Talk:PLATO (computer system)/Archive 1

Other PLATO Games
I'm a PLATO user from the 1977-1980 era and was wondering if we could add some of my personal favorites of the PLATO games to the list. The ones I remember the best are Omegaland, Galaxy, Adventure, and RobotWar. Many of us who were children as users ended up in computer science because of our early experiences with PLATO. -- artdeco


 * I had access to PLATO back in 1981-82 for a few months. I played Moria, Labyrinth, Flatbush. I also played another game that I cannot remember the title. You could play as Federation, Klingon, Romulan or Orion, and you went about through space trying to destroy enemy ships.--RLent 21:15, 14 June 2006 (UTC)


 * Sounds like Empire (which I never played). You can play it again on cyber1.org !  &mdash;Tamfang 21:59, 14 June 2006 (UTC)

There were also a number of excellent games teaching basic math skills. How The West Was One, Make-A-Monster, and Pizza Factory stand out in my mind. I believe Sharon Dougdale (sp?) was one of the major writers for those, although it was a long time ago and I could be mistaken. We had PLATO terminals in my grade school, South Side Elementary, and it was a great way to drill without getting too boring. If we can get more concrete information, I think these deserve a mention. Can't find any of them on cyber1 so far, though. Critterkeeper (talk) 02:53, 21 August 2008 (UTC)

Apronym vs backronym?
PLATO is a backronym, not an apronym. The words were invented to fit the name they had already come up with.

Last section references "interested educators ported the engine"; if you're referring to the various variants of the Tutor language, that isn't exactly porting "the engine". There were various offshoots, such as GIST's USE (?), Authorware, and CMU's version of Tutor (written by Bruce Sherwood and Dave Andersen, who had been system programmers at CERL PLATO), but PLATO itself is written in CDC assembly language - porting it has always been judged to be too difficult, except to run it on an emulator.

The only real "ports" of PLATO are NovaNET (running on a Cyber emulator on a DEC Alpha), an internal product never released, internally called NovaNET-in-a-Can, and the CYBIS system cloned from the FAA system running on a (different) Cyber emulator at cyber1.org.

Sorry, I'm sort of new to wikipedia, not sure if this is the right way to bring up issues like this. I suppose I should create an account. Contact me at sep@shout.net if you want.


 * "PLATO is a backronym, not an apronym...'" is exactly what I thought, until I looked up apronym which states that an apronym can be either an acronym or'' a backronym, as long as the meaning of the word is similar to the meaning of the phrase that forms it. So yes, PLATO is a backronym, because the phrase was invented later, but it's also an apronym. It's also a backronym for "Please Leave All Terminals On" because turning the plasma displays on and off every day wore them out more than leaving them on all night. At least at Iowa State, policy was to never turn them off, and there was a sign above all terminals to this effect, using the backronym phrase. Aumakua 04:06, 9 January 2006 (UTC)

Better information: PLATO is an acronym. Here's the authoritative quote, from Prof. Donald Bitzer: There is often confusion about where the name came from. During the first week of starting the project a few of us met to discuss a name. I came up with the name PLATO Standing for Programed Logic for Automatic Teaching Operations. We worked it both ways. Fixing the name and the words to match. One word we changed was Automated to Automatic since automated sounded too much like self mated. That was Peter Braunfield's suggestion and we used it. (from email from Prof. Bitzer dated 2008-02-21, quoted by permission. I also have an old (mid 1960s, undated) U of Illinois document that supports this.

Yes, the "backronym" claim shows up in assorted places, including -- somewhat surprisingly -- a University of Illinois physics department web page. But it is clear from Prof. Bitzer's comments that these claims are incorrect. I have adjusted the intro text and the article title accordingly.

If someone can find a copy of "D. Bitzer, P. Braunfeld, W. Lichtenberger, PLATO: an automatic teaching device, I.R.E. Transactions on Education, December 1961", that would be very interesting. It's the first published paper on PLATO.

