 ARPA RECEIVED THE PROPOSAL AND THERE IS BIG TROUBLE. THEY WERE SUPPOSED TO
 PAY A $25 DELIVERY CHARGE.  THEY WOULD NOT PAY IT AND LET US REIMBURSE THEM.
 LICKLIDER SAYS HE DOES NOT WANT TO GET INVOLVED WITH MONEY.  THEY WILL
 GET BACK TO US TO LET US KNOW IF THE FREIGHT COMPANY WILL BILL STANFORD.
 OLIVIA.
 

∂22-MAR-75  1239		2,BG
 The most recent version of the MTC section of the ARPA proposal is
 FR2B[2,BG].

∂22-MAR-75  0422		1,TAG
 YES, I PUT FINISHING THE ARM ABOVE REPAIR OF THE IMLAC, WHICH IS LIKELY
 TO BREAK IN SOME NEW AND INTERESTING WAY SOON AFTER IT IS REPAIRED ANYWAY. I
 EXPECT TO GET IT WORKING EARLY NEXT WEEK. BY THE WAY, I FINALLY OBTAINED ALL THE
 PARTS FOR MOUNTING THE LOUNGE TV ANTENNA ON ITS OWN MAST WITH A ROTOR, AND THE BILL
 IS $63.92. I HAVE THE RECEIPTS IN MY OFFICE.

∂21-MAR-75  1450		DOC,AJT
 GOAL.NOT[R,AJT] is a note about what I propose to do.  RWW will
 incorporate it in some form in the proposal, if you approve.
 a.
 I might think about using Geomed as a display program, just for
 dramatic effect, in the assembly task mentioned.

∂19-MAR-75  2024		CAR,HPM
 A new version of CRYPT which works if TTY: is specified as the input
 file is available as CRYPT.FAI[HAK,HPM] if you want it. The version
 on the system has been changed to it.

∂17-MAR-75  2253		network site AI
 
 From:  MINSKY@MIT-AI 03/18/75  01:52:49
 Diffie asked me to write reccomendation to stanford.
 I think he is first rate and should do very well.
 Is there still time for letter?

∂07-MAR-75  1109		THE,AJT
 I shall be glad to hear your ideas; but why are you prepared to give only
 half time support during the summer when it is/has been standard practice to
 give full time support? Is this because you are unhappy with what I'm doing?
 If so, say so, and I will argue with you, and make invidious comparisons...
 a.
 p.s. one of the reasons that has decided me against staying next year
 is that I really don't feel that I have been getting much (moral) support
 from you of late.  The way I look at it is this: I am one of the few people
 who seem to take seriously the idea of tackling real-world problems using the
 formalism of FOL, and as such I had expected that you would react with a 
 little more enthusiasm towards my efforts than you have.  In some (non-self-
 serving) sense I expect a little consideration for trying...  I have always
 had the feeling that you find my ideas banal, not to say boring.


∂19-MAR-75  1526		1,PDQ
 I will need some time to tie up loose ends and find another
 job before I leave.  Could you suggest how much time would
 be reasonable?

∂19-MAR-75  0027		S,LES
 I don't have the Formal Reasoning part of the proposal.
␈ CC: jmc;rww


∂17-MAR-75  2300		DOC,TOB
 Are you going to the Workshop on Cognitive
 Robotics?   If you are interested, it's
 still open to give a keynote address.
 Tm

∂17-MAR-75  0639		network site ISI
 Date: 17 MAR 1975 0639-PDT
 From: FIELDS at USC-ISI
 Subject: KNOWLEDGE REPRESENTATION LANGUAGE
 To:   WINOGRAD at SU-AI
 cc:   MCCARTHY at SU-AI
 
 THE RECENT SU-AI PROPOSAL MENTIONED AN EFFORT
 ON A KNOWLEDGE REPRESENTATION LANGUAGE, AND GAVE A POINTER
 TO AN SU-AI FILE CONTAINING A DOCUMENT. I HAVE BEEN
 UNABLE TO RETRIEVE THE DOCUMENT. COULD TOU PLEASE
 SEND ME SOMETHING TO READ? THANKS
 BEST
 CRAIG FIELDS
 -------

∂17-MAR-75  0637		network site ISI
 Date: 17 MAR 1975 0636-PDT
 From: FIELDS at USC-ISI
 Subject: TELEPHONE PROTOCOL
 To:   RUBIN at SU-AI
 cc:   MCCARTHY at SU-AI
 
 JOHN MCCARTHY TELLS ME THAT YOU HAVE
 DEVELOPED A PROTOCOL FOR LINKING COMPUTERS THAT CAN
 TELEPHONE EACH OTHER AND ANSWER CALLS. THIS IS
 OF GREAT INTEREST TO ME BECAUSE THAT IS PROBABLY JUST
 RIGHT FOR MESSAGE TRAFFIC AMONG INTELLIGENT TERMINALS., ETC ETC.
 COLD YOU SEND ME WHATEVER YOU HAVE IN WRITING?
 THANKS
 BEST
 CRAIG FIELDS
 -------

∂16-MAR-75  2053		105,SGK
 Some things I might do while you are away for you to consider:
 
 Write the Cusp document
 Write an ETV manual
 Design a new editor
 Begin to design a PUB replacment
 Continue work on ETV.
 
