|
|
Totally from memory - and probably wrong.... (bear in mind this was early 1960's):
The first thing that struck me was that 1500 Operating was much more of a show-biz performance - but underlying the "flash" was a very basic machine.
In BHS we had a massive 1500 comprising:
 | 6 * 10Kc tape decks |
| 6 * 30Kc tape decks |
| 600 cpm Card Reader |
| 1,000 lpm Analex Printer |
| Paper tape reader - 300 characters per second |
| Paper tape punch - 100 cps |
| Removable Disk Device (looked like a Juke Box - rarely used) |
| Massive Fixed Disk Device - millions of characters |
| 40,000 characters of Memory - each of 6 bits |
Usually 2 operators on shift, with one mounting tapes whilst the other did Card Loading and Console operation. This was the first area for "flashness" to become apparent - the Console man would leap from his console chair and slide across to the Card Reader to unload/load cards.
There was usually a PLT (Program Library Tape) loaded on one of the small 10 Kc (10,000 characters per second) decks - but invariably, wherever it was - the next test/programmer required it moved to another. So "flash" 10 Kc tape loading and unloading was paramount. This PLT contained what would probably now be called ROM - Print and Arithmetic Tables (more of those later) - and initial values for low-order memory. Most programs read from this to ensure the machine was set up correctly - hence these tapes got worn out quite quickly.
More "flashness" came from the buttons used on the 1500 (which was in fact a re-badged RCA 301) - they were of many colours and lit up brightly when pressed - the Console especially looked fantastic.
It was divided into 3 sections left to right, with a bank of red lights at the left to indicate errors, a central section of ??Blue?? Buttons - to enable various Registers to be "selected", and at the right - a Yellow section of Buttons allowing Memory Addresses to be entered or displayed.
Beneath these 3 sections was a row of Green Buttons which allowed special features to be set up
At the far right was a single "Start" button.
This Console went from desk level up to perhaps 10 inches above that - and above that was the Central Processor with a door that had to be banged with your fist (held magnetically shut) - and when opened this displayed the contents of the Registers that couldn't be seen on the Console - e.g. Test Positive, Negative, Zero - Indicator - The Carry Indicator etc..
How a Test session was started - I cannot remember - was it something like...
| Add about 20 "Loader" cards to the front of any card deck supplied by the programmer |
| Load a Program Library Tape on some specified deck, and other tapes as directed |
| Cannot remember - but was that all to ensure that tables and memory were set up correctly (Print and Arithmetic Tables) - I do remember that the 1500 didn't have an arithmetic unit so all maths was done via tables in memory. For example to add 45 to 23 it would take the units first and make an address of 53, and in address 53 - would be an 8. Do the same for Tens - address 42 held a 6 - answer 68 - Carry Register came into its own where things like 8 and 7 were being added. |
| We'd set a green indicator button ON (which one?? why??) set up instructions on the console to read a card and branch to it - that would get a few cards read, at which point the Console Op would set the Green indicator button OFF and hit start again - and the job would run |
| Often the job would crash and the errors were signified by the red error lamps with wonderful name like MRPE (Memory Register Parity Error), DRP (D Register Parity Error), which sometimes were actual hardware failures and the programmer would have asked the Op's to "set P back 10 on any Parity errors" - which involved lots of finger manipulations because any numbers entered via the console had to "maintain Parity" (was it Even Parity?? I cannot remember) - so using 1,2,4,8 bits - encoding the number 2497 required a 2 bit with parity bit, a 4 bit with parity bit, an 8 and a 1 bit with no parity bit and 1,2 and 4 bits with parity bit |
| Another request from programmers was to "run with ALI on" - ALI being the Alarm Inhibit Button - which meant that whenever an error occurred - the program just kept running - until it could go no further. |
| I cannot remember the mechanics of it now - but there was a "Print Table" in Memory which somehow or other facilitated printing. This and many other Memory Tables could easily get overwritten - they were not protected in any way. If a program overwrote the Print Table, and the programmer had asked for a print of Memory, fuses would blow in the printer - requiring the attentions of an engineer to fix. I think the Analex printer we had, had 132 print positions - and there was a "lockout" that prevented all 132 being blown - but programmers often managed to blow less than 132 but sufficient to rile the engineers. |
| More to be determined here......................... |