Are you _sure_ the ABC was electronic? I was under the impression that the 1930s machines were all (almost all?) relay computers. Konrad Zuse is the pioneer whose name is oft mentioned here... As I see it, the sequence of events is as follows: The 1940s saw the valve (vacuum tube) computers begin to emerge - in some order (still in debate) ENIAC, Univac and the hush-hush British project, Colossus (hush-hush because it was part of the war effort), all appeared in 1943 I think. Colossus currently claims to have been first, but it is hard to verify with all the wartime secrecy that surrounded it. The Manchester machine claims to be the first _stored program_ machine. It was the first electronic computer, and I think also the first computer, to hold its software in main memory. It was far too small for this to be sensible - the purpose was to demonstrate the principal of it with a view to using similar hardware and software designs on larger machines in the future. All previous machines had a main store for data, and a programming panel for patch leads etc. to hold instructions. Soon after the Manchester machine ran, the EDSAC project in Cambridge pulled ahead with a large scale stored program (Von Neumann) machine. Professor Wilkes, who ran this project, said in one talk he gave that they had wanted to include floating point in EDASC, since it was already available on many relay computers, but this had to wait until a later design... Philip.