 I will only comment now that I think that E is a lost cause.  Any minor
 improvement to E costs about 4 times as much as it should because of the
 lack of organization. And in any case we need a significantly hairier
 editor than E will ever be.  I don't particularly advocate writing a new
 editor for this system, but enough is known about the new system that we
 should start thinking about its editor.  
 
 My personal plans are to stay around here for the summer if not for another
 year.  I don't know what I will decide to do about my education during that
 time.  I do want to move into a place where I have space to hack which
 means that I am going to ask you for something closer to a reasonable wage
 than what I am making now.  It doesn't seem right that you should have to
 practically support me because of random Stanford rules or whatever decided
 how much you were to pay me.

∂16-MAR-75  1331		ACT,REG
 Gary Morganthaller called to say that Tymshare is thinking of building a new
 assembler which will handle multiple relocation counters.  This feature is
 necessary to handle extended addressing in the KL10 properly.
 
 Now, FAIL already has half that feature, and with a little work it might be
 improved to the extent they desire.  My understanding is that if Tymeshare
 wants to use FAIL and if we (meaning me) want to do this mod, then they'll
 recompense us with machine "tyme".  Should I pursue this?

∂16-MAR-75  1145		S,WD
 	Have you considered mailnet.  The "letter protocol" requires a folder-stuffer
 and a postage meter connected to the XGP.   Th "tape protocol" requires a more
 elaboratβ device for putting mag tapes into packages.  This might be easier with
 the IBM3850.

∂15-MAR-75  2101		network site SRI
 Date: 15 MAR 1975 2058-PDT
 From: BOYER at SRI-AI
 To:   JMC at SU-AI
 
 Thanks for the invitation. I am planning to come.
 -------

∂14-MAR-75  1610		REP,DEW
 John, I have the outline of a solitaire system in SOL[REP,DEW].  It is
 fully specified except for 2 disaster routines although not fully
 formalized.  The formalism I'm using makes it hard to express he
 xx the things I want to say.  I'll work on it more and have it in
 better shape before long.  I'm planning on leaving Wednesday for a 
 vacation so would like to talk to you on Monday or Tuesday.  In
 particular I'd like to know where to go with the solitaire thing.
 I tried to simulate a game but in the middle it took about 10 minutes
 per card to figure everything out and I reverted to just doing what I
 thought the system would do.  We won the game but the cards were
 favorable.    Dave

∂14-MAR-75  1613		AP,EK
 Fred Moore from Homebrew Computer Club wanted me to tell you
 that they are meeting here at seven on wed nite.   Paul

∂12-MAR-75  1124		1,PAW
 monday fine for 36,000 mile servicing a Peninsula Mazada...patte

∂12-MAR-75  1027		RAP,MG
 Subject: Location of Meeting
 
 The next meeting  of the discussion  group (at which Jim  Morris will
 describe work  he's doing with Ben Wegbreit)  will be in the ISHIZAKA
 room (S 117)  in the  INTERNATIONAL BUILDING  at SRI on  march 24  at
 3:30 pm. Coffee and cookies will be served.

∂11-MAR-75  1530		S,WD
 TIDBIT:	The documentation which traveled with Skylab weighed on the order of one
 hundred kelos.  Most of this doWmentation, primarily stowage lists, could be
 replaced by a houskeeping computdr.  Maybe such a system has enough in common
 with a home computer that NASA could be touched for some funds.

∂11-MAR-75  1139		1,MG
 Subject: Next Meeting of Program Reasoning Group. 
 
 At the next meeting Jim Morris will describe some recent work that he
 and Ben Wegbreit have done.  The  meeting will be on march 24 at 3:30
 p.m. in SRI (I'll fix up an exact room later)

∂10-MAR-75  1711		DOC,TOB
 I hope that you will be coming to the
 workshop on Cognitive Robotic Systems
 next week, March 20-22.  If you feel 
 like saying a few words as a keynote
 speaker, that is still open.
 Tom

∂10-MAR-75  1125		network site ISI
 Date: 10 MAR 1975 1125-PDT
 From: ABRAMSON at USC-ISI
 Subject: HEART-SHER POSITION PAPER
 To:   DAVID-WALDEN at BBN
 cc:   ABBOTT at ISI, ABRAMSON, AMAREL, BERNSTEIN at BBN,
 cc:   BURCHFIEL at BBN, BRYAN, CERF, CHEATHAM at HARV-10,
 cc:   COOPER at BBN, CHI at ISI, DERTOUZOS.MAC at MIT-MULTICS,
 cc:   FARBER at ISI, FEIGENBAUM, FRALICK at SRI-AI, FRANK at ISI,
 cc:   HEART at BBN, HOLT at SRI-ARC, HUMMELL at ISI, KIRSTEIN,
 cc:   KLEINROCK, MAGILL at SRI-AI, JMC at SU-AI, MEDRESS at BBN,
 cc:   NEWELL at CMU-10A, NILSSON at SRI-ARC, OMALLEY at BBN,
 cc:   PIRTLE at I4-TENEX, PRATT at ISI, SALTZER at MIT-MULTICS,
 cc:   STOCKHAM at UTAH, SUTHERLAND at BBN,
 cc:   TEITELMAN at PARC-MAXC, UNCAPHER at ISI, WALKER at SRI-AI,
 cc:   WARSHALL at SRI-ARC, WATSON at SRI-ARC,
 cc:   WIENER at RAND-RCC, WINSTON at MIT-AI, WOODS at BBN,
 cc:   CARLSTROM at ISI, KAHN, WALKER, RUSSELL, LICKLIDER, CARLSON,
 cc:   FIELDS, STUBBS, BLUE
 
