Towards the end of my final year at Cambridge, I went to deliver some late work to the office of John Rushbrooke in the Cavendish. He asked me what I was going to do after graduating, and when I told him the plan was to go into computers, he looked sadly round his office. The desk, the table and most of the shelves were covered with large rolls of punched paper tape on their way to or from Titan. John sighed and said ‘You want to keep away from those things’. As I quickly discovered after moving to the Data Handling Division (DD), this reaction was typical of many physicists of his, and older, generations. Originally, experiments in particle physics had fitted on a benchtop, and the results had been a few numbers scribbled in a log book. Later, when bubble chambers took over, the results were in the form of thousands of large photographs of particle tracks. These had to be visually scanned on special plotting tables, and the ones with interesting tracks had to be very precisely measured. DD had started life largely to support this scanning activity, including a modest amount of computerised calculation, in due course adding pattern recognition software to reduce the need for human eyestrain.