Talk:ENIAC/Archive 2

Archive 1

Archive 2

Archive 3

Mention of hydrogen bomb belongs in lead?

Re: this edit, GroveGuy believes it is important to include in the lead that ENIAC's first problem was for the hydrogen bomb. I see this as more of a parenthetical that would be more appropriate to mention and fully discuss later in the article. Any other thoughts on this? WP:LEAD, and especially the section on "Relative emphasis", is relevant here. Robert K S (talk) 12:59, 4 February 2011 (UTC)

If true, then it would seem absolutely correct to mention that in the intro; the current intro gives a very different impression which is seriously misleading if in fact ENIAC was first used for something quite different than is stated now. quota (talk) 13:52, 4 February 2011 (UTC)

But be careful! "First used" is ambiguous phraseology. It could mean "initial purpose", but here, it does not. That a hydrogen bomb-related problem happened to be the first task to which the ENIAC was put to work was an accidental or coincidental sort of thing. It wasn't "essential" to the ENIAC in that it doesn't explain any of the major who-what-when-where-why-how questions that a well-written lead is supposed to address. In that sense, it's sort of interesting trivia, a fun aside, but not proper material for the lead. Yes, it's fascinating information in an "oh-by-the-way" sense. Good to have in the article? Surely. But the lead? Does it address one of those major questions? Given that it was a one-time thing, and the ENIAC wasn't built specifically for it, why is it appropriate content for the lead? In my opinion, if included there, it is more misleading than helpful. Robert K S (talk) 22:02, 4 February 2011 (UTC)

I believe that the change in sponsorship of ENIAC belongs in the lead paragraph. Today ENIAC is on the front page of CNN and they again promulgate the myth of the artillery firing tables.One year into the ENIAC project it came to the attention of the atomic bomb people. From that point forward the lion's share of the computations done on the project were atomic bomb calculations. Other calculation got the leftover time on weekends and holidays. It went from a $61,000 to a $486,000 project over three years. Notice that the patent dispute says "The Los Alamos calculations which commenced December 10, 1945, were the first problem placed on the ENIAC machine. When the first problem was put on the machine, it was the first time that the machine as a whole was being used.". The atomic bomb was the largest and the most secret project of World War II. Let's do the right thing here and restore ENIAC's association with the bomb. GroveGuy (talk) 22:48, 8 February 2011 (UTC)

We're debating *which* way the ENIAC was devoted to killing people? Does it really matter? Weaons all have the same ultimate purpose. Artillery tables must have paled in importance after the bomb came out. --Wtshymanski (talk) 22:58, 8 February 2011 (UTC)

GroveGuy, sorry, but your response is only misinformation or confusion. There was never a "change in sponsorship" of the ENIAC. From beginning to end the project was financed by the Ballistics Research Laboratory based in Aberdeen, Maryland. It is not true that "the lion's share of the computations done on the project were atomic bomb calculations". I can't tell whether your response really confuses the atomic bomb and the hydrogen bomb or whether you're only mixing up terminology. In any case, the first problem to be run on the ENIAC which took advantage of the entire machine was a problem run by Stanley Frankel and Nicholas Metropolis for the hydrogen bomb. The gradual increase in the project's budget that you make reference to was by request of the Army as additional accumulator units and features were requested so that the machine would be capable of handling broader classes of problems, not specifically hydrodynamics problems or problems of interest to nuclear weapons scientists. Your statement "Other calculation got the leftover time on weekends and holidays" seems to be conflating into your own new theory other elements of the story: John Mauchly on a few occasions made use of the machine on weekends and holidays to run hobby problems of academic interest, such as Dick Lehmer's number sieves. The fact that "the atomic bomb was the largest and most secret project of World War II" is of no moment whatsoever since both atom bombs were dropped in August 1945 without the aid of ENIAC, which wasn't finished until months afterward in November. Also, I don't think there's any evidence that the hydrogen bomb people ever ran a problem on the ENIAC again after that first run; perhaps they did, but we're not here to speculate—we are bound with what is verifiable in the record. Robert K S (talk) 01:28, 9 February 2011 (UTC)