Paul Koning (talk) 20:54, 22 February 2008 (UTC)

Not 6600
I didn't know about the "donated" part so I can't comment on that, but the machine at the University of Illinois in 1975 definitely was not a 6600. It was a Cyber 73, a substantially less powerful machine. It's a dual CPU but single functional unit machine, unlike the multiple unit 6600. Somewhat later (as is mentioned a bit further down in the article) a second machine was added, that one a CDC 6500 -- basically the same thing as the Cyber 73. The 6500/Cyber 73 is a dual 6400.

The code inside PLATO actually shows this; it was not at all optimized for a 6600.

Fascinating Article
This is a very well-done article on an intriguing topic. I need to learn more about computer history. Superm401 | Talk 16:35, Jun 2, 2005 (UTC)

plasma display
I'm puzzled by the statement that the plasma display "incorporated both memory and bitmapped graphics". What memory? --Anton Sherwood 01:38, 9 January 2006 (UTC)

It's weak, but true. The display acts as its own screen memory, its electricaly write-only and optically read-only. The only competing no-memory graphical display tech. of the time was the Tectronix storage tube. --Pdpruyne 09:02, 11 January 2006 (UTC)

I believe in this case if you take "The Display" to be the Plato terminal. The terminal actually had a processor that could be programmed (Mickey Mouse clock) and memory for a lineset and charset.Spandox 20:33, 12 April 2007 (UTC)


 * Partly true for the CDC terminals, and the PLATO V (PPT) terminal. Not true for the original "Magnavox" terminal.  The Magnavox uses hardwired logic, while the others use a microprocessor.  Also, CDC did not use the plasma panel but instead used plain old CRTs, with refresh memory as in other video terminals.  In any case, linesets (vector-drawn scaleable characters) are never done by the terminal; they are handled by the central system and sent to the terminal as a sequence of line drawing commands. Paul Koning 21:35, 12 April 2007 (UTC)

Vector or raster?
The original documents I used to make this article stated the terminals used a 512x512 display, which I also found several other documents referring to, the Atari one for instance. However the article now states that it used a vector display instead. I have edited it, perhaps incorrectly, to suggest that the two were the same thing.

Does anyone know the real story here? Did PLATO II/III use vector and PLATO IV raster?

Maury 21:57, 9 February 2006 (UTC)


 * Yes, if I understand the terms correctly. PLATO III used a CRT.  &mdash;Tamfang 22:11, 9 February 2006 (UTC)

PLATO III was refreshed directly out of the CDC 1604's core memory (and hence was raster) via coaxial cable which could be run over long distances. Washington Elementary School had some PLATO III terminals.


 * So you are saying there was a change between the PLATO II and PLATO III? Maury 12:52, 10 February 2006 (UTC)


 * I never saw PLATO II, but I assume it also used CRTs or something similar.
 * I dimly remember hearing that P1 supported one user, P2 two, and P3 many. &mdash;Tamfang 20:55, 10 February 2006 (UTC)


 * At age 10 or so, I once used a hacked P3 terminal to watch tv on a Sunday evening while stuck at CERL with dad. &mdash;Tamfang 20:55, 10 February 2006 (UTC)

PLATO used a Raster plasma display. But it used "vector graphics" in much the same way we do now. Vector graphics (A lineset in PLATO speak) was a scalable graphical entity defined by the connections of lines. Compare to a raster graphic (charset) which is your standard bitmap. Both linesets and charsets could be loaded into the terminal and displayed locally. Linesets could be scaled and rotated. Spandox 20:24, 12 April 2007 (UTC)


 * I would not say "raster graphics" when talking about plasma displays. "Raster" implies a TV-like line scanning display.  That term fits for PLATO 1 through 3, and for the CDC PLATO 4 terminals.  For the PLATO 4 plasma panel based terminals (Magnavox and PPT) I would say that they are bitmap graphics displays but not raster-scanned.


