Pleasant@Rutgers (11/09/82)
HUMAN-NETS Digest Saturday, 6 Nov 1982 Volume 5 : Issue 100
Today's Topics:
Queries - Knuth's Art of Computer Programming &
Whetstone Benchmark Programs &
Laboratory for Human-Computer Interaction,
Artificial Intelligence - An Apology & Parallelism and AI,
Technology - WorldNet (4 msgs),
Computers and People - Computer Names (2 msgs)
----------------------------------------------------------------------
Date: 2 November 1982 15:35 cst
From: Heiby.APSE at HI-Multics
Subject: Knuth's AoCP
In Knuth's "Art of Computer Programming" volume 1, published by
Addison-Wessley, the preface says that the work is in seven volumes.
I have never seen any volumes other than 1-3. Do the others (any of
them) exist? Are they being written? Did they exist and are now
out of print? Obsolete?? Thanks. Ron.
------------------------------
Date: 4 Nov 1982 1136-PST
From: JBROOKSHIRE at USC-ECLB
Subject: Whetstone Benchmark Programs
In a recent exchange on INFO-PC there was a discussion about the
Whetstone benchmark, and reference to runs which had been made and
the location of the program used (FORTRAN - at USC-ISIB in
<INFO-PC>whetst.for, FTPable). I got the program and have been
doing several local studies with it. NOW - in the October issue of
DIGITAL DESIGN there is a short (almost page) discussion of the
Whetstone SERIES of benchmarks that were developed at the behest of
the British government, and exist in at least two other languages -
ALGOL and BASIC, and perhaps others. My question is: does anyone
know of the availability of these other modules/versions, or can
anyone provide literature references I could follow to get more info
on these tests? Any help greatly appreciated, and I will summarize
results and make them available for those interested.
Jerry Brookshire
------------------------------
Date: 26 Oct 82 19:05:39 EDT (Tue)
From: Mark Weiser <mark.umcp-cs@udel-relay>
Subject: Laboratory for human-computer interaction
We are helping Nasa to equip a laboratory for studying human factors
in human-computer interaction. This means not ergonomics (such as
screen tilt and chair to console distance) but software and display
content aspect, menu vs. commands, high-res bitmapped vs. regular,
color or not, and anything you and we can think of. At the moment
equipment is of interest, since it takes such incredible lead times
to buy equipment for Uncle Sam. My question to you all is:
What other human-computer interaction laboratories are out
there, and how are they equipped?
I think probably the whole community will be interested, so
reply to the bulletin board (with cc to me please).
[Mark has agreed to accept your replies, compile and summarize them,
and send the summary back to Human-Nets and WorkS. Please send your
replies to Mark without cc'ing these lists. Thanks - Mel]
------------------------------
Date: Friday, 22 October 1982 15:56-EDT
From: AGRE at MIT-MC
Subject: HUMAN-NETS Digest V5 #96
By the way, I should make a public apology for forgetting the name
of the head of Columbia's Non-Von project, David Shaw. He and his
project are every bit as important as the others I mentioned. Also,
the discussion was indeed not one of AI rescuing computer
architecture, but rather one of computer architecture rescuing AI.
(Both are of course simplifications.)
- pHil
------------------------------
Date: 28 Oct 1982 1515-EDT
From: David F. Bacon
Subject: Parallelism and AI
Reply-to: Columbia at CMU-20C
Parallel Architectures for Artificial Intelligence at Columbia
While the NON-VON supercomputer is expected to provide significant
performance improvements in other areas as well, one of the
principal goals of the project is the provision of highly efficient
support for large-scale artificial intelligence applications. As
Dr. Martin indicated in his recent message, NON-VON is particularly
well suited to the execution of relational algebraic operations. We
believe, however, that such functions, or operations very much like
them, are central to a wide range of artificial intelligence
applications.