 
 IN RESPONSE TO YOUR SAN DIEGO POSITION PAPER OF MARCH 7:
 
 WE HAVE STARTED SOME WORK ON A NARROW PROJECT WHICH  IS CONTAINED
 IN THE AREA DESCRIBED IN YOUR PAPER.  WE HAVE BEGUN TO DO SOME
 SPECULATING  ABOUT USING PACKET BROADCASTING WITHIN A MULTIPROCESSOR
 SYSTEM  --- A PACKET BROADCASTING UNIBUS IF YOU WILL.  WE HAVE ALSO 
 HAD SOME CONTACT WITH RAJ REDDY AND SAM FULLER AT CMU ON THIS
 AND WE HOPE TO EXPAND INTO SOME JOINT WORK HERE.  IN ESSENCE WE 
 FEEL THE CONNECTIVITY PROBLEM WITHIN A SYSTEM OF THOUSANDS OF
 MICROPROCESSOR AND MICROMEMORY CHIPS MIGHT BE LOOKED AT FROM
 THE POINT OF VIEW OF PACKET BROADCASTING.  EACH ENTITY TO BE
 CONNECTED MIGHT HAVE A MICRO-TCU, CONSISTING OF A BUFFER, MODEM AND 
 MICROTRANSCEIVER (ORIGINALLY RADIO, BUT WE ARE NOW TENDING TOWARD
 INCOHERENT LIGHT EMITTERS SUGGESTED BY KAHN) EMITTING MICROWATTS.
 
 AS WE SEE IT YOUR SUGGESTION IS COMPOSED OF TWO IDEAS ---
 (A) PACKET SWITCHING
 (B) PACKET BROADCASTING
 AND THE FLEXIBILITY YOU MENTION IN PROVIDED BY THE CONNECTION FREEDOM
 OF IDEA (B) MORE THAN IDEA (A).  FOR A GOOD TREATMENT OF THE
 DISTINCTION SEE METCALFE'S PROJECT MAC REPORT --- PACKET BROADCASTING.
 
 SO FAR WE HAVE A ROUGH DRAFT OF THE WORK OF ONE UNDERGRADUATE ON THESE 
 IDEAS.  IT IS RATHER UNSOPHISTICATED, BUT IF YOU WANT TO LOOK AT IT 
 (18 PAGES) SEE <ALOHA>MULTI-MICRO-PROC-PAPER AT ISI.
 
 -------

∂09-MAR-75  1650		S,WD
 	Ves Marinov called in connection with visa to the Soviet Union.

∂07-MAR-75  2159		105,SGK
 Sorry about the SX-70.  When I heard you were leaving I thought you might
 want it, but hadn't had the chance to transport it or anything else I had
 packed in my bag.

∂7-MAR-75  1341		network site BBNB
 Date:  7 MAR 1975 1603-EDT
 From: SUTHERLAND at BBN-TENEXB
 Subject: more new program comments
 To:   abbott at ISI, abramson at ISI, amarel at ISI,
 To:   bernstein at BBN, burchfiel at BBN, bryan at ISI,
 To:   cerf at ISI, cheatham at HARV-10, cooper at BBN, chi at ISI,
 To:   dertouzos.mac at MIT-MULTICS, dickson at BBN,
 To:   farber at ISI, feigenbaum at ISI, fralick at SRI-AI,
 To:   frank at ISI, heart at BBN, holt at BBN, hummel at ISI,
 To:   kirstein at ISI, kleinrock at ISI, lebow at BBN,
 To:   magill at SRI-AI, marill at CCA, jmc at SU-AI,
 To:   medress at BBN, newell at CMU-10A, nilsson at SRI-AI,
 To:   norton at SRI-ARC, omalley at BBN, pirtle at I4-TENEX,
 To:   pratt at ISI, saltzer at MIT-MULTICS, stockham at UTAH,
 To:   teitelman at PARC-MAXC, uncapher at ISI, walker at SRI-AI,
 To:   warshall at SRI-ARC, watson at SRI-ARC, weiner at RAND-RCC,
 To:   winston at MIT-AI, woods at BBN, wintz-purdue at I4-TENEX,
 To:   licklider at BBN, carlstrom at BBN
 cc:   sutherland at RAND-RCC, kahn at ISI
 
 I have made a further collection of various documents I put together
 in the last year which comment on new program issues.  In case anyone
 wants to read them they will shortly appear as
 
 [bbn]<documentation>new-program.thoughts
 
 in the bbn (bbnc) documentation directory.  Anyone interested can
 print theH out and read.
 I'm afraid that they may be somewhat rambling at times , but there
 they are.  Comments are welcome.
 