 * In all cases, PLATO 4 has a 512x512 bitmap display, with text being handled by bitmap fonts, and graphics either by bitmap manipulation or by drawing lines as bitmapped vectors. It's not a "vector display" in the classic sense of a display that draws continuous lines on a CRT. Paul Koning 21:35, 12 April 2007 (UTC)

In the plasma display each pixel was self sustaining. If it was off, it stayed off. If it was on, it stayed on. Each pixel could be individually addressed. In fact, as a joke, one of the PLATO system programmers sent random dots out over the system to all the terminals so that individual pixels would flash on and then immediately flash off before anyone had a chance to look directly at them and tell that they were really there.

There were a number of commands in the tutor language that knew how to display lines, circles and other basic shapes and fonts. They could be manipulated and complemented with all permutations of Boolean intersection. Data for one terminal was distributed efficiently over the 1200 baud channel. Typically there were four terminals to a voice grade line with special conditioning. Position and diameter was sent for a circle, and just the end points for a line. In fact, a series of line segments could be sent just by sending a series of vertices.

There is quite bit of interesting history about the graphics because it was nothing like ASCII. In fact it was transmitted over the network as audio sound. You could record the data with a tape recorder and if you played the recorder into a terminal properly, it would display the graphics.

Also text notes could be animated because the characters in a text line were displayed directly to the screen space rather than a within a text line structure. Two characters could be displayed in any Boolean on/off mode, overlapped and even manipulated by special characters which could change the Boolean write, erase, xor, overwrite, and other modes. The text line could be made to wonder up the screen and then proceed backwards as each additional character was displayed. Fonts could also be changed in position one pixel at a time up, down, left, or right. This led to bizarre and wonderful character oriented graphics based on the Boolean overlapping of characters. Font sized funny faces were popular, composed out of a series of characters, or characters could move across the screen as a text line was displayed as they were erased and rewritten by Boolean overlapping of a sequence various characters.

There is a lot more that could be talked about in just the graphics.

HonestGent (talk) 09:38, 14 September 2008 (UTC)  (John Sellers, formerly with the PLATO login, “sellers of Arizona”)

OK, several comments.

First, I have used three different Plato displays. One was a nice, orange plasma terminal, which has 512x512 raster graphics. I also used a black/white CRT that had 512x512 raster graphics. Finally, I also used the console of the Cyber computer which was a *vector* CRT, with 512x512 resolution. So, the answer is that plato was both raster *and* vector.

Secondly, the above post talked about the graphics being transmitted as audio sound. This is just confusion, all modems (modulator-demodulators) take digital information and convert it into sound so that it could be transmitted over the phone network, this is not unique to Plato, nor does it have anything to do with ASCII.


 * But PLATO "rolled its own" communication protocol. You can NOT use an ordinary tape recorder to record and play back the typical half-duplex, full-duplex communication of the time for several reasons having to do with hand-shakes, line conditioning, initializations, et al.  With PLATO, on the other hand, you could disconnect the terminal when the "PLATO is UP" was displaying, record what was coming over the line with an ordinary tape recorder and then hook it to the terminal.  Playing the recording would display whatever would have been displayed had the line been hooked to the terminal, including animations.


 * Key presses from the terminal were routed to the Cyber and submitted individually to the Cyber job queue as real, honest to goodness full fledged, Cyber jobs. End to end round trip from terminal to Cyber and back was a guaranteed 1/4 second echo response time.  That characteristic was critical to PLATO and was not supported by most phone line modem communication protocols of the day. HonestGent (talk) 22:07, 15 September 2008 (UTC)

Finally, on the subject of ASCII, Plato did not use it, it had 6-bit characters, 10 of which could fit in the 60-bit words that it used. This was all inherited form the computers they ran on, e.g. the CDC 6600. Wrs1864 (talk) 12:49, 14 September 2008 (UTC)

