perm filename JULY.MSG[ESS,JMC] blob
sn#113553 filedate 1974-07-31 generic text, type T, neo UTF8
∂29-JUL-74 1257 J,REF
May I see you this afternoon?
Thanks --Bob.
∂26-JUL-74 2008 CAR,HPM
Re the affect of denying me hardware on my programming:
I've been denyed hardware ever since starting to work on the
cart, and it's mainly that fact that has kept me hardware hacking,
trying to put something together out of scraps. What I've been
bargaining for in the current case is something that makes such
hacking unnecessary, namely a complete, working, transmitter. If you
saw fit to allocate a modest amount of technician time to the cart,
I'd even promise to keep my hands off hardware altogether, and work
diligently, and full time, on the programs needed to make the cart
go.
Re the need for hardware:
The forty or so pictures on Baumgart's area are already
starting to nauseate me, and they fill his allocation in spite of the
fact that he has the largest disk area at the lab. It's pretty
obvious that even at this stage my programs are being optimized to
work with that set, at the expense of the wider variety of real road
scenes. A working transmitter would provide thousands of times the
variety available now.
∂26-JUL-74 1134 P,JRA
I've been incorporation the ideas in M. Gordon's thesis into
my LISP notes; it works very nicely, giving an intuitive motivation
for Scott's work, and lends support to a natural extension of LISP
to user-defined data structures. It should be reasonable within a week;
if you would like to see it I would be happy to oblige. I am still very
interested in teaching 206 this fall if at all possible.
∂26-JUL-74 0936 1,KIC AT TTY57 0936
JOHN, I WOULD LIKE TO MEET WITH YOU FOR A DISCUSSION OF MY THESIS PLANS AND WHAT
TO DO NEXT ETC.ANY TIME THAT IS CONVENIENT TO YOU SUITS ME FINE. KICHA
∂24-JUL-74 1612 FOO,DBA
(DE BOO (BAR BDR ARG)(BAR (BDR ARG)))
(DE COO (CAR CDR ARG)(CAR (CDR ARG)))
(LAP BOO SUBR)
(PUSH P 1)
(MOVE 1 3)
(PUSH P 2)
(CALLF@ 1 0 P)
(CALLF@ 1 -1 P)
(SUB P (C 2 0 2 0))
(POPJ P)
NIL
(LAP COO SUBR)
(HRRZ@ 1 3)
(HLRZ@ 1 1)
(POPJ P)
NIL
α-conversion, anyone?
∂23-JUL-74 1349 1,KNM
dr mears of the aaas phoned and he needs a tentative program on your part of
the conference on robotics by august 1st. a 2 page abstract would be fine.
∂19-JUL-74 1953 1,PDQ
Continuation of previous message: ... and have a good lead on NASA support
for roving vehicle studies. Perhaps I can follow up on that lead next
week in DC.
∂19-JUL-74 1953 1,PDQ
;I just returned (Fri 8 PM) from a Viking meeting, and have a
∂19-JUL-74 1350 1,KNM
PROBLEM
KOTOV AND WIFE WILL BE ARRIVING MONDAY AT 2:21--SINCE YOU ARE GOING TO BE GONE,
WHAT IS TO BE DONE ABOUT THEM? FROM TALKING TO THE SECRETARY AT U OF WASH,
I ASSUME THEY EXPECT TO BE MET. ALSO WHERE ARE THEY GOING TO STAY--I DON'T THINK
THEY HAVE MADE ANY ARRANGEMENTS. I MAY OR MAYNOT BE ABLE TO MEET THEM AT AIRPORT
AS I GO IN FOR MINOR SURGERY MONDAY AND DON'T YET KNOW THE TIME.
∂19-JUL-74 1353 1,KNM
MESSAGE
SHUBIK (THE RUSSIAN FRIEND OF FEIGENBAUM) HAS CALLED 3 TIMES TODAY AND IS MOST
ANXIOUS TO GET A HOLD OF YOU. HE HASN'T LEFT A NUMBER.
