trb (11/18/82)
I hope to end the discussion about inconsiderate irresponsible people who send anonymous mail and netnews, before it starts (again). I have been the butt of such baloney. Here is my learned advice: Do not post netnews commenting on personal messages, just read them if you like, delete them, and forget about them. If you are crazy enough to try to persue the villains, then do so, but don't post netnews about it. Attention-seeking anonymous people thrive on having innocents scurrying about in circles. Deprive them of this pleasure and they have little reason to continue doing their dastardly deeds. Enough said. Andy Tannenbaum Bell Labs Whippany, NJ (201) 386-6491
tjt (04/28/83)
Just wanted to add my 2 cents to the recent praise of PLATO:
I spent 5.5 years at U of Illinois enjoying the experience of PLATO,
and although UNIX is by far the better software development system, there is
almost no comparison between the user interface of PLATO and almost anything
else I have seen in 5 years at Bell Labs. Every time I think of it, I wonder
why the PLATO system never caught on elsewhere. Then again, the need for
a massive CDC machine with several PPU's (Peripheral Processing Units) might
present an obstacle for most potential users. I would be IMMENSELY interested
in any information regarding the current situation and future fate of PLATO.
Regarding the notesfiles system, the biggest difference on PLATO
was that you had to remember the name of each and every notesfile you wanted
to peruse, although I'm sure that's long since been fixed. HOWEVER, being
fairly tolerant, I have not been particularly put off by the obvious
deficiencies in the 'readnews' interface. I instead revel in the much
greater (and more interesting) community of people who interact with
USENET (although I always call it netnews). I'm suprised I haven't
heard of a link between PLATO notesfiles and USENET. Is anything like
that in the works?
USENET is communication personified. (gee, that might be a nice slogan)
===Tim Thompson===BTL/Holmdel===hocda!tjt===grunwald (04/29/83)
#R:hocda:-25400:uiucdcs:10600102:000:3944
uiucdcs!grunwald Apr 28 18:27:00 1983
No, You still need to know the name of any notesfile that you want to read
on PLATO. There is no way to get a listing of all the files unless you are
a systems administrator or an "account director" (a person who hands out
file space -- you can't just go around creating files on PLATO. Weird)
And, No, there has been no link between USENET and PLATO. I've suggested this
a few times, and Ray Essick is willing to put in the PLATO Notesfile to
UNIX Notesfile conversion routines on the UNIX end of things, but there has
been no interest in this on the PLATO end. At least not enough to get any-
thing done.
As for the desirablility of PLATO. Well. Let us say that PLATO has some of
the absolutely worst constructs in it from a C.S. point of view. What I
would hope for is that the "spirit of PLATO" continues. By this, I mean the
user interface presented and the greatly enhanced input parsing.
For people who never used the PLATO system, this might need some
clarification. PLATO uses "arrows" for input. When you run across an "arrow"
command in a program, a little arrow (similar to ">>") pops up. You enter
your input. At this point, it gets complicated, but in short, the programmer
specifies allowed inputs. If your answer is not one of the allowed inputs,
it tells you "no". If your answer is a mispelling of an answer, then it
"marks up" your answer so you can correct it. It is also possible to say
"accept misspellings of answers as if they were spelled correctly."
Additionally, it was possible to specifiy that certain words were synonyms,
that some could be ignored, and so on.
All of this is wonderful for Computer Assisted Instruction (CAI), and it's
nice that these interfaces exist. However, PLATO limits you in so many other
ways, that I would be more for taking the good things from it and putting
them on UNIX (e.g. the PLATO arrow/answer concept).
I have heard that other people have written subroutines which would give
you an estimated guess as to the "closeness" in the spelling of two words,
but I've never seen on on UNIX. The one that PLATO uses was based on compacted
information about each word in a responding phrase (number of vowels, etc etc).
As for the future of PLATO --
Control Data signed a contract with the University of Illinois sometime ago,
and they market a version of PLATO for their Cyber 170 series of machines.
They have many CDC operated systems where they rent resources to users, and
it is also possible for users to buy the software themselves (this is done
by the University of Delaware and the U. of Florida). All told, I think that
there are about 20 PLATO systems in the world, including South Korea, South
Africa, Belgium, England, the Netherlands and Australia.
CDC PLATO is rapidly diverging from UIUC PLATO, and they are planning on
generalizing the structure to enable porting it to other CDC machines.
(hope this isn't too long winded for you all)
There is also a product called "Micro-Plato", which is marketed by CDC,
although the beginning work was done at UIUC. Basically, they use an
intelligent PLATO terminal (512x512 graphics display) with a z80 inside
to run a varient of the PLATO programming language, TUTOR.
More recently, work is being done on a "Cluster PLATO" system, wherein
you hook these "micro-plato workstations" to a motorola 68000 system through
a network, thus saving the cost of a disk drive for every terminal, while
also allowing developement work to be done on the 68000 system. There is a
similar project underway at Control Data. I stopped working for them at
this point, and haven't been keeping track of what's happened in the
last year.
There's all sorts of other things happening, but I think that I signed a
non-disclosure agreement with them a while ago, so I'm going to play it safe
and not tell you (sorry).
Dirk ("weaned on PLATO") Grunwald
usenet: pur-ee ! uiucdcs ! grunwald
PLATO: grunwald/sinspect/cerl