What's that got to do with it? Diodes are analogue parts - the output (current) is a continuous function of the input (voltage), not a discrete one (to me the difference between an analogue and a digital component). In fact, Allison, you were saying only a few days ago that you don't need any amplification to make an analogue _computer_ (with which I agree - although some of your examples I wouldn't call computers). That is an excellent concept. Thank you - I'll remember that. For non-electronic digital computers, where do Facit mechanical calculators lie? I have one (which is driven by an electric motor but could conceivably use any motor) which has algorithms for optimised decimal multiplication and non-restoring decimal division. It is not programmable, but it is pretty complex - and all done mechanically. For pneumatic computers, I think some pipe organs of the turn of the century came close - you could program some buttons to set various combinations of ranks for fast selection during performance. However, the more complicated schemes of this nature (popular around 1920) used electrical as well as pneumatic logic elements. Philip.