So that's what The Roach Organization is...
207.179.172.217 14:07, 26 April 2006 (UTC)

some dates
This was posted to pad (on cyber1.org) two weeks ago by asher/cerl:

Just received my latest copy of the Physics Illinois News last week (sent periodically to U of I physics grads) and they had over 2 full pages devoted to PLATO, including a picture of Dan Alpert at a terminal as well as the PLATO lab in room 220 in Loomis. They also had an interesting timeline from 1960 - 1974 for the PLATO project that was apparently originally printed in 1974. Here is what they had:

1960 June     PLATO I (one terminal) 1961 January  PLATO II (two terminals) 1961 March    First remote terminal (30 miles) 1961 Spring   First teaching attempt using PLATO 1962 Spring   First college credit for students taking a                 course using PLATO 1963 Fall     First stage of PLATO III completed 1964 December First on-line editing possible 1966 March    PLATO III (20 terminals) 1967 Summer   TUTOR author language in use 1968 March    PLATO IV system design started 1968 June     First time-shared authoring and student use 1968 Winter   Four remote demonstrationg centers in                operationg (12 to 14 terminals in each) 1969 Summer   150 hours of instructional material developed by this date 1970 Summer   720 hours of instructional material developed 100,000 student contact hours of use 1971 May      Delivery of first Digivue display from Owens 1971 June     Delivery of first PLATO IV terminal from Magnavox 1972 Summer   1,600 hours of PLATO III instructional material in about 70 courses with 154,000 student contact hours to date 1972 Summer   40 PLATO IV terminals in operation; intensive PLATO IV lessone development; remote on-line demonstrations in Canada, Europe, and USA 1972 Winter   250 PLATO IV terminals in operation at                approximately 40 locations (15 on the UI                campus and about 25 off campus) 1973 June     PLATO III phased out 1973 Fall     Remote on-line demonstrations in Sweden and the Soviet Union; 1,500 hours of available PLATO IV lesson material in 50 teaching areas; 25,000 student contact hours between September and December 31, 1973 1974          400 terminals in operation at approximately 70 locations (20 on the UI campus and about 50               off campus)

Note, in all the articles in this newsletter, there was no mention of Avatar or Empire. Go figure.

Oh, another interesting thing I had not seen mentioned before is that they said the ILLIAC was the system that originally ran PLATO. Back when I was an undergrad lab assistant, I actually ran across some of the original operating manuals and schematics for the ILLIAC buried at the bottom of an old filing cabinet in Loomis Lab. I kind of wish I had a copy of them now...

I found the link for the newsletter to view it online. It can be found at:

http://www.physics.uiuc.edu/alumni/PIN-current.pdf

Look at the article on the bottom of page 1 and the articles on pages 4 and 5.

- Marc


 * I think the proper link you want is this one: http://www.physics.uiuc.edu/alumni/PIN-2006-No1.pdf -- the link you provided just always points to the current issue. SunSw0rd (talk) 19:15, 26 March 2008 (UTC)

More Complete PLATO/NovaNet timeline
University of Illinois Computer-based Education Research Laboratory (CERL) Evaluation Office

