Originals of this book are very easy to get a hold of. The military printed millions? of these. They are so common that at one point I was using the unbound pages as packing material. -- Will

Paul Koning wrote: Why not just duplicate the real thing! Tubes are cheap now days! $6.00 each for the orginal tube, Since the ADDER block is the problem, that I don't expect to be a major problem to duplicate. My guess about $250 today to duplicate.

Indeed it does. Back when I was at Purdue, I knew Saul Rosen who went WAY back. He referred to the push to prioritize Atanasoff as the cult of Atanasoff. But it's so clear that this weird saga really is about the legal/social and not the technical. Both groups innovated in different ways and both are important to study to get a good picture of history and of the evolution in design. Everyone had to address the fundamental question of memory. Eckert and Mauchley knew that it was an eventual issue; Eckert wrote a paper (in 1944 IIRC) where he described a magnetic disk (and even suggested a stored program on it). But in the ENIAC, they largely punted on the question of memory. But Atanasoff's rotating capacitor drum really does stand out as an innovation in memory design. Even if it wouldn't scale well, the rotating, regenerative memory structure is very similar in concept to the recirculating, regenerative delay line memories, like that used in the UNIVAC I. This is the sort of thing that led Eckert to remark that the processing part of the ABC would never have been particularly reliable. His design for the ENIAC was radically different. It bore much similarity to his earlier experience in tube-based counters. And his design techniques were the only reason a machine with 18,000 tubes could be expected to operate for any time at all. Ultimately, we can see seeds of future technology in both machines, but neither machine bore much resemblance to what came after either in terms of architecture or the design of basic functional elements. BLS

I can't help but see Eckert as one of the truly brilliant engineers of the 20th century. Thanks, but my wife gets really annoyed at how long it takes me to compose considered responses to her. :-) Indeed, almost everything was quite incremental when you look at what everyone else was doing. Just a sampling with no intended disrespect or suggestion of insignificance for those left out: 1933-1941 - Mauchly's experiments with neon bulbs for digital circuits (I couldn't find an exact date but it was while he was at Ursinus which was between 1933 and 1941.) 1937 - Shannon's MS thesis on binary numbers and Boolean algebra - George Stibitz' Model K binary half adder with relays 1939-1944 - Development of the Harvard Mark I (Aiken) with relays and rotating decimal components patterned in part after Babbage 1940 - The Bell Labs Complex Numerical Calculator extending Stibitz' binary relay system to a full calculator 1941 - The ABC implementing binary with active electronics 1942-1945 - Development of the ENIAC using active electronics in a "rotating" decimal way and even more general than the Harvard Mark I 1944 - Beginning of EDVAC design and introduction of the stored program idea using active electronics in binary 1945 - von Neumann's "First Draft" paper making the design ideas for the EDVAC widely known 1946 - The summer school on computers at U Penn attended by those who went back and developed the Manchester Mark I and the EDSAC 1948 - The ENIAC reconfigured to operate as a type of stored program machine 1948,9 - Manchester Mark I operational 1949 - EDSAC and EDVAC operational So given all these events over a period of 15 years or less, who invented the computer? That was rhetorical, in case it wasn't obvious. Unless we say it was Babbage for the idea and partial implementation of the automatic sequencing of mathematical operations in a mechanical system or was Eckert, and independently Zuse, for the stored program, I don't think there's really anyone who can be said to have invented the computer. But rather the best approach to "compute these numbers by steam"* grew organically through the work of numerous very bright people. I agree. There are many people who are pretty much unknown but who were the intellectual foundations of everyday life. It seems that it's easier to popularize someone when they build something that can be shown on TV than when their contribution is strictly intellectual. Einstein and Hawking were interesting people which drew peoples' attention to their work. It's like with Church and Turing. Both developed universal models of computation at about the same time, but Turing is by far the better known. A lot of it comes from his work at Blechley Park, his work on the pilot ACE, his excentricities (like chaining his coffee cup to the radiator and riding his bike with a gas mask on), and of course the tragic circumstances of his death. Well, intended for the lay person or not, I'd like to see something like that. BLS * The computing numbers by steam is a paraphrase of a statement Babbage made as a student that led to his work on the differential engine and later the analytical engine. He was staring at a book containing mathematical tables and a friend asked him what he was thinking and that was his answer. Hmmm come to think of it "Computing Numbers by Steam" would make a good title for a book on computer history...

Definitely. In fact, as I understand it, Eckert had quite a bit of experience with high speed counters in the radar world. That's what made the decade counters such a natural design for him, as opposed to a binary approach that would have saved tubes. Where I think he really shown on the ENIAC however, was in getting the level of reliability he did out of the components of the day. My understanding is that it was one of the earliest applications of worst-worst design and of derating the tubes, running them at lower filament and plate voltages than normal, to extend their lives. BLS

"Brian L. Stuart" wrote: Something I've wondered about with ENIAC is just where did all those 18000 tubes go (as in, where were they used in the machine)? With 20 accumulators of 10 decimal digits each, each digit being a decade ring counter of flip-flops, that's: 20*10*10*2 = 4000 triodes or 2000 duo-triodes. Somewhere I got the impression that number was about doubled for support circuitry around the ring counters, which would put the count around 8000 or 4000, depending on how one is counting. The accumulators account for a substantial portion of the machine (the majority of the rack panels). I know each panel had a fair bit of control-sequencing circuitry, and the power supplies may have accounted for quite a number, but still, where did all the other tubes go? And how feasible would it have been to replace those ring counters with 4-bit binary decade counters to save a lot of tubes? Perhaps it wouldn't have helped much because of the way the values from each counter were transmitted around the machine. Binary counters were around by then of course, I'm not sure when binary counters were first wrapped into decade counters ( by late 40's at latest). It would be fun to examine the ENIAC schematics for these sorts of questions.

Brian, ENIAC was a decimal machine, not binary. Mauchly wanted the Navy to be able to understand the proposal so they made it decimal. It was also not a stored program machine; it was rewired for each problem. They were well into the design and construction when they realized that there was a better way to do this so their next machine, the UNIVAC I, was binary, stored program, and had about 30% as many tubes. ENIAC, The Triumphs and Tragedies of the World's First Computer by Scott McCartney is my favorite on Eckert and Mauchly and their computers. Jim Brian L. Stuart wrote: