I’ve been writing software for a long time. In 1963 I wrote my first program, whilst studying Mechanical Engineering at university, in Fortran II, after a quick read of Daniel D. McCraken’s 1961 publication A Guide to Fortran Programming.
The program was based on specification scribbled on the back of a punch card by a Maths lecturer. After a week I was able to attempt to run my first masterpiece. It compiled cleanly – except that I did not understand that Fortran II, as implemented on the IBM 1620 in the Maths department, did not support what it called mixed mode arithmetic. It would not convert integer to float or vice versa.
The IBM 1620 is a story in itself. Three large boxes. A punch card reader/punch; a punch card sorter and a panel mounted on a desk which also housed a teletype. Input was on punch card. Output was on punch card or teletype. The OS, compiler and run-time libraries were on several beautifully coloured punch card decks. In the UK in 1963 this was cutting edge.
After completing my degree I took a job in a pretty prestigious US company in London. This lasted 6 months. In 1967, I left, telling my employer that the job they had asked me to do would be better served by a computer system and went to work for a large European company as a “Scientific Programmer”.
We wrote in PL/I, IBM’s merging of COBOL and Fortran. The work was fascinating. We worked directly with marketing people. Two examples: measuring and predicting the sales impact of a round of money-off coupons on the sale of tooth-paste; predicting the correct time to pick fields of peas based on analysis of “Tenderometer” and “Maturometer” readings taken at each field each day and transmitted to the data centre on paper tape.
To be continued…
I’ve got another 50 years of being a software professional to go. Let me know if I should stop here.