IBM System7use of Teletype as console 2549

Jay Maynard

I'd bet there were some politics involved, maybe even legalities too, having to do with the N.A. telephone monopolies in the 1950's and 60's, such as what terminal devices were allowed to connect to the lines. I think TELEX had a monopoly of sorts too.

'Inherited' a S-7 once (it came with the job). Never knew much about it except that the paper tape reader had been broken for years. Ran a pretty big network at the time, about 250 terminals spread across 3,000 miles or so (mainframe ran PARS-F). Most of the terminals were 110 baud but there were a few fast ones, something like 300 baud. I could be wrong but I thought the code was a variety of the earlier BCDIC, at any rate it was a 6-bit character set on the lines, not 7 or 8, no room for lower case.

This was around 1980 or so. Even the IBM CE's didn't want much to do with the S-7 let alone the 'console' and the original 'sysgen' had been lost when new management fired the entire tech services department. The Pars-F application required a shut-down to do end-of-day processing, so every morning, the S-7 would be re-booted. This was done from a tape (mag tape not paper tape) cbuttette, of which we had only two. It had been years (like 15 or more years) since anybody there remembered the paper tape being used.

One day one of the cbuttette tapes snapped which made everybody a little nervous. I went to Radio Shack, bought a consumer tape deck and some blank cbuttettes and made a bunch of copies, some to go off-site. They booted fine. The tape drive (I forget the model number) had what looked like a standard consumer plug, jury-rigged to a little wiring harness. Got another plug from Radio Shack and managed to get the $29 tape deck to subsbreastute for the original one. There was lots of other weird old gear at that place, so I never had time to learn how to do a S-7 sysgen. Just told management that we were secure because of the off-site cbuttettes.

I had a few unsolicited job offers over the 'phone from as far away as Europe - various airlines that somebody had pbutted my name to as a PARS expert, which I really wasn't. Still, I liked how lean and mean PARS was and it was easy to follow since all the source code was freely distributed. Our application programmers thought they knew were doing buttembler, but it was only a subset - linkage and IO was all taken care of for them so all they knew was a handful of arithmetic, move and compare instructions. A guy who had been for training in St. Louis told me that a 360-50 there ran PARS for a network of 1,200 terminals!

p