In particular, we are currently developing a parallel version of the
PROLOG language for NON-VON (in addition to parallel versions of
Pascal, LISP and APL). David Shaw, who is directing the NON-VON
project, wrote his Ph.D. thesis at the Stanford A.I. Lab on a
subject related to large-scale parallel AI operations. Many of the
ideas from his dissertation are being exploited in our current work.
The NON-VON machine will be constructed using custom VLSI chips,
connected according to a binary tree-structured topology. NON-VON
will have a very "fine granularity" (that is, a large number of very
small processors). A full-scale NON-VON machine might embody on the
order of 1 million processing elements. A prototype version
incorporating 1000 PE's should be running by next August.
In addition to NON-VON, another machine called DADO is being
developed specifically for AI applications (for example, an optimal
running time algorithm for Production System programs has already
been implemented on a DADO simulator). Professor Sal Stolfo is
principal architect of the DADO machine, and is working in close
collaboration with Professor Shaw. The DADO machine will contain a
smaller number of more powerful processing elements than NON-VON,
and will thus have a a "coarser" granularity. DADO is being
constructed with off-the-shelf Intel 8751 chips; each processor will
have 4K of EPROM and 8K of RAM.
Like NON-VON, the DADO machine will be configured as a binary tree.
Since it is being constructed using "off-the-shelf" components, a
working DADO prototype should be operational at an earlier date than
the first NON-VON machine (a sixteen node prototype should be
operational in three weeks!). While DADO will be of interest in its
own right, it will also be used to simulate the NON-VON machine,
providing a powerful testbed for the investigation of massive
parallelism.
As some people have legitimately pointed out, parallelism doesn't
magically solve all your problems ("we've got 2 million processors,
so who cares about efficiency?"). On the other hand, a lot of AI
problems simply haven't been practical on conventional machines, and
parallel machines should help in this area. Existing problems are
also sped up substantially [ O(N) sort, O(1) search, O(n^2) matrix
multiply ]. As someone already mentioned, vision algorithms seem
particularly well suited to parallelism -- this is being
investigated here at Columbia.
New architectures won't solve all of our problems -- it's painfully
obvious on our current machines that even fast expensive hardware
isn't worth a damn if you haven't got good software to run on it,
but even the best of software is limited by the hardware. Parallel
machines will overcome one of the major limitations of computers.
David Bacon
NON-VON/DADO Research Group
Columbia University
------------------------------
Date: Friday, 22 October 1982 19:22-EDT
From: Vince Fuller <VAF at CMU-20C>
Subject: WORLDNET ('True Names')
There has been some discussion of this book on SF-Lovers recently,
specifically concerning the general (non)availability. Check recent
SF-Lovers issues for details and info about how to obtain the book.
--vaf
P.S. It's worth reading, if you can get a hold of it.
------------------------------
Date: Friday, 22 Oct 1982 22:00-PDT
Subject: Re: Why not AT&T for WorldNet?
From: guyton at RAND-UNIX
Why not AT&T for WorldNet? Sigh . . . where do I begin?
First, this is not the Telecom Digest. We're not talking about how
well AT&T runs our phone system, or if Sprint or MCI are any better.
My comment about "Ma Bill" was unfortunate.
I guess the key point that was missed is that Worldnet must be a
well connected community of smaller networks, where each subnet can
be of a radically different communications technology.
Take a good look at the DOD InterNet community. They've got leased
phone lines for the 'ol Arpanet, satellites for some Satnets, a few
radio-packet nets, and a lot of Ethernets and some others that I
don't know how to describe. All of them up and running and talking
to each other with the same protocols (TCP/IP).
In this context it makes no sense to have any one owner of the
entire network, regardless of how well they run their communications
system.
At the same time, I'm trying to lobby for development of a very
cheap communications medium (ham packet radio is an example). This
is not crucial for WorldNet to survive, only for it to grow quickly.
You made a good point when you said the US telephone system is much
better than those systems of the rest of the world. I agree, most
other phone systems are very bad. Therefore we must not depend upon
the phone systems for WorldNet, or we will be excluding most of the
world.