∂19-JUL-74 1348 1,KNM
PHONE MESSAGE
19 JULY 13:48
PHIL JACKSON CALLED AGAIN AND WANTS YOU TO CALL 322-0171--LET PHONE RING AS
HE MIGHT BE IN THE BACKYARD GARDENING.
∂19-JUL-74 1344 1,KNM
PHONE MESSAGE
19-JULY 13:44
MR MADILEN OF YARDLEY, PENNS. CALLED IN REFERENCE TO THE YEAR END SEMINAR OF
AAAS ON AUTOMOBILES AND THE ENVIRONMENT. HE WAS REFERRED TO YOU BY PAUL
ROBBINS. HE WANTS YOU TO CALL (215) 493-4033 AND WILL BE IN UNTIL 3 OUR TIME.
∂19-JUL-74 1039 1,KNM
PHONE MESSAGE
19-JULY 10:45
PROG. BECKMANN PHONED AND WOULD APPRECIATE IT IF YOU CALLED HIM BACK TODAY AT
(303) 444-0841.
∂18-JUL-74 1244 105,SGK
Jeff Eastman called you early this morning. He said he had sent
you a letter, and would like to arrange a time to come out here
and talk to lab folk. His number is 322-9206
∂17-JUL-74 1350 1,PJH
john, have you any comment on the contents of META>AXMS ? I would
appreciate any criticism. Iwill be back
in two weeks. pat.
∂16-JUL-74 1350 THE,AJT
you never did give me your detailed comments on FOLMAN.PUB[DOC,AJT] -
the FOL manual. a.
∂14-JUL-74 1208 ACT,REG
[The following notice is provided for your amusement, or whatever. It came
from the gripe file. - REG]
∂14-JUL-74 1006 NET,GUE AT TTY121 : GORDON LIDDY @ network site RTGT
WHEN IS THE NEWS SERVICE GOING TO BE PUT UP FOR US
REMOTE USERS TO READ, IF EVER????
CC: bh;me;jmc
∂14-JUL-74 0251 SW,REG
We now have a copy of Grimm's Fairy Tales - GRIMM[ACT,REG] gobbled from the 158.
It needs some editing.
I don't know how widespread this is on the 370, but that file has 80 byte records,
61 records per track. This allows the storage of 4880 bytes/track. The max
capacity of the track is 13100 bytes - so their wasting nearly two thirds
of the (max) available space due to a bad choice of record size.
Our system by comparison, stores about 10000 bytes/track. The lost space
goes into interrecord gaps. Also, the 370 users store card images, i.e.,
records are padded with spaces. All this means that on the 370 it takes
396 tracks to store this file, whereas here it takes 128 tracks. And, by
the way, they're crying to SCIP for more disk space....
∂12-JUL-74 1037 network site ISI
Date: 12 JUL 1974 1036-PDT
From: LICKLIDER at USC-ISI
Subject: PDP-10 For Sale?
To: JMC at SU-AI, McCarthy
cc: Fields, Russell, Licklider
Rumor says Stanford has a PDP-10 for sale. IPTO contractors
are short of PDP-10 resources and IPTO, though short of funds, is
therefore interested in PDP-10s. Is there any substance to the rumor?
Regards
Lick
-------
∂12-JUL-74 0727 network site CCA
Date: 12 JUL 1974 1027-EDT
From: DALE at CCA
Subject: UNICON AVAILABILITY
cc: TOM
- - - -
TOM MARILL IS OUT OF TOWN, SO I WILL TRY TO ANSWER YOUR QUESTION
AS BEST I CAN.
WE HOPE TO HAVE ACCESS TO THE UNICON IN MID-SEPTEMBER. IF IT
WORKS OKAY, THEN IT SHOULD NOT TAKE LONG TO GET IT INTEGRATED
INTO THE DATACOMPUTER SYSTEM.
THERE WILL ALSO BE A MASS STORE AT CCA IN THE FIRST
QUARTER OF 1975.
