My first encounter with computer-based learning was in 1975 when I trialled packages produced by the Chelsea Science Simulation Project while on teaching practice. Access to the minicomputer running the packages was via modem and teletype, and response times were measured in minutes rather than seconds. Three of the four packages were little more than computer-based lookup tables but the fourth, despite the severe limitations of the technology, inspired me to take computers seriously as tools for learning. The application was a simulation of crossing different strains of fruit fly, a technique beloved of geneticists but impractical for school biology. I was impressed by the use of a computer to achieve something highly educational that was impractical by any other means.