Robert - You are quite right. I read Goldstine's book and was struck by the fact that while the Ballistics Research Laboratory initiated this project Goldstine says on page 214 the problem being computed was hydrodynamics. Goldstine says on page 226 "that first run" was a million punch cards. That's a thousand boxes of cards. It boggles the mind. And to have a project's budget doubled and doubled and doubled again seems pretty extraordinary. So, reading between the lines, so to speak, it seemed to me the h-bomb people had taken over the project. I don't have a source to quote that exactly says this, but it seems to me that that's what happended. And if we're looking for consensus, note that both quota and Wtshymanski support me. GroveGuy (talk) 07:21, 9 February 2011 (UTC)

Goldstine's book doesn't literally say what the calculations were for; he points out that he and other ENIAC personnel weren't cleared for weapons research and so no-one told them exactly what the calculations were used for, but Los Alamos wasn't working on improving watch dials. --Wtshymanski (talk) 14:41, 9 February 2011 (UTC)

While we don't know exactly what the mathematical-physical nature of the Los Alamos problem, it is not true that we don't know its general nature as a simulation of a theory for the hydrogen bomb--Teller's theory, in fact. Edward Teller testified about the problem in Honeywell v. Sperry Rand, years before Goldstine wrote his book. See, e.g., McCartney page 104. As for GroveGuy's "reading between the lines", it is not accurate. It is true that Los Alamos heard about the ENIAC's successful completion and was eager to run a problem on it, and was granted priority to get the first major problem as soon as the machine was finished being tested out. It is not true that this problem was responsible for the expansion of the ENIAC contract and the increase in its budget. Actually, the budget increase and expansion of the project to enhance the computer's capabilities was anticipated from the very beginning of the contract by Colonel Gillon, among others, as Goldstine, Mauchly, Eckert and others related on numerous occasions. Cheers, Robert K S (talk) 00:44, 10 February 2011 (UTC)

Lack of Information

The history section is incomplete. There's no mention at all about when it stopped being used or what happened to it exactly. Gune (talk) 08:19, 14 February 2012 (UTC)

Tube failures

The article states:

Some electronics experts predicted that tube failures would occur so frequently that the machine would never be useful. This prediction turned out to be partially correct: several tubes burned out almost every day, leaving it nonfunctional about half the time. (According to a 1989 interview with Eckert the continuously failing tubes story was a myth: "We had a tube fail about every two days and we could locate the problem within 15 minutes.") Special high-reliability tubes were not available until 1948. Most of these failures, however, occurred during the warm-up and cool-down periods, when the tube heaters and cathodes were under the most thermal stress. By the simple (if expensive) expedient of never turning the machine off, the engineers reduced ENIAC's tube failures to the more acceptable rate of one tube every two days. In 1954, the longest continuous period of operation without a failure was 116 hours (close to five days). Given the technology available at the time, this failure rate was remarkably low, and stands as a tribute to the precise engineering of ENIAC.

I have highlighted both Eckert's recalled failure rate and the rate previously in the article (before the addition of the Eckert comment) and they look identical to me. Perhaps Eckert was recalling only the later rate of failure instead of initial problems before the solution of leaving the machine on was discovered... -- RTC 01:10, 15 February 2006 (UTC)

In a television interview, I remember either Eckert or Mauchly admitting they knew about the possibility of poor reliability. To deal with it, they adjusted the design to improve reliability. Most notably, they were very conservative and chose to operate tube filaments considerably lower than their ratings. Can anyone confirm this? Madhu 23:14, 21 March 2006 (UTC)