PLATO/NovaNET History and Operation Overview

History 1960--PLATO project of Coordinated Science Laboratory (CSL) begins --Computer-Based Education (CBE) system (PLATO I)          first demonstrated 1961--PLATO II - World's first time-shared CBE system is in use --Computer generated graphics used in instructional simulation --First CBE authoring system (EE Logic) 1962--First accredited college course supported by           PLATO (EE) 1964--PLATO first supports on-line editing, multiple classes 1965--Entire accredited college course taught solely by PLATO (Library Science) 1966--PLATO III - Classroom-sized CBE system (20 terminals) 1967--Computer-based Educatation Research Laboratory (CERL) created to continue research and development of CBE. CERL takes over PLATO from CSL --Artificial Intelligence techniques used for response judging --General-purpose authoring language (TUTOR) in use 1971--First commercial version of plasma display (invented at CERL) delivered 1972--PLATO IV - Wide-area-network CBE system in use (250 terminals initially) 1974--Control Data Corporation installs first commercial PLATO IV system 1975--First PLATO microprocessor terminal 1980--PLATO microcomputer classroom used for accredited course 1982--CERL Cluster system in use -- low-cost local CBE system 1987--CERL NovaNET system in use -- low-cost central CBE system 1987--University Communications Incorporated (UCI) begins marketing NovaNET 1988--Portal Software allows use of personal computers for NovaNET 1991--PLATO IV system (1,997 terminals at peak) replaced entirely by NovaNET 1993--NovaNET used for central management of local CBE --University of Illinois begins shut-down of CERL 1994--NovaNET equipment and operation turned over to UCI.

Service 27,269,601 user-service hours delivered on CERL CBE systems from 1961 to 1 July 1993 22,269,319 hours on PLATOs I-IV, 5,000,282 hours on NovaNET)    Over 1 million user-service hours delivered every     year from 1978 to 1993

Terminals 3,390 connected on 1 July 1993 36% at elementary/secondary schools 25% at colleges and universities 19% at community colleges 11% at business and industrial sites 5% at adult education centers 4% at CERL for research and system support <1% at local and federal government sites

Users  (On 1 July 1993) 73,827 student records (71,864 individual,            1,963 multiple-use) 7,680 instructor records 4,007 author records

Materials available to PLATO/NovaNET users (on 1 July 1993) 50,929 files 68% used for system management, data, and communication + 2,787 files are "bulletin board" style note files + 1,228 files are personal mail files for individuals 32% used for instructional material + about 9,371 hours of material of which about 6,107 have accredited classroom use

(Compiled by Allen Avner, 1 July 1993) The above information was the last of a series of information sheets issued by the CERL Evaluation Office during the years 1967 through 1993.

Current usage of PLATO
The Anchorage School District currently uses a version ("PLATO Web Learning Network") to teach basic classes. The software is apparently run off of district web servers and is accessible from any computer once an ActiveX plugin is installed. Stellertony the Bookcrosser 17:42, 2 January 2007 (UTC)

This REALLY needs some pictures
Does anyone out there have some more images we can put in the article? It strikes me as the one obvious problem. Maury 16:23, 3 March 2007 (UTC)

Improvement
This is a really nice article, but it could use some improvement. It lacks many citations. The sections are overly long and could be trimmed. And there should be a legacy section which is usually in articles like this. Just some suggestions. I'm really not expert enough in the subject to do anything so I don't want to sound critical, but given the importance of this system and the amount of info that's here, I think with sufficient clean-up this could be a featured article —Preceding unsigned comment added by 72.82.227.246 (talk) 18:26, August 28, 2007 (UTC)

Rick Blomme
The article is not complete without some mention of Rick Blomme, a programmer of legendary skills, to whom PLATO owes a great debt. I would put something in, but what I know of his contribution is hear-say and may not be accurate to the particular facts. Does anyone have some good references?

I did witness some of his unusual abilities. I once observed him running a PLATO typing drill. He was interactively making corrections to his typing and yet his accumulative speed was STILL 110 WPM. HonestGent (talk) 03:32, 15 September 2008 (UTC)


 * I did a typing test the other day, correcting as I went along. At one point I thought I had missed a word (because the panel in which I was typing had acquired a scroll bar, so that its word-wrapping no longer matched that of what I was copying from) and stopped to look for the error. I only scored 107 that time. —Tamfang (talk) 04:38, 16 September 2008 (UTC)

As I was adding an info box to this page today (I was a PLATO lesson programmer at UD) I read some of the references that were cited. In one, Rick Blomme was mentioned as one of the major initial contributors to the design of the TUTOR language. I put his name in the infobox & cited the reference. Philhower (talk) 20:49, 15 September 2008 (UTC)