AS FOR MAKING THESE TWO MASS STORES AVAILABLE TO DATACOMPUTER
USERS, THAT IS AN ADMINISTRATIVE PROBLEM THAT I DON'T KNOW
MUCH ABOUT. I DON'T KNOW, FOR EXAMPLE, HOW MUCH OF THE
UNICON HAS ALREADY BEEN ALLOCATED BY ARPA OR AMES.
I'M SURE THAT TOM WILL BE ABLE TO TELL YOU MORE WHEN HE
RETURNS AT THE END OF THE MONTH.
DALE STERN
-------
∂12-JUL-74 0655 network site ISI
Date: 12 JUL 1974 0655-PDT
From: LICKLIDER at USC-ISI
Subject: Meeting of July 23rd
To: Feigenbaum, Fredkin.MAC at MULTICS, JMC at SU-AI,
To: Minsky at MIT-AI, A310AN02 at CMU-10A, Uncapher
cc: Licklider
Time and place of the July-24 meeting were
indicated in a message you should have received yesterday or the
day before. Please let me know if you did not get it.
July 23, which is intended to be a discussion of longer-term
issues among this (addressee list) smaller number of us,
I'd like us to meet at my apartment. It is in a building called
'Prospect House' about half a mile from ARPA (the Architect Building
at 1400 Wilson Blvd. in Rosslyn). The mailing address is Apt. 850,
1200 N. Nash St., Arlington, Va. 22209, but the entrance is actually
on Oak St---1200 N. Oak St., but I think it has no number on the
door because it is the only entrance in that block on the east side
of Oak. (All that for the info of a taxi driver.) To get to Prospect
House from the airport, go up the river on the parkway that runs
along the river, do not cross any bridge to Washington, do not
take the fork to the George Washington Parkway that offers itself
a half-mile after you have gone under the 14th St. bridge, do
take the left-hand fork (which leads to the Memorial Bridge), and
just before getting to the Memorial Bridge take the road off to
the right that runs down under the bridge right next to the George
Washington Parkway. Then, as soon as you get back to the main
road after going under the Memorial Bridge, get onto the middle
lane and avoid the fork to the right that goes to the George
Washington Parkway. (I am explicit about this because it is very bad
to get onto the GWP. You don't go quite to Maryland before you can
get off, but almost.) Having avoided the GWP, you are getting close.
Take the first off-ramp, which is marked 'Rosslyn'. Turn
left at the end of the ramp---thruough an aperture in the center
strip and onto the far side of the road, which immediately crosses
over an overpass. Then take the second right, which is 14th. St.
Go to the end of it (2 blocks, up steep hill) and turn left on Oak.
Prospect House is on your left, then. If the guy who designed the
road system in this area ever applies to you for a job, hire him.
Program flow clontrol would be a snap for him.
I think it will suffice if we start at lunch,
say at 12:00. We can eat in the restaurant
on the top of Prospect House, move to apt. 850 for afternoon talk,
have coctails and eat again at the restaurant,
and finish up in the evening in the arartment. Keith will be
available only for the evening discussion, I am sorry to say.
In this meeting, I want to go over several issues with
you, issues pertaing to the nature of DoD support for computer
science and artificial intelligence, ARPA and its image of itself,
long-term funding issues (government-wide, country-wide, world-
wide), possible new programs in CS and AI, what the trends are
in the computer industry, prospective PMs
and IPTO directors in the future, and so on. I do not intend,
of course, to try to monopolize the meeting. It will be shaped
mainly by what you consider important. Part of my motivation has to
do with the fact that I have had to stay pretty close to the
grindstone these first six months and feel that I ahve not communicated
sufficiently with the key people in the community. I hope this
discussion will be a start on the road back to strong interaction
between IPTO and the contractor circle.
Regards
Lick
-------
∂10-JUL-74 0711 network site ISI
Date: 10 JUL 1974 0711-PDT
From: LICKLIDER at USC-ISI
Subject: Meeting of July 24, P.S.