 Bert Sutherland
 -------

∂7-MAR-75  1217		network site BBNA
 Date:  7 MAR 1975 1511-EDT
 From: DAVID-WALDEN at BBN-TENEXA
 To:   ABBOTT at ISI, ABRAMSON at ISI, AMAREL at ISI, BERNSTEIN,
 To:   BURCHFIEL, BRYAN at ISI, CERF at ISI, CHEATHAM at HARV-10,
 To:   COOPER, CHI at ISI, DERTOUZOS.MAC at MIT-MULTICS,
 To:   FARBER at ISI, FEIGENBAUM at ISI, FRALICK at SRI-AI,
 To:   FRANK at ISI, HEART, HOLT at SRI-ARC, HUMMELL at ISI,
 To:   KIRSTEIN at ISI, KLEINROCK at ISI, MAGILL at SRI-AI,
 To:   JMC at SU-AI, MEDRESS, NEWELL at CMU-10A,
 To:   NILSSON at SRI-ARC, OMALLEY, PIRTLE at I4-TENEX,
 To:   PRATT at ISI, SALTZER at MIT-MULTICS, STOCKHAM at UTAH,
 To:   SUTHERLAND, TEITELMAN at PARC-MAXC, UNCAPHER at ISI,
 To:   WALKER at SRI-AI, WARSHALL at SRI-ARC, WATSON at SRI-ARC,
 To:   WIENER at RAND-RCC, WINSTON at MIT-AI, WOODS,
 To:   CARLSTROM at ISI, KAHN at ISI, WALKER at ISI,
 To:   RUSSELL at ISI, LICKLIDER at ISI, CARLSON at ISI,
 To:   FIELDS at ISI, STUBBS at ISI, BLUE at ISI
 cc:   WALDEN
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Position Paper for ARPA IPTO Meeting
 in San Diego, March 1975
 
 Control and Monitoring Systems Using Packet Technology
 
                    F. Heart, L. Sher
               Computer Systems Division
               Bolt Beranek and Newman Inc.
 
 
 
 
 
 	A major advantage of packet-switching technology is that
 one is able to use less communications bandwidth by more effectively
 multiplexing many different users over single pieces of bandwidth.
 To date, the packet-switching technology has been considered mainly
 in relation to communications between people and computers, computers
 and computers, and people and people.  There is an entirely separate
 class of possible applications which effectively uses some of the
 techniques learned during the development of the ARPA Network packet-
 switching technology:  applications for monitoring and control.
 
 	Many large and complicated systems in both the civilian and
 military sectors are built with a very large number of wires, e.g.,
 automobiles, ships, airplanes, and buildings.  An airplane has many
 miles of wire connecting all the controls and indicators with all
 the actuators and all the sensors.  A ship has a similar abundance of
 wires.  An office building uses many wires for alarm systems,
 heating systems, lighting systems, etc.  Whereas such wires--on the
 spool--are usually inexpensive, the related costs are not.  In all
 cases, wires require installation, connectors, servicing provisions,
 failure-mode provisions, and documentation, any or all of which may
 cost much more than the wires themselves.  In aircraft, additional
 related costs arise from considerations of weight, reliability,
 shielding from electrical noise sources, possible retrofits and
 modifications, periodic reconfiguration (e.g., passengers vs. cargo),
 interchangeability of the system pieces (with those in other
 aircraft types), inventory of spares, etc.  All such costs are
 impacted if packet technology is substituted for "multiwire", and
 it would appear that the impact could be made to be favorable in
 almost every case.
 
 	A multiplexing scheme using packet technology has the key
 attribute that it can compress a large portion of the complexity
 of a monitoring and/or control system into a set of near identical
 little boxes, and such boxes have been getting smaller and cheaper
 at a stunning rate.  Thus, one can easily imagine a set of
 standardized modules, one used at each control, indicator,
 actuator, or sensor.  Interconnecting these modules would be a
 single communications medium, using guided or unguided electro-
 magnetic or acoustic energy, e.g., a wire, a coaxial cable,
 a waveguide, the skin of an airplane, the electrical power line,
 the hydraulic power lines, optical fibers, etc.  Sensors and
 controls would launch addressed packets into the communications
 medium.  Indicators and actuators would listen for and acknowledge
 correct receipt of any packet addressed to them.  Packet routing would
 probably be quite different from ARPANET protocols, and all
 instruments* would probably share a common communications channel,
 as in the ALOHA protocols.
 
 
 ←←←←←←←←←←←←←←←
 *For this purpose, an instrument is any indicator, sensor, actuator,
 or control.
 
 
 
 
 	Refinements of numerous types are possible:  Sensors could
 send packets at a rate suitable only for resolving changes in the
 parameter sensed.  (If there is no change, sensors could just send
 packets periodically.)  Data of low or high bandwidth, if not
 needed in real time, could be time compressed or expanded as
 needed in order to get into and out of the packet format.  Polling
 could be used in all, some, or no data transmissions.  A
 system could run continuously-interleaved self-diagnostics.  The
 communications medium could have various forms of redundancy making
 it highly immune to failure.  It could be optical, which would
 be fully immune to electrical noise.
 
 	Adding any sending, receiving or sending/receiving element
 would require a near-irreducible minimum of effort.  For example,
 one could easily imagine an airplane in which any of the cockpit
 instruments could be plugged into any cockpit instrument location.
 