I have read several times the book Colossus, and think it's interesting to mention that in the use of valves, Tommy Flowers is credited with recognising that the way to keep them running reliably is never to turn them off. I think the book states that it was the warm up and cool down of valves that degraded the metals inside them until they eventually failed; isn't this a bit like light bulbs by the way? Anyhow I don't think this needs mentioning in the main article but just thought I'd slip it in for information that Colossus had addressed the reliability of valves before ENIAC came to the same solution. I also am tickled by the debates over the definition debate to keep ENIAC as a first... I can't actually spot what defines a machine as general purpose, but as far as I can see ENIAC did the ballistic calculations it was created for, then was headhunted to perform atomic calculations; is that what makes it general purpose? I'll re-read the book as I am pretty sure Colossus was reprogrammed to take on different tasks. Hmm. Mungo Shuntbox (talk) 07:31, 29 July 2012 (UTC)

First sentence claims

The article states that "[Eniac] was the first Turing-complete, digital computer capable of being reprogrammed to solve a full range of computing problems". This is not the case however. The German Z3 computer, built in 1941, was Turing complete. I present this source: http://ieeexplore.ieee.org/search/wrapper.jsp?arnumber=707574

The source, from 1996, claiming otherwise is outdated, since Z3 was proven Turing complete in 1998 —Preceding unsigned comment added by Bugfindersolas (talk • contribs) 11:58, 10 October 2008 (UTC)

As many commentators have noted, this article is somewhat misleading regarding the claims of first this or that. A modern computer has 3 attributes: 1) digital, 2) electronic, ie, operating without moving parts, and 3) programmable, which means running off of an electronically stored program. Neither the ENIAC (as first designed) nor the Colossus were programmable. The first machine that satisfied all 3 criteria is generally presumed to have been the Manchester Baby. Jfgrcar (talk) —Preceding undated comment was added on 22:31, 24 February 2009 (UTC).

In the LEO article I added, I thought fairly reaonably, "contender as the" to the bald assertion that ENIAC was the "first general-purpose electronic computer". This has now been reverted twice. Opinions? Really this talk is more suitable to be discussed in the LEO talk pages but since have been told by the editor making the reverts to discuss it here, that is what I shall do. I see less harm in this article making the bald statemenet (though I disagree with it) but I can't see why it should spread out into other articles. It was marked as "weasel words" but I don't think indicating, OUTSIDE OF THE ENIAC ARTICLE ITSELF, that it is not a bald fact. SimonTrew (talk) 15:16, 28 February 2009 (UTC)

It is indeed "the first general-purpose electronic computer". Regarding "a" vs "the": If the ENIAC was "a" general-purpose electronic computer then it had to be "the first." There are no claims that the ABC, the Colossus, or the Z3 were "other" general-purpose electronic computers. The ABC and the Colossus were clearly not general-purpose, without multipliers, branching or even a sequence of instructions. The Z3 was clearly not electronic, running at 5 Hz. Programmability is essential; however the significance of the stored-program is greatly misunderstood. Its importance was speed. Paper tape was the conventional method of sequencing machines at the time. It could never, ever keep up with electronics. The breakthrough was a sequencing method that was NOT mechanical, allowing machines to run at Megahertz speeds. The ENIAC managed to do this without using precious RAM; only plugboards and switches. That did not make it a less capable machine; in fact it was much faster than the first few EDVAC-type machines like the Baby. I think it is proper to demand that a computer must be programmable and general-purpose, and both the ENIAC and Z3 qualify. Of those, one was electronic and 1000 times faster. Zebbie (talk) 20:46, 16 March 2009 (UTC)