To: FEIGENBAUM at ISI, HART at SRI-AI, MCCARTHY at ISI,
To: JMC at SU-AI, A310AN02 at CMU-10A, PHW at MIT-AI,
To: Amarel at BBN, Cheatham at ISI, Fredkin.MAC at MIT-MULTICS,
To: Srin at ISI, Uncapher at ISI
cc: Fields, Licklider
In previous message, please read 'your respective programs'
for 'our respective programs'.
It would be good for me to have, from each of you, and
indication of who will be attending from your project. For
the presentation of each project's goal-objective
plan, I have in mind 15 minutes, including questions of clarification
but not including any protracted discussion.
Regards
Lick
-------
∂10-JUL-74 0657 network site ISI
Date: 10 JUL 1974 0656-PDT
From: LICKLIDER at USC-ISI
Subject: Meeting of July 24 on AI and AP
To: FEIGENBAUM at ISI, HART at SRI-AI, MCCARTHY at ISI,
To: JMC at SU-AI, A310AN02 at CMU-10A, PHW at MIT-AI,
To: Amarel at BBN, Cheatham at ISI, Fredkin.MAC at MIT-MULTICS,
To: Srin at ISI, Uncapher at ISI
cc: Fields, Licklider
From returns received thus far, I conclude that
the date is generally acceptable and that the meeting is
on. ARPA 8th-floor conference room, Architect Building,
1400 Wilson Blvd., Roslyn, Arlington, Va., 9:30 AM.
It has been suggested, and I like the idea, that
the meeting be opened up to include research group leaders (where
different from PIs) as well as PIs. That would include such
notables as Balzer and London of ISI, Martin and Vezza of MIT, Wolf of
CMU. I want to leave the matter of extending the invitation to
research group leaders to the decision (and, if affirmative, the
implementation) of the PIs, each making the decision in his own
case. [Pat Winston, John McCarthy, Allen Newell, and perhaps others
might consider themselves PIs for both AI and AP for this purpose.]
Now let me try to say what motivates the meeting for me. It
is a complex of things. I might start by saying why I did not ask
for a meeting earlier: It seemed to me that there had been several
meetings of AI people and several of AP people, most of them not very effective for one reason or
another, and that most of you were, or could well be, somewhat
put out or put off by an on-again-off-again quality, or a sporadic
franticness, or an ad hockery, in ARPA's efforts to 'organize'
the two program areas. I have been deliberately trying to
cool it and to let the dust settle before taking further initiative.
The basic motivation behind the meeting is that IPTO is
in a poor position, vis-a-vis ARPA Hq., in not having written and
coherent plans for the two program areas, AI and AP, and that I
want to correct that defect. What I see as being needed is an
internal ARPA planning exercise for each of the areas, not a big
deal in which all the researchers convene for two months
and write a planning book. On the other hand, it doesn't make much sense
for IPTO to construct a plan out of whole cloth. The plan has to
reflect the goals, objectives, time scales, and so on of the fields.
So one of the things I want us to do in the meeting is to talk
about goals, objectives, time scales, and the like, using specific
examples as well as general definitions, and get ourselves
to a point at which we are communicating effectively in the planning
area that has become so important in ARPA in recent years. I
do not suppose that, at the end of the meeting, I'll have all I
need to put together the two plans I need, but I expect to be
partly on the way, and in good communication with each of you.
In the process of getting into the area just discussed, I expect to
give you the picture of ARPA that I have built up in my six months
here.
The second factor in the motivation of the meeting (for me)
is more immediately pertinent to AP than to AI, but it is likely
to be pertinent to AI in the future. It has its roots in ARPA's
interaction with Congress. One outcome of last spring's interaction
was guidance from one committee to the effect that, when we next
present the AP program, we should do so with a listing of definite
accomplishments. I want to discuss with you what kinds of accomplish-
ments are judged appropriate for presentation to a congressional
committee (or staff thereof), what kinds of accomplishments are
feasible on the short time scale, how long or short the time scale
actually is, etc. In this connection, I do not want to disrupt the
ongoing research that is not likely, in any event, to produce
'bullets', but I shall try to convey my assessment that it is in our
interest to redirect some parts of the work if they would produce
definite achievements upon redirection and not otherwise.
