1972-9: Purdue, Plato & PC’s

From 1972-1976 at Purdue I was Assistant Professor of Computer Instruction and Director of Computer-based Education in the School of pharmacy, a rather unique position.  I helped other faculty at Purdue use the University of Illinois PLATO system developed by their Center for Education Research, CERL.  The hardware there was the brainchild of Don Bitzer with a lot of help from Roger Johnson, and the TUTOR programming language came from Paul Tenczar.  This system was way ahead of its time.  When we were using it a CDC6600 mainframe was used with some of its PPU power (peripheral processing units) allocated to running a time multiplexed output channel to thousands of graphics terminals scattered far and wide.  We were in Indiana.  There were terminals coast to coast and even offshore.  At every terminal, every keypress was sent by local wiring or modem all the way to Urbana, processed and then echoed to the screen or used to initiate a graphical presentation.  This was a 512 by 512 plasma graphics station made by Magnavox with a slide projector built in on top, touch sensitive screen, and an audio output option (see the Wikipedia photos).  Courseware written in TUTOR would do  simulations as well as traditional CAI tutorials with questions and answers, slides, whatever.  Stanley Smith and Bruce Sherwood were prolific U of Ill. profs who taught Chemistry and Physics courses as did many other innovative instructors using the system.  See also this synopsis.

At my location we had 8 of these workstations as well as other AV-based equipment.  I advised faculty and trained their graduate students to write courseware to teach medicinal chemistry and physical chemistry and other subjects for pharmacy students in the Purdue School of Pharmacy and Pharmacal Sciences.  My expertise beyond programming was in making designs unambiguous, user friendly, and cognitively tractable.  It was necessary to get faculty to think more in terms of mastery learning and specific modular instructional objectives.  The last thing we wanted was to frustrate students who had never used a computer before.  We had hundreds of sessions per week of self paced individualized instruction, and we ran 2 or more shifts a day to fit in all the students doing this for credit.  The system had an early version of email, called ‘notes’ that allowed any user anywhere to leave messages for others far faster than snail mail.  Our own work with PLATO became widely known and was a showcase example of educational innovation at Purdue, long since forgotten.

PU1

PU2There was a problem trying to get this sort of thing accepted on a wide scale.  Courseware was labor intensive.  It could definitely be amortized and shared, but I never have met a faculty member who did not think they could explain their subject better than the next guy and/or did not fear losing their job in the classroom.  There was also the expense of setting up a big time sharing system on or off campus.  These were expensive systems back then.

Intel announced the 8008 and the 8080 soon after.  BYTE Magazine was launched, and MITS created the Altair computer kit.  Heathkit had theirs.  It was like the modern birth of the Maker movement we see today all happening 40 years ago.  Finally, Digital Equipment Corp. launched the LSI 11 micro line to complement their PDP-11 minicomputers.  I visited Xerox PARC in about 1975 where I saw the Alto workstations running SmallTalk with mouse, graphics, and minicomputer.  I’m pretty sure I got there before either Jobs or Gates.  This is also where they got their inspirations.  Adele Goldberg showed me around.  She invented SmallTalk and launched the whole object oriented programming revolution.  I remember reading her SmallTalk descriptions and being baffled by the new paradigm.  Incidentally, Bob Metcalfe at PARC was responsible for bringing Aloha Net down to earth for ethernet.  What a place to work.

I talked to NSF and Magnavox, and they were supportive of the idea of making what I called a Stand-alone CAI workstation.  But, unfortunately, my School of Pharmacy at Purdue was not interested.  So I chose to return to graduate school and launch into this flat out.

I heard about the UCSD PASCAL project also running on LSI-11s, and I went there in fall of 1976. Surprisingly, the department (Applied Physics and Information Science) was not ready  for this kind of interdisciplinary work, and the PASCAL project had its own agenda such that there was little interest in building a fancy graphics station.  I was determined to learn how to design hardware.  So I got a job nearby at a hardware company called Linkabit.  I learned a little about hardware there, but was mostly programming microcode.  I did get the chance to design a prototype GUI for communication station operators, and I modeled it after PLATO and Alto using a plasma display.  It had drop-down menus, message boxes, and touch input (in 1978!).  However, I was still not learning digital design.  After two years I joined a company doing CAI and designing a 68000 workstation in San Francisco.  They did not last long, and our branch closed even sooner (WICAT-World Institute for Computer Aided Teaching).  So I next went to SLAC where they hired me as an engineer, and I got the chance to finally do some hardware.