Enough flaming for tonight,
-- Jim Guyton
------------------------------
Date: 23 October 1982 07:42-EDT
From: Gail Zacharias <GZ at MIT-MC>
Subject: Why not AT&T for WorldNet?
Date: 17 Oct 1982 2104-EDT
From: ZALESKI at RU-GREEN (Michael Zaleski)
...
Phone service in many third world countries is at a level that
Americans would find totally unacceptable.
We must be in pretty bad shape to find comfort in being more
technologically advanced than many third world countries...
As for the general point you are making, one might mention that
Hitler made the trains run on time. I.e. technical competence is
not the point here, power is. Something as important as a
communication network should not be controlled by any single group.
Witness for instance the recent events in Poland, where a revolution
was stopped by turning off the phones.
------------------------------
Date: Saturday, 23 October 1982, 15:42-EDT
From: Robert W. Kerns <RWK at SCRC-TENEX at MIT-MC>
Subject: HUMAN-NETS Digest V5 #97
I just couldn't let these claims go by undisputed.
Date: 17 Oct 1982 2104-EDT
From: ZALESKI at RU-GREEN (Michael Zaleski)
Subject: Why not AT&T for WorldNet?
In a recent message to Human-Nets, one reader expressed a wish
for a "World-Net", to tie all sorts of computers worldwide
together. In this message, the author stated a belief that it
should not be owned by one company and specifically stated that
"Ma Bell" should not be the owner. I must honestly say that I
find this attitude toward the phone company hard to understand.
Compared to ANY other phone system in the world, the U.S. has
THE best.
I have been victimized by ma bell's utter incompetence in the use of
computers so many times I am convinced that letting them have
anything to do with world-net would be the kiss of doom. I'd rather
let IBM do it.
New England Telephone does not even do on-line service-order entry.
They take down, on paper, the information about your service
request. They then mis-transcribe it (when they don't lose it
altogether), fail to add relevant information to the record (such as
the fact that they already called you and asked you something), and
take months to get things straightened out. (Not to mention their
billing software!) I'll spare you the case histories, but in my
experience, a service order *ALWAYS* results in a complete mess, and
wastes about a day and a half of my time.
It may or may not still be true that France is months behind in
installation, but they are actively working on correcting the
problem, and are working with such advances as distributing
terminals instead of phone books. NET has shown no visible sign of
correcting their inadequacies in the many years I've been a
customer . Comparing them with other phone companies is not the
point. Compare them with a non-monopoly, and they are clumsy,
incompetent, inflexible, and uncooperative. Hardly the kind of
company one would want to put in charge of the most exciting new
forms of communications around.
Another minor point:
They do, but my experience with MCI showed me they also provide:
- An extra nuisance at dialing time.
This is because Ma Bell doesn't provide the signalling necessary
to do it any other way.
-- Michael Zaleski, mhtsa!mzal@UCBVAX or "Zaleski@GREEN"@Rutgers
Bell Labs, Murray Hill, NJ
As far as I'm concerned, Bell Labs and the installed physical plant
are the only parts of Ma Bell worth keeping.
------------------------------
Date: 22 October 1982 10:47 cdt
From: heiby at HI-Multics
Subject: Net addresses in ads
I have seen several advertisements in the personal computer field
which give a mail address on Compuserve or The Source. This is, of
course, most prevalent in Compuserve's magazine, but I've seen it
elsewhere, as well. Ron.
------------------------------
Date: 22 Oct 1982 1102-PDT
Subject: Use of network addresses by businesses
From: WMartin at Office-8 (Will Martin)
The only publicized solicitation of net mail that I have seen has
been by columnists in Electronic Engineering Times. There were two
who gave Source and Compuserve mail addresses; the current issue
only seems to contain one -- Phil Koopman, TCP893 on The Source.
Of course, it's hard for ARPANET people to credit either The Source
or Compuserve as being "real" networks...
Will
------------------------------
End of HUMAN-NETS Digest
************************