The preparation that I would like most from you is your
consideration and formulation of the goal-objective structures of
our respective programs in AI and in AP. I'd like to devote
a considerable part of the meeting to your presentations
of those structures, with the objectives set on a time scale. I
fully understand, or think I do, that such plans can rarely be
expected to proceed for long without being changed, and I have not
thought of holding anyone to a stated plan if changes will
improve the expected result. I just want you to realize that I am
in a position in which one always has to have a plan, and to be without
one is to live dangerously, to be with one that does not approximately
match what is actually going on is to invite disaster. At the same
time, I want you to know [it is true] that IPTO really wants to get
great research and exploratory development done without unnecessarily
bugging the guys who do it.
That has to be it for now. I am late for work. Maybe there
will be more from me, maybe not. Please send message
or telephone if there are questions or problems.
Incidentally, Craig Fields is going to help
me in AI and AP work, but I am going to be the ARPA PM for the two
fields for some time.
Regards
Lick
-------
∂9-JUL-74 1429 network site SRI
Date: 9 JUL 1974 1423-PDT
From: FIKES at SRI-AI
Subject: SIGART Statement
To: jmc at SU-AI
Statement received and being included.
Rich
-------
∂08-JUL-74 0750 1,MSW
WILL CALL YOU AT 4PM
∂8-JUL-74 0739 network site ISI
Date: 8 JUL 1974 0737-PDT
From: LICKLIDER at USC-ISI
Subject: Meeting Place
To: JMC at SU-AI
cc: Licklider
Please forgive me for not responding to or at least
acknowledging your suggestion of the meeting place at Lahe Mohonk.
It looks like a beautiful place, and Louise took an instant liking
to it (on the basis of the literature you sent, for which thanks).
Because I am having to be away too much the next few weeks, I felt
it necessary to have the meetings of the 23rd and 24th here in
Washington. My plan is to have the smaller one (23rd) in my
apartment, which is only a few blocks from ARPA, has a view,
and is near a restaurant, and the larger in the ARPA Conference Room.
As i promised, I'll get out a note on purposes, issues, etc. All
responses reeived to date are favorable to the proposed dates.
Regards
Lick
-------
∂05-JUL-74 0906 I,REF
I have a minor stylist question about writing programs in LISP. In 206, the
esthetic beauty of the purely recursive function was stressed. In practical
programming, however, this meets up against the barrier of finite core, finite
push-down stacks, and finite time. In other words, a lot of things are done
quicker and in less space with iteration. I've been writing this program as
sort of a combination of the two. Can you give me any idea as to the validity
of each approach?
I've got a working (fingers crossed, of course) routine that picks out a move
from the board, and also incorporates a back-up structure. I'll probably begin
testing data tomorrow.
Looking ahead, I've been considering the problem of documentation. Is it necessary
to go into detail as to the behavior of each function, or can the explanation be
more general?
Thanks for your help. The functions are in my area J,REF if you want to look at
them.
Bob
∂4-JUL-74 0259 network site ISI
Date: 4 JUL 1974 0258-PDT
From: LICKLIDER at USC-ISI
Subject: Meeting of PIs of ARPA/IPTO-Supported AI and AP Projects
To: FEIGENBAUM at ISI, HART at SRI-AI, MCCARTHY at ISI,
To: JMC at SU-AI, A310AN02 at CMU-10A, PHW at MIT-AI,
To: Amarel at BBN, Cheatham at ISI, Fredkin.MAC at MIT-MULTICS,
To: Uncapher at ISI
cc: Licklider
If you have a long memory, you will recall that I have
been trying to arrange a meeting of the PIs of AI and AP. The
purpose is to examine the trends and issues in the two fields and
ARPA's support of them and, particularly, to chart areas in which
it may be possible to demonstrate visible achievements of clear
value to DoD. In AP, as I have indicated, IPTO has a definite
need for several such achievements for inclusion in reports
to Congress next spring---which translates into 'December' in IPTO.
