Good point. Many early clocks were (or contained) primitive analogue computers, so I think you win there... sort of. Two books to look at: "The Mediaeval Machine" by J. Gimpel and "A History of Engineering in Classical and Mediaeval Times" by D. R. Hill. Gimpel will fill several holes in your timeline - Su Sung made quite a complicated astronomical clock in c. 1090; the middle ages saw a sizeable crop of similar machines in the west, culminating in that of Giovanni Dondi, under development from 1348 until 1364. Hill's treatment of clocks is also interesting. He points out that the Classical civilisations had (presumably inherited from the ancients) a system by which the hour changed in length depending on the date so that sunrise to sunset was always twelve hours. Thus ordinary timekeeping clocks had to combine time and date in an analogue computer to get hour-number out at the end. Some of the mechanisms Hill describes get this quite wrong! (I don't recall any of these clocks also taking account of lattitude...) OK. I'll go for the Z1 as the first _general purpose_ computer. Philip.