Varisystems, Compugraphic (William Maddox)
infomagic
infomagic at localisp.com
Tue Jul 26 10:41:22 CDT 2005
Here's a little background on Varisystems. I'll try to dig up some more info on the hardware, but it will take some time (weeks...).
I started working for Varisystems on August 9, 1979, just a few months before they moved from Hauppague, NY (Long Island) to Bohemia, NY (also Long Island). Varisystems was founded by Bob and Owen Shevlin. IIRC, Bob was the computer guy and Owen was the business guy (President).
Varisystems did its own design, purchasing, manufacturing, QA, and software. Bob Shevlin had developed a proprietary bit-sliced CPU, made of discrete TTL logic. It had its own bootstrap loader, micro-code, and assembly language, for which Bob had written an assembler. He had also created his own character generator, and all software tools were created in numerous languages, naturally with all accented caracters available.
Varisystems also developed exceptional editing and composition software to run on this home-made hardware. TheVarisystems
units were able to drive the optical output machines from various manufacturers, including Mergenthaler, Bobst, and others. They had a good business going.
At about the time I was hired, the Shevlin brothers were in the process of selling Varisystems to Jacuqes Bobst & Fils, a prestigious Swiss phototype company. Bobst's optical typesetters were of the highest quality and expensive, but their editing systems were far less capable. In the acquisition, Bobst got a lot of good front-end design and equipment, and the Shevlins got a lot of money. The rest of us got shafted - as the last of the projects were shipped off to Switzerland, we were all laid off. That was late in 1980, or early 1981.
I had been hired as a Tech Writer to develop the user manual for the "1200" editing terminal, the first standalone editing station they produced. The first version went out with the first production machines in October 79. A second, moreprofessional version, was
issued the following spring. I also participated on the design team, providing input on "human factors" for various functions and software messages. The "1200" was intended to offload editing work/time from the expensive "production" machines that drove the output devices. Jobs were transferred via 8" floppies. The "1200" unit was well received in the industry.
While I was there, Gavon Balharry (a consultant from Australia) developed the Arabic version of the editing software. It let you choose whether English or Arabic was the "primary" mode, and when you "inserted" text, you could insert either the primary or secondary language. The inserted language always followed its own rules. So if English was primary and you inserted Arabic, it inserted at that point from right to left, and if necessary word wrapped to the next line, also according to the language! All kashidas weredetermined on the fly and as the
next character was typed, the kashidas would change to their appropriate next form. I worked closely with Gavon, doing QA testing as he cranked out new compiles... Quite amazing - I think he was truly the only Genius I have ever met.
Anyway as I try to recall the hardware, I only remember 8-bit instructions, nothing 16-bit. In fact, the big microprocessor threat on the horizon was the Z80.
I'll be cleaning out the garage later this summer, so I'll keep an eye open for any leftover assembly listings... Wish me luck!
-John
More information about the cctalk
mailing list