 	Except for serial vs. parallel data paths, this concept of
 a packet-carrying "information bus" is not unlike the PDP-11's
 Unibus, which has since been widely copied.  A similar kind of
 "information bus" has now been proposed by Hewlett-Packard for
 industry-wide adoption, so that various instruments of different
 manufacturers--and of one manufacturer--can easily be plugged
 together.
 
 	Some of the interesting questions about using packet
 technology for control and monitoring systems are the following:
 
 	1.  What are the projected costs of the modules which
 interface instruments to the packet-carrying communications medium?
 
 	2.  In what kinds of applications are the costs first likely
 to be acceptable?
 
 	3.  Can interface modules become so standardized that one
 type or a very few types could serve many users and uses?
 
 	4.  In what application areas might it be reasonable to
 expect this technology to transfer from the military to the
 civilian sector?
 
 	We can now speculate on some of these issues.
 
 	On costs (question #1), we must make some assumptions.
 First we assume a communication medium that has a useful bandwidth
 of somewhere between 1Mb and 100Mb (even though for some
 applications, 1Mb would be overkill).  This bandwidth would
 probably permit error-free intercommunications of a hundred or
 so to a few thousand common kinds of instruments.  An interface
 module for this purpose, if built in 1975, would probably have
 10 to 30 IC's and would probably cost $1000.  Current trends,
 however, suggest that in 5 to 10 years, one or two IC's at $10
 is not unreasonable, particularly if they are not too highly
 specialized.  One can easily imagine this interfacing module
 incorporated either in the instrument or in a connector.
 
 
 
 	On applications (question #2), one cannot avoid 
 spacecraft and aircraft as the most likely candidates, both having
 extensive, stringent requirements and a high cost overhead per
 installed wire.  (But, as noted, there is a possible commonality
 of parts with other monitoring and control systems, which suggests
 that in more cost sensitive applications, high parts costs might
 be offset by low design and development costs.)  The rising
 popularity of fly-by-wire techniques also bodes well for the
 high damage-resistance possible in a packet-mediated control
 system.
 
 	Standardization of interface modules (question #3), appears
 to be limited by two primary factors, both of which may be
 overcome, at least partially:  (a)  Different kinds of communication
 media, e.g., fiber optics vs. coaxial cable, obviously require
 different kinds of coupling to them.  The solution here appears to
 be a two-part interface module, one part which does the logic
 electronically and the other part which couples energy to and from
 the communications medium.  (b)  Applications requiring vastly
 different minimum speeds of communication are easily imagined.
 The solution here may be two or three different speed-families,
 or possibly using a single fast family of logic modules for all
 applications, slowing the bit rate (within each packet) by
 suitable fast-in slow-out shift registers where necessary.
 
 	Fallout of this packet technology into civilian life
 (question #4) may appear in such places as automobile traffic
 control, mass transit and personal rapid transit (PRT) systems,
 monitoring and control in large buildings, weather sensor arrays,
 fire sensor arrays, and commercial aircraft.  It is interesting
 to observe that multiplexing to reduce the number of wires has
 already been done in the 10-channel audio and the light switches
 at each seat in a jumbo jet and has been seriously proposed and
 demonstrated for automotive use by DuPont using digital multiplexing
 over plastic fiber optics*.
 
 
 ←←←←←←←←←←←←←←←
 *Design News 30 (No. 4), 34, (Feb. 17) 1975.
 
 
 
 
 
 
 	We suggest that now, as the needed technology is so rapidly
 appearing, the use of packets for monitoring and control should be
 studied.  The primary issue for study is a technologic-economic mix:
 What kinds of development are needed to assure early yet prudent and
 economic deployment of such a capability?  Other important issues
 are a quantification of costs and benefits for likely areas of
 application, and key technologic hurdles.
 
 
 -------

∂06-MAR-75  2147		105,SGK
 Is this a good time to ask you about borrowing some money for a few weeks?
 I have applied for random Master Charge card.

∂06-MAR-75  1641		LSP,AJT
 Now that I've committed myself to going away in September, can I be
 reasonably sure that you will support me during the summer? (It's a 
 question of necessity rather than convenience: in the absence of that
 support I shall not be able to stay during the summer).
 thanks. a.

∂06-MAR-75  1356		1,ECH
 Dear Pro. J. McCarthy,
   Unfortunately, I have to leave your very attractive laboratory.
   Thank you very much for giving me a chance to work at AI Lab., and good bye.
             March 6,    Eiichi Miyamoto

∂06-MAR-75  1156		1,BH
 Re ACMization of DECUS paper, here are what I think are the important ideas:
 1. Even programs which do not obviously need a display capability (e.g.,
    SPELL) can be vastly improved by exploiting display features.
 2. Display support should be built into the system at a low level in a
    simple and uniform way, not programmed separately in each user program.
 3. It is okay and maybe even preferable to centralize the intelligence in
    the terminal system.
 4. Very high baud rates make a qualitative difference in what you can do.
    (e.g., WHO, NS, E)
 5. 2-D display capability is intimately connected to the ability to control
    multiple procedures effectively.  Users need this ability.
 Of these, I think the last is the most important, followed by #2.  Do you
 have any comments?

∂05-MAR-75  1532		LSP,AJT
 reminder - committee meeting tomorrow (Thurs) 2:00 in Shepard's office:
 Jordan 356
 a.

