A personal project by Eric Waddell

In 1969, I was 18 years old, having just completed high school. I decided to work for a year before heading off to McMaster University. Because I had taken computer science in Grade 12, I found a job in the Forecasts & Analysis Department at Gulf Oil. With only 10 people, our department was miniscule compared to the accounting department which took up three full floors in our building.

In those days, if you bought gas at a Gulf station with a credit card, the transaction was recorded at the station onto an IBM punched card. Tens of thousands of these cards made their way to our building every day, and an army of people worked to process this ever-growing mountain of cards. It was sheer bedlam as the cards snaked their way throughout the building, floor by floor until they finally made it to Data Processing.

Gulf had recently installed a brand new IBM 360 mainframe. It was a stylish, bright blue mainframe with sharp modern right angles. The beast sat in a glass walled “sanctuary” on a gleaming, raised white floor. With clean, modern lines and not a cable in sight, it was blue, grey and gorgeous.

One of my first projects was to forecast the amount of heating oil required over the next winter for each town an Northern Ontario. I had 10 years of historical data including year-by-year average temperatures and consumption. In Fortran, this analysis would have been trivial, but I was never able to get any mainframe time. Someone suggested I talk to the people who looked after the old “unit record” equipment.

Relegated to an old part of the building, the room was crammed with the old unit record equipment – monster-sized clunkers with big ugly cables and wires all over the place. There was no raised flooring. Most of these machines had been in use since the 1940s and, unlike the ultra-modern mainframe next door, they were all the same ugly dark-grey colour complete with art deco, rounded corners from the 1930s. The whole room smelled of machine oil which keep the clunkers from grinding themselves into piles of metal dust.

IBM 405 Accounting Machine
1934 – 1949

In those days, computers cost millions of dollars, so getting time on Gulf’s mainframe was next to impossible. One of the old-timers, who was getting ready to retire, listened to what I needed and said I could do what I needed using a card sorter, a collator, and a rudimentary calculator-printer. He let me borrow some of his old, musty manuals and I was soon “programming” by plugging wires into plug-boards. I was in business. Instead of storing my data of a disk drives, I used punched cards for everything.

I was fascinated by these electrical-mechanical monstrosities. I was amazed that people of my parents’ and grand-parents’ generation had managed to create entire data processing systems long before computers became available. Although I went on to developing software for mainframes, minicomputers, embedded processors, PCs, and eventually web servers, I’ve always been fascinated by the history of some of the old equipment.
This articles discussed the advances made between Herman Hollerith’s invention of the Tabulator in 1892 and the beginning of the 1950s when general purpose computers made their first appearance.