IBM Department  

Back to the Times Newspaper

This attempts to give you an idea of how an early (around 1960) "punched card" installation operated...

We had IBM:

bullet

024/026 Punch/Verifiers

bullet

??? Interpreter

bullet

077 Collator

bullet

??? Sorter

bullet

519 Reproducer

bullet

609 Calculator

bullet

412 Tabulator

.... and this is how they were used:

Punch/Verify:

Master Files/Tables - as we now know them - were punched onto 80 column punched cards - usually with some form of coloured stripe across the top, so that visual identification was possible - e.g. Employee File may be coloured Green, whilst Customers were Blue.. etc.. - (for more details about the codes used see - http://www.maxmon.com/punch1.htm)

Transactions (perhaps - hours worked this week) were also punched onto these cards.

So, Master Files and Transactions were punched into cards - and were then verified on Verifiers (basically a verifier was a Punch with "hole detection rather than hole creation abilities") and the same info was entered and verified by two different people, thereby reducing errors to an acceptable number. You can just see on the punched card above, right at its top (where any coloured stripe would be) - printing - which interprets/identifies the code which is punched into each column.

Interpreters:

That printing could be put there by the Punch operation - or later by an Interpreter - we used both methods - (for example where cards got punched (not by the punch-girls) - maybe because they wrecked and needed to be re-created - then a manual punch (about the size of a typewriter) would have been used - and they didn't "interpret").

Reproduction:

Also - from time to time - Master Files - got old.. meaning they had passed through machines so often - they were starting to "wreck" too often. Time for the Master File to be "reproduced" - that's when the 519 Reproducer was used. It also did much more than that - but you get the idea..

Sorting:

"Files" (maybe from last night/week's run, or just - straight from the punch-girls) - needed to be sorted into sequence - ready for the next process - that's when we used the Sorter. This where I spent a lot of my time - because - a normal "job" would be to get Sub-Accounts sorted into Account Code sequence - meaning:

bulletGet the 4,000 Sub-Account Cards
bulletSort them into sequence on card column 7 - at 600 cards per minute - but when a "hopper" got full you had to unload it, or a card or 5 got wrecked (and you had to re-punch them)  - so this "maximum speed" - got significantly reduced.
bulletThen - do it again - on card-column 6
bulletThen - do it again - on card-column 5,4,3,2 and 1

Approximate time spent..... 2 hours!! Just to get 4,000 accounts into ascending sequence!!

You'd probably have to do this for the Master Files AND for the Transactions... and then you'd perhaps Merge the two..

Merging/Collating:

This relied on two "files" being in the same sequence - and requiring to be collated/merged - for example there may have been a 3 way merge - where Master Account details were to be followed by Sub-Account details which were to be followed by Transactions.  the end of this process you'd perhaps end up with something like this:

bulletMaster Account 6458700/0000, XYZ Corp, Little Dingwell, Hlu-Hlu, Natal, ZA
bulletSub-Account 6458701/0000, White Rhino Park
bulletTransaction 6458701/4523 advert 4 by 6
bulletTransaction 6458701/4524 advert 8 by 2
bulletTransaction 6458701/4525 advert 3 by 5
bulletSub-Account 6458764/0000, Johannesburg Zoo
bulletTransaction 6458764/2467 advert 6 by 4
bulletTransaction 6458764/6890 advert 3 by 7
bulletMaster Account 8475700/0000, Little Ranch, Hudson Prarie, MN.... etc.....

Calculating:

The merged/collated file would now need to go to the Calculator - where the advert costs would be calculated and punched into each Transaction card

Tabulating/Printing:

All of that would now go into the Tabulator where - perhaps - invoices would be printed, and maybe Taxes added to the Net Total etc... Of course, the format of the punched cards (like file/table formats now) differed from installation to installation - that's where "programming" came in - in the form of plugboard wiring.

Our 412 looked similar to this 407...   thanks to http://www.columbia.edu/acis/history/407.html for the pictures - also this one which shows what we had to do to wire up a plugboard.. . Once tested and proven, the plugboard could be "secured" either by screwing a cover over the wires, or by replacing the easily removable wires with more "secure" ones that "locked" into the hole in the board. A common problem would be that the e.g. 47th print position wasn't printing - during a Statements or Invoicing run - and all that had happened was that the wire on the plugboard had fallen out.

 Back to the Times Newspaper

_____________________________________________________________________________

Links in this site

Shepperton Bing Crosby Triffids Olney 519 Reproducer Newman Street Natsopa Shedland MRPE-DRPE Vibeke Maltravers Melton 1900 Programming Pergammon Letts Howell & Hall Farquhar Arrow Electric Tom Curley Boulting George Pitcher May Flower John Flower Angela Flower Joan Collins Navarone Moor Hall P3/1 ICT 1500 Initial Orders Scicon Bell & Howell British Leyland Charles Letts Philipp Bros The Times ICT/ICL 1900 Bob Hope Charles Little Arnold Keppel Bob Tester 609 Calculator 412 Tabulator P3/2 RCA 301 FP2 Barry McGinn Aramco Chicago Bridge & Iron Gaudern

__________________________________________________________________

 Visit my other sites....

Go to:  Home Web-Site | Lived | Schooled | Family | Pools | Computers | OddsnSods | Photos web-site | POWWEB-1 | POWWEB-2  | Christianz Interiors

Last update - September 9th, 2009 - email to djmikecurley@gmail.com