∂05-MAR-75  1129		1,JB
 I'd like to see you about my work for the coming quarter.

∂26-FEB-75  1034		1,MG
 Is there any news about money for Dr. John Darlington to visit here?


∂04-MAR-75  0957		ACT,REG
 Yes, I saw your comments to Les.  I believe the revised proposal
 incorporates most, if not all, of your suggestions.  There are still a
 couple of things I'd like to improve.  Lick says essentially that ARPA
 doesn't want to spend any money on operating systems "unless the need is
 specific and thoroughly justified." I believe we've made a case for the
 display work, because Lick himself already believes in it to some extent.
 I feel the file system thing is a little shakey.  I'd like to firm it up
 by adding more arguments in its favor (I've already added examples of
 why the present schemes are inadequate and that some of the MULTICS work
 is relevant so we don't have to reinvent things) but I feel there's a
 lack of conviction that the proposal is "thoroughly justified."

∂03-MAR-75  2044		ACT,REG
 A draft of the revised proposal is PROP75.A[DOC,REG].  I welcome your comments.

∂03-MAR-75  1942		1,BH
 In RCV, typing "C ?" or "T ?" to an option request will tell you the name
 of the current output file and allow you to specify another.


∂3-MAR-75  1154		network site AI
 Date: 3 MAR 1975 1457-EST
 From: PHW at MIT-AI
 To: jmc at SU-AI, les at SU-AI, minsky at MIT-AI
 
 Hi,
 I have not received your vote for the C&T lecture and I need them
 desparately.  Could you please rank the following and return via
 the net immediately.
 
 bobrow
 buchanan
 clowes
 fikes
 goldstein
 green
 hayes
 papert
 reddy
 reiger
 sacerdoti
 sandewall
 schank
 siklossy
 sussman
 wegbreit
 wilks
 woods

∂03-MAR-75  0956		1,MG
 Subject: Richard Waldinger talks at PARC today. 
 
 The meeting  is at Xerox  PARC today at  3:30 p.m. Richard  will talk
 about his recent work on achieving several goals simultaneously. 
 
 Mike. 
 
 P.S. We still need a volunteer for next time. 

∂28-FEB-75  1119		S,WD
 	The proposal is LISP.PRO[WD,JMC].  I believe Patte will have left a
 copy on you desk.

∂27-FEB-75  1146		L,FWH
 Subject: What Richard will talk about at next discussion group meeting. 
 
 At the next meeting  (3:30 on 3 mar. at PARC)  Richard Waldinger will
 describe   his  recent   work  on  how   to  achieve   several  goals
 simultaneously.  He'll illustrate his methods by showing how they can
 be used  to generate a  program for  interchanging the values  of two
 variables. 
 
 MG


∂1-MAR-75  1711		network site SRI
 Date:  1 MAR 1975 1718-PDT
 From: BOYER at SRI-AI
 Subject: AI FORUM - Delegate safety at IJCAI4
 To:   sigart at CMU-10B
 cc:   erik at MIT-AI, jmc at SU-AI
 
 
 IJCAI4
 FROM: Vesko Marinov
 IMSSS, Stanford University				March 1, 1975
 Stanford, CA94304
 
 
 	The safety of the delegates to an international scientific
 conference is, in my opinion, essential for the normal conduct of the
 conference.  Furthermore it is important for the success of a 
 conference that anybody qualified and wishing to attend should feel free
 to do so.
 	I welcome the choice of Tbilisi as the site of IJCAI4, but only
 under the condition that all prospective delegates should feel safe
 to attend.  However, there are reasons to believe that the safety of
 some delegates, particularly those of East European origin, may be
 jeopardized by going to Tbilisi.  
 	Therefore I suggest that the organizing committee, including
 the Soviet organizers, coordinate opinions with the Soviet authorities
 and issue an oficial statement guaranteeing the safe attendance of
 all delegates.  I appeal for the support of the AI community in obtaining
 such a guarantee.
 
 -------

∂01-MAR-75  1706		REP,DEW
 There's a finished copy of my paper on your desk.  Please throw the old
 one away as soon as possible since it had some errors in it thathave been corrented
 .  Also the last example is the most instructive one.  I would like
 suggestions again on what to do next.   Dave

∂01-MAR-75  1320		1,ELF
 I AM NOW TO DAMN SCARED TO TALK TO YOU AND IM GETTING SICK OF IT

∂28-FEB-75  1659		2,JH
 My contribution to expanding news75.pro is on [2,jh].  The expanded parts
 are mostly on the area of man-machine interaction. Jim

∂28-FEB-75  0615		network site ISI
 Date: 28 FEB 1975 0614-PDT
 From: CARLSTROM at USC-ISI
 Subject: IU WORKSHOP
 To:   JMC at SU-AI
 cc:   CARLSTROM
 
 THANKS FOR YOUR NOTE.  I'VE TALKED WITH BINFOORD ON THE PHONE  AND
 I THINK EVERYTHING WILL BE FINE.
 
 REGARDS,
 DAVE C.
 -------


∂27-FEB-75  1537		PNT,DCL
 YES-MARCH 11 AND 13 ARE FINE. WHAT IS THE ROOM NO. OF YOUR CLASS?

