N.B. This poor excuse of a home page exists because it is better than nothing and will get some of the stuff here into search engines.
In chronological order:
[1963] The first computer I ever programmed was a process control computer my father designed, Bailey Meter 756. It's notable for:
The CPU executed off a drum. Sort of like a disk, but with words placed around a cylinder. See The Story of Mel.
Instructions and data were in decimal, as the target market and intended users knew noting about binary, octal, or hexadecimal numbers.
Dad said, "It's so easy to program that a 12 year-old can do it." He then proceeded to teach me (I was 12 at the time) to program the data logging processor.
It appears to be the first commercially successful parallel processor. From Dad's patent, 3,266,023: Parallel Program Data System:
The programing simplicity achieved through use of a plurality of independently operating parallel subsystems renders the data handling system particularly valuable to a process industry such as an electric utility company. Each subsystem operates from its own independent program which may be easily modified by a power engineer without the assistance of a computer expert. The program in any of the subsystems may be changed without interference with the other subsystem programs, or any one subsystem may be removed from service completely either intentionally or because of equipment failure without interference or inter-action with the other subsystems.
The system was introduced in 1962. One was installed in 1967 to monitor an Australian power plant and was shutdown in the late 1990s. By then programming was done within a spreadsheet, I guess something like a drum knowledgeable assembler. The eight ton system was given to the Australian Computer Museum Society, the source of the link above.
A tangent - we lived in Ohio then, as did a budding computer scientist named Don Knuth. Years before, his introduction to computing was through the drum based IBM 650. He dedicated his The Art of Computer Programming books to the IBM 650 computer.
[1967] Physics class in Riverside High School
Mr. Spondike played a record (yeah, 33 1/3 rpm LP) put out by Bell Labs. The album was on
their research into having computers synthesizing human voice. A first step was to
synthesize music. One track on the record was an instrumental version of "Daisy" or "A
Bicycle Build for Two." Another track had the computer singing it. I found them
fascinating but was disappointed that very few people would ever hear that early
work.
The next year the movie 2001 - A Space Odyssey was released and HAL sang it as he was being shutdown. I need to dig deeper - I'm sure I heard the Bell Labs version in the theater. ("Oh, they couldn't have. Well, Daisy, okay, but not the Bell Labs version. But Arthur C. Clarke. Auggh - sit still, don't cheer. Don't make a fool of yourself.")
It appears that did get replaced with HAL singing Daisy in his voice. The Bell Labs version lives on elsewhere. THE HISTORY: Daisy's Bell & IBM 704 and Bell Telephone Labs - Daisy Bell '63 are also interesting.
[1968] MERCURY - simulates a satellite (NASA manned Mercury capsule) orbiting the
Earth
This is the first program I ever wrote for fun and
curiosity. I wrote it as my first programming class wound down and I intended to step
it through the Gemini (spacecraft that could change its orbit and rendevous with other
objects) and Apollo (Earth to Moon and back) programs. However, it didn't take long
before I didn't have time to work on it more. Besides, my orbits kept decaying.
[1969] LIGHT - displays light patterns (and ping-pong) on a
KA10 console
This is my first PDP-10 program I remember writing for the first PDP-10 at CMU.
[1970?] I realized I was a computer wizard one day
It was quite a surprise to me, read about it and a disk patching tool that paved the
way.
[1972] DATE75 - a date overflow before Y2K.
It will return in late January 2052.
[1979?] warstory6 - Wow, someone saved my USENET post
about computer produced music
I probably stumbled across this while looking for something else I had done. I was duly
impressed that someone thought one of my posts was worthy of preserving on the web. On it's
own it's worthy to include here, as are the other stories.
[1989] The ETA Saga
ETA was a CDC spinoff that produced a mini-supercomputer that looked like it would be
serious competition to pretty much any of the other vendors' offering. Inexplicably, they
got shutdown just before they were about to become real. We were surprised, relieved, and
even grateful. The folks at ETA couldn't believe it. On the first anniversary, Rob
Peglar, a manager at ETA, posted this account of what happened. It is jaw dropping.
The above link goes to yarchive.net, whatever that is. I made a copy so I can't lose it.
Thoughts on Large Language Models and the Future of
Programming
This draws on my career to answer a coworker's question about the future of programming in
the new era of Large Language Models where you "just" tell an AI what you want a program
to do. The devil will still be in the details.
The Soul of an Old Machine
It's not entirely obvious above that DEC's PDP-10, especially the first model, the KA10,
is my all-time favorite computer. Folks who have only known Windows and Macs cannot know
the fondness some of us oldtimers have toward that system. It executed only some 150k
instructions per second - but they were good instructions! We assembly language
programmers loved rich collecion of medium strength instructions. That's clear enough in
my Tulip pages the DATE75 reference above link to them. The TOPS-10 operating system was
command line oriented, TOPS-20 and Unix came along a later with much better interfaces
(shells), but back in the days of 110 baud Teletype terminals, we couldn't dream of a
mouse, and well, they sure beat card punches.
At a lower level, the console switches made a great binary abacus, the lights often told a story about what the system was doing, the more technical lights above the CPU bays were a huge help in understanding CPU failures. Yes, we fixed CPUs then - they cost $160k and held hundreds of logic cards. I got pretty good at repairing CMU's.
One thing I didn't fully appreciate is that users were also were very fond of that old system until I read The Soul of an Old Machine about a PDP-10 at the end of its life. I still choke up every time I read it.
I would not - could not - go back to the PDP-10 as my everyday system, but it's really sad that as technology and users changed over the decades people can never develop the same fondness we had for those old machines.
This is supposed to be at inwap.com/pdp10. My browser says "404" today (2024 Sep 24), but whois says the domain is current. I saved this in 2016, it's long past time to get this on my web site.
Contact Ric Werme or return to his home page.
Written 2022 Oct 31, last updated 2024 Sep 24.