[Regarding ENIAC being "the first general-purpose electronic computer"] I and many others regard that as an opinion rather than a fact, hence our discussion. [Regarding the "a" vs. "the first" distinction] That makes no sense. Either it was the first general purpose electronic computer or it was just another computer. [Regarding the "no claims" that previous computers were both electronic and general-purpose] There are many such claims. My digital watch runs at not much more than 5Hz, are you claiming that is not electronic? [Regarding the assertion that the ENIAC's important advance was its speed] No. Its importance is the Von Neumann architecture, and particularly, the ability to write self-modifying code-- something we still all rely on (a program loader is, after all, self-modifying code, although the technique outside of that particular function is now generally frowned upon). [Regarding paper tape sequencing being the conventional method of sequencing] Some Bletchley peripherals used photoelectic cells to read tape quicker than could be done mechanically. So, no prior claim there. [Regarding the ENIAC being faster than the Z3] Being faster doesn't make it the first— in fact almost by definition you have accepted there were earlier computers. The claim that Z3 was not electronic, I think that's indisputably correct. I think others here have well expressed their terms of definition more succinctly than I could. I don't really mind the claim of being "first" on the ENIAC page (well I am not too happy about it, but it will just be a constant edit war); I dislike that other pages for Colossus, LEO, etc get changed when I put down that they "have a claim to be the first" or language like that. Others might see that as weasel words, but they are not: the claim may, in some eyes, not stand up to scrutiny, but the fact is many have claimed it. SimonTrew (talk) 21:08, 16 March 2009 (UTC)

Regarding Zebbie's remark about the "electronic" primacy claim and SimonTrew's follow-up, "My digital watch runs at not much more than 5Hz, are you claiming that is not electronic?": What Zebbie is saying is that, in order for a computer's priority claim of being electronic to have any substantial meaning, it must be able to operate so as to take advantage of electronic speeds. The mere presence of electronic components does not render a device "electronic"; if it did, then the differential analyzer in the basement of the Moore School was an electronic computer, because Eckert modified it to use electronic torque amplifiers. The ABC computed electronically but operated at ~60 Hz electrical speeds because of the way its memory rotation was timed. I wouldn't argue that that disqualifies calling it an "electronic computer", but even so, it was not a general-purpose electronic computer, and nor were the Z3 or the Colossi. Robert K S (talk) 20:15, 17 March 2009 (UTC)

I see now my discussion of articles (a and the) was, well, too wordy. I was trying to convey a logic, and that is:

0) Given that the ABC, the Z3, the Colossus, the Mark I and the ENIAC were computers,

1) The ABC was not general-purpose;

2) The Colossus was not general-purpose;

3) The Z3 was not electronic;

4) The Harvard Mark I was not electronic;

5) The ENIAC was both general-purpose and electronic.

6) Therefore it is safe to say that the ENIAC is the ONLY general-purpose electronic computer in this group.

Since it is "the only", it is indeed "the first", and not simply "a."

Of the above statements, Items 1 through 5 are "bald facts;" I don't know of a challenge to that. There have certainly been many arguments over statement zero (0). Perhaps they are not all "computers." (Most of the arguments spring from a lack of precision in the definition of the word "computer.") Zebbie (talk) 22:32, 20 March 2009 (UTC)

Everything in the above comment is true - but irrelevant because it omits one of the contenders. The SSEM ("Manchester Baby") was digital, electronic, and programmable, the three criteria for a computer listed at the top of this thread. At the time the SSEM was built, ENIAC was not yet programmable (the source in the article for the addition of "a primitive read-only stored programming mechanism" comes from September 1948, and talks of the new facilities being available "by December"; the Manchester Baby had already run its first program in June). 93.97.184.230 (talk) 20:50, 29 July 2012 (UTC)

Inexact information?

The comparative table of the first computers contains some information that may be inexact.

1. ENIAC became operational in November 1945, when it ran a series of calculations for the hydrogen bomb. ENIAC was secretly operating a few months before it was introduced to the public in July 1946.

2. According to the book "Origins of Cyberspace: A Library on the History of Computing, Networking and Telecommunications" by Diana H. Hook and Jeremy M. Norman, Zuse Z4 did not have conditional branching in its original 1945 version. Z4 became Turing-complete only in its final version, which became operational in 1949-1950. Well, one may argue that the same odd hack that is described in Raul Rojas's paper on Z3's purely theoretical "Turing-completeness" could work also with Z4, but that's not the way how Konrad Zuse himself had implemented the branching later.