∂27-FEB-75  0808		E,ALS
 The trouble is with the MAIL-E interface in that MAIL does not pass on to E
 the information that E needs. I Will look into it however and see if I can
 get it fixed.	ALS


∂26-FEB-75  1355		1,MG
 Subject: Next meeting of discussion group
 
 The next meeting, at which Richard Waldinger  will talk, is on monday
 3 march at 3:30 p.m.in PARC. 
 
 After  that we run out  of volunteers. Would anyone  like to describe
 what they  have been  up to  lately or  what  their current  problems
 and/or unimplemented future plans are? 
 
 Mike

∂26-FEB-75  1337		THE,AJT
 since you appear to be leaving on the 7th., it is essential
 that you let me know whether you can make it to my committee
 meeting on Wed. 6th. provisionally at 2:00. please let me know
 a.

∂25-FEB-75  1631		NS,ME
 I looked at the proposal again but have no suggestions.
 I'll be gone till Friday night.

∂24-FEB-75  1515		THE,AJT
 Hintikka can't come(aaarrrgh!) to the Thursday meeting. Other possible
 times are: any day next week in the morning, or between 2 and 4. Let me
 know a possibility list and I'll try to achieve a magical correlation with
 other people's.
 sorry about all this... arthur.

∂24-FEB-75  1211		1,QIB
 TENTATIVE:  A.I. SCIENCE LUNCHEON WITH LEDERBERG, FACULTY CLUB, 12:00,
 MARCH 10 - MAY I CONFIRM?

∂24-FEB-75  1206		1,QIB
 REMINDER:  TAXI WILL PICK YOU UP AT THE FRONT ENTRANCE TO THE FACULTY
 CLUB AT 1:00 TO TAKE YOU TO THE AIRPORT - 2-25 (TUESDAY).
 
 PROF. FEIGENBAUM NEEDS A COPY OF THE LETTER ON THE ARPA PRINCIPLE
 INVESTIGATORS MEETING IN MARCH.
 

∂24-FEB-75  1152		VCG,DCL
 Were we supposed to meet today (Mon. Feb-24) to continue?
␈ CC: JMC;TOB;RWW;TW;DCL

∂23-FEB-75  2224		1,DBX
 I have put a copy of my write-up on recursion to your desk. I hope it
 is more or less correct now. We are going to discuss it with Richard
 and Bill to-morrow and I hope we can talk on Friday, which will be my
 last day here.

∂23-FEB-75  0055		S,LES
 It is now Sunday Morning and I still have not received any contributions
 for the ARPA PI Meeting Summary, except for one written by Brian McCune
 on behalf of Cordell.  I should be pissed off, but it is hard to get
 worked up when this kind of performance has become standard.
 
 I have hacked something together in ARPAPI.75 [D,LES].  It is mainly the
 list of bullets that I compiled several months ago.  Perhaps if you show
 it to them they will be moved to write something more up-to-date.
 
 According to the Meeting Announcement, we were supposed to send it to
 FEINLER @ NIC by last Thursday.  Hopefully, it will happen early this
 week.
 
 I will try to check in by phone on Tuesday, or sooner if there is a
 phone at the place where we are staying.

∂21-FEB-75  1811		network site SRI
 Date: 21 FEB 1975 1811-PST
 From: BOYER at SRI-AI
 Subject: scientism
 To:   jmc at SU-AI
 
 I wonder if there is still a plan for a meeting tonight at
 your house on the subject of a bay area scientism club.
 -------

∂11-FEB-75  1225		1,MG
 John Darlington (a collegue of Rod Burstall who's at present visiting
 IBM Yorktown  Heights) would like to visit Stanford  and SRI in a few
 months time.  Richard Waldinger tells me that SRI can only offer $150
 and as this  won't cover the cost was wondering  whether Stanford had
 any money available which could be combined with the above to produce
 a reasonable sum. 
 
 Mike. 


∂08-FEB-75  1719		1,SGK
 Call David McQueen 513-8787295 office 255-3098. Feferman has folder.


∂11-FEB-75  1231		network site ISIA
 Date: 11 FEB 1975 1229-PST
 From: NBS at USC-ISIA
 Subject: TIMESHARING COOP
 To:   JMC at SU-AI
 cc:   NBS
 
 DEAR DR. MCCARTHY,
 
      IREAD A COPY OF THE PROPOSAL FOR A TIMESHARING COOP,
 AND IT SOUNDS VERY INTERESTING TO ME.  THEREFORE I WOULD
 
 BE WILLING TO AND INTERESTED IN TALKING ABOUT IT WITH
 OTHER PEOPLE, AND IN HELPING TO SET IT UP.
 
      I AM AN UNDERGRAD AT STANFORD, AND HAVE A FAIRLY EXTENSIVE BACKGROUND
 
 IN COMPUTING.
 
