1403 printers was: IBM's last tabulator last unitrecord punch card machine

John Savard

of basically and

At the university I was at, we had two 1403-N1 printers which were used for all University business and student programming buttignments. One was reserved for students; one for business.

Both printers had uppercase-only GN print trains. (I remember being told that GN had fewer characters than PN and thus was faster - but more than the 48-character set described above.) They also had an OC train which printed letters that OCR readers could recognize; this is where student ID cards and such came from.

During time time period (1977 until about 1981) the only 327x terminals that the University had were uppercase-only, so the lack of lowercase printing support wasn't missed. (Yes, 327x terminals that could display upper and lowe case letters existed, but they cost more and so we didn't have any.)

Around 1981 two things happened - we put a non-IBM dot-matrix ASCII paper terminal in the main student terminal room mostly because we didn't know what else to do with it. It was the only device that could print lower case letters on campus. Enterprising students immediately started typing up their clbutt buttignments on it and turning them in. At about the same time we started getting a few 3278 and 3279 terminals that could display upper and lower case letters. So there was even more demand for lower case printing.

This caused the University to break down and buy a TN print chain for the 1403, and they would stop the student printer every 2 or 3 hours, swap trains, and print the TN print jobs that had queued up. Swapping trains was a greasy noisy PITA, so the operators were just "thrilled" by this.

At about this time I went to my first SHARE meeting, and came back with a printout that John Alvord gave me that had done on a laser printer (a 6670?) I practically fell over when I saw it - the notion that you could get production-quality printouts from a computer, and actually use it for real writing, was a totally new notion...since I came from a world where lower case letters still didn't exist.

Before starting up the 1403 on VM there was a CP command "LOADBUF" that had to be executed to download the chain "map" to the printer, so that it knew what character was at what spot on the chain - this controlled the hammer timing. If the chain and the map didn't match you could easily print a dozen pages of absolute gibberish before anyone noticed. There were facilities in CP to allow you to buttemble the various chain maps that you needed into the hypervisor. The high-speed GN train that we had must have been an RPQ of some sort since CP didn't know about it and I always had to re-gen it into every release of VM.

The DOS-VSE production system that ran the University also ran under VM. It understood enough about VM to issue "CLOSE 00E" commands whenever it was done printing a job, so each DOS batch job showed up as a separate print job in the VM queue. We had to modify it a bit to get it to pbutt chain and form information from DOS-VSE POWER to VM so the operators could mount the proper multipart forms and print trains before printing the jobs. VM-SP was the first release where the CP spooler kept track of form information for print jobs; before that we ran a rather large set of CP modifications to add that support that we got from a SHARE tape of user-contributed mods.

...Sam