In AI, there is not quite the same immediate pressure, but it will
be desirable to have the achievement/accomplishment situation well
in hand. Beyond this specific IPTO need, which I want to get
squared away this month and then reviewed in early November, there
are several general issues that we should discuss.
On the basis of telephone communications with some but not
all of you, it appears that July 24 is the best date. I'd like very
much to have the meeting then, and to have it here in
Washington, in the ARPA 8th-floor conference room. Please let me
know ASAP whether you can make the meeting at that time and plance.
To avoid confusion, let me explain that I am asking for
another, separate meeting of a small group that partly overlaps
this one, and that the date of the other meeting is July 23.
I do not want the juxtaposition of the two meetings to cause
a problem about which is which or when. The AI-AP meeting
is to be the 24th unless that date is impossible for some of you.
Now a very brief note about the outcome of the 'Apportionment'
exercise. Both AI and AP came through unscathed, I
am happy to report, and the 'National Software Works' (Software
Production Technology) got some additional funds that it needed.
As the exercise was winding down, however, ARPA was hit by an overall
deferment of a rather large amount of money, and a big chunk of that
devolved upon IPTO's program element called 'Advanced Command-Control-
Communication Technology' or 'ACCCT'. 'Deferment' means that
the funds may or may not come back in the fall. Informed guesses
about the probability of their return vary so greatly that I shall not
pass on a figure. Anyway, that event does not impinge directly
upon AI and AP.
Except for the ACCCT problem and for the fact that the
exercise ran on for several weeks instead of taking one afternoon
as I had expected, I am happy with the outcome of Apportionment.
In ACCCT, however, I am (we are) feeling bruised and bloody, and
if management by exception is the procedure of choice ACCCT will be
my exception for a couple of weeks.
I'll try to get a note off to you shortly suggesting
some topics to think about in preparation for the meeting. Please
let me know by net message (Licklider@ISI or Licklider@BBN) or
telephone (202 OXford-4-5921) whether or not you can make the meeting
on the 24th of July.
Have a good Fourth.
Regards
Lick
-------
∂04-JUL-74 0553 1,YAW
thanks for the last three years john.bye.
∂03-JUL-74 1926 MA,JAM
you mean loses a char on echo or on xmit?? (on one line in 10)...tag
∂03-JUL-74 1508 1,PJH
John, Ifinally manaaged to get on the net. the file META>AXMS (for> read .)
has some initial comments and a few axioms on it . My typing is very bad,
and the syntax changes rather often, but this will hopefully improve.As you
see, i have upper&lower case but not all your lovely set theory symbols,
unfortunately,. Thanks for writing to Keith, he is now being very cooperative.
Pat@Essaex(!)
∂03-JUL-74 1507 1,PJH
John, I finally managed to get on the net. There are some initial
∂03-JUL-74 0933 1,JAF
Your explanation makes sense to me, I withdraw my objections. Have
you explained to LOU/TOB . jerry
∂3-JUL-74 0331 network site ISI
Date: 3 JUL 1974 0330-PDT
From: LICKLIDER at USC-ISI
Subject: Meeting Date, Vision Application
To: JMC at SU-AI, McCarthy at ISI
cc: Licklider
Have focussed on July 23-24 for pair of meetings, 23rd
for discussion of longterm issues with you, Minsky, Fredkin, Newell,
Feigenbaum, and 24th for meeting of PIs of AI and AP. 23rd okay
for first group, 24th okay for most of second group and practically
definite now.
Have asked Kent Kresa, Dir of ARPA office
concerned with remotely piloted vehicles, to suggest dates for
briefing on vision and automatically controlled planes.
I did receive both your messages.
The SUMEX machine is coming onto the network. There has
been a suggestion that the SU IMP be moved to another location at
SU. I do not want to incur additional and unnecessary expenses,
but I want to leave optimization up to you, Feigenbaum, Cerf, et
al---i.e., to the people who are near the problem and understand
it. Please let me know if there is any part of the matter IPTO
should involve itself in.
Regards
Lick
-------