 I CAN BE REACHED VIA THE FOLLOWING:
 
 ARPA-NET       NBS@ISI   (PLEASE START MESSAGE WITH "FOR ED FRANK"
 
 SLAC-TRIPLEX   SUGGEST TO EHF$EA
 
 UNCLE SAM
 LETTER SINK     ED FRANK
                P.O BOX 2008
                STANFORD, CA
 
 MA BELL        415-321-6143
 
                 SINCERELY,
                 ED FRANK
 -------


∂11-FEB-75  1407		DOC,TOB
 There are two positive contributions by
 Don Gennery.  As a course project, he made
 a search for calibration solution that was
 successful, where Hannah had been unsuccessful
 before.  At SRI, he made a horizontal 
 rectangle bounder; JMT's opinion is that he
 did a good job.  That he had good ideas, was
 competent and a good executor, hard worker.
 Along the lines of Perkins, not a deep thinker
 but reasonable and competent.
 There is of course the negative information
 about not working his first year here.
 Tom


∂08-FEB-75  0139		100,100: sgk @ SAIL
 Have invited Schroeppel & c. to stay here the nite of the 22nd.  Say if OK soon pls.


∂05-FEB-75  1503		100,100: S/WD @ SAIL
 	I don't know if you found the note.  Telco was here about the phone problem
 while you were away and claims that you need an interface for you Phonemate.


∂05-FEB-75  1415		1,ELF
 READ MADMAD.MAD LAST PART[1,ELF]


∂05-FEB-75  1132		1,MG
 Subject: Change of time or place of meeting. 
 
 I've just  discovered that  the conference  room at  SU-AI is  booked
 every monday between  4:15 and 5:15. Thus either the time or place of
 the next meeting of  the program reasoning  group must be changed.  I
 propose that it be moved to tuesday  18 feb. (the day after it was to
 be).  Please let me know if  this is O.K. If  it isn't maybe we could
 still have the meeting on monday but at SRI or PARC? 
 
 I'll confirm the new time and place in a day or so. 
 
 Sorry about this. Mike


∂04-FEB-75  1202		P,JRA
 cs206?

∂04-FEB-75  1137		1,MG
 Subject: Future meetings of Program Reasoning Group. 
 
 The next meeting is  in the conference room at Stanford  A.I. Lab. at
 3.30  p.m.  on mon.  17 feb. Balint Domolki  (who's visiting the  A.I.
 Lab. from Hungary) will talk. 
 
 The meeting following that will be at Xerox PARC at 3.30 p.m. on mon.
 3 march.  Richard Waldinger will talk. 
 
 The debate described in previous  messages (which was conceived of by
 Bob.  Boyer)  is  to be  organanised  by  Richard  Waldinger. Further
 suggestions concerning it should be sent to Richard at SRI. 
 
 Sorry about the  vast list of names  that non-SU-AI people have  been
 getting at  the top of  previous messages -  I had no  idea they were
 appearing!  I hope I've now found the right hack to suppress them. 
 
 If you would  like to  stop getting these  messages, or  if you  know
 someone who'd like to start getting them, please let me know. 
 
 Mike

∂04-FEB-75  0758		100,100: sgk @ SAIL
 Heave left your camera by the imlac.  I bought more film for you, and
 then proceeded to take nine shots...  Will get more eventually.  TNX
 Essay debugging is progressing despite E's efforts to drive me to the
 nut house.


∂03-FEB-75  1102		THE,AJT
 what do you know about the funny Schloss Laxenburg placein Vienna - can
 you give me some idea of th sorts of things that they're doing, and of 
 your own opinion of the place. thanks. arthur

∂03-FEB-75  0951		1,MG
 Subject:  **** FREE COFFEE AND COOKIES ****
 
 Today's meeting of the program reasoning group - in which J moore
 will discuss the new improved LISP theorem prover - will be at 
 3.30 in SRI's I building room s109 (the keidan room).
 
          ************************************
          * There will be  FREE refreshments *
          ************************************
 


∂02-FEB-75  1839		1,PMK AT TTY106   1839
 John,
 Would you have some time to talk with me tomorrow? I saw you when
 I came in tonight but you looked tied up, then you disappeared!
 
 I plan to be here most of tomorrow, so if you have some free
 time would you please let me know.
 
 Thanks, Peggy


∂31-JAN-75  1354		network site ISIA
 Date: 31 JAN 1975 1334-PST
 From: LICKLIDER at USC-ISIA
 Subject: Hieronymus
 To:   McCarthy at SU-AI
 cc:   Licklider
 
 	John, I do not know him well enough to be helpful.
 He seemed bright and interesting when he was here for a short visit,
 but I did not get enough of a fix on him to be sure whether I'd
 want to hire him, myself, or urge you to keep him
 hired.
 
 	My consideration of your proposal was interrupted by
 paperwork requirements levied by the new Director and by
 the need to get out several Memos Requesting ARPA Orders before
 COB today.  I'll get back at it shortly.  I'll have to ask for
 considerable sharpening of the objective structure; it now reads
 as though there are a lot of miscellaneous objectives to be reached
 in a year or two, but that it is not clear what roads they are on,
 what major goals they are subordinate to.  Also, I'll have to ask for
 a major reconsideration of the equipment section, which sounds as though
 it is heading for another ideosyncratic system.  (I have an idea
 about how to handle that, but I can't describe it now.)  Sorry to 
 have to noodle about on the proposal.  Al Blue got us an extension
 to cover the time required for noodling without losing the money.
 But we'll have to work pretty fast.  I assume I should communicate
 with LES -- with carbons to you?
 
 				Regards
 
 				Lick
 -------