3. I am not sure that "Modified ENIAC" deserves a separate entry in the table. All those computers, including ENIAC, were modified significantly a number of times through their history, but there was no such machine called "Modified ENIAC". These modifications were just a part of ENIAC's development, followed by a series of other modifications.

I am not going to change any of these details, but I think someone who is an expert in computer history should check it out. Laplandian (talk) 20:48, 14 December 2011 (UTC)

I agree that it would be appropriate for someone who is an expert on the history of computing to improve the table. It is, however, a free-standing entity within Wikipedia, a 'Template' under the titile of "Early computer characteristics" that is 'transcluded' into this article and a number of others. I am therefore going to copy this section into the talk page for it which is at Template talk:Early computer characteristics. --TedColes (talk) 12:07, 14 February 2012 (UTC)

What is the source for (1) above? As regards (2)I am concerned that Turing compatability is given undue prominence in the table. Perhaps it would be better replaced with the number of machines constructed, but that would, I think, be one for all except Colossus Mark 2. Alternatively 'Memory size' might be useful, if an agreed definition of memory in these machines could be agreed. Re. (3), I agree that the 'Modified ENIAC' row of the table could well be deleted similarly the Mark 1 and Mark 2 Colossus could be merged into a single row. --TedColes (talk) 09:20, 6 April 2012 (UTC)

No comment on (1) or (2) (no quibbles from me) but as to (3): while it may be true that a number of computing machines were modified throughout their lifetime, the ENIAC modification was a striking one that significantly changed the way the machine was used forever. In one sense, it hobbled the machine by preventing the full range of use for which it was designed. On the other hand, it permitted for faster set-up times. The machine no longer required more than a day's worth of re-cabling setup for a new problem. Essentially the only thing that needed to be changed was to enter the "program" by means of setting the dials on the function tables. This modification fundamentally changed the way the machine was used, and the machine was never used the old way again. Robert K S (talk) 16:51, 30 November 2012 (UTC)

Architecture!

The one thing this article lacks is a description of how ENIAC actually worked. A fetishistic listing of its size, weight, number of valves and panoply of functional units does not tell you this. Rather, you need to know about how the bits fitted together, and how they were configured and controlled. This is difficult, because ENIAC is so different from modern post- Baby computers, but it has to be done. This is absolutely crucial to understanding ENIAC's place as a transitional form between things like differential analysers and the modern concept of a computer, and deciding whether you think it was truly a programmable machine: before von Neumann's 1948 modifications, there was no stored program, but rather an implicit program embodied in the way the largely autonomous functional units were cabled together.

The only good account of ENIAC's architecture written for modern readers that i've found is in a paper by Atsushi Akera:

Exploring the architecture of an early machine: the historical relevance of the Eniac machine architecture - A. Akera, M. Marcus (1996); IEEE Annals of the History of Computing 18(1):17

Which is sadly behind a subscription wall. I've read it, but still don't understand it well enough to be able to write a good description of the architecture. Does anyone fancy doing this? Pleeeeeease?

-- Tom Anderson 2007-07-17 21:50 +0100

Re: "But it has to be done"--such a functional description of the ENIAC, any more than one or two paragraphs (which I think would be insufficient for the type of description you seek), is quite beyond the scope of this article. The 1981 Burks & Burks article for the Annals is well over a hundred pages and barely scratches the surface in this regard. Those looking for information about how the ENIAC worked would be advised to seek more complete, more authoritative sources than a Wikipedia article. It is enough for this article to say what the computer was, what function it served, who built it, and some of the particulars of its technological and political history. In other words, its significance to the history of computing as a whole. (In the details of its implementation, it lends little to that history, as virtually everything about its logical and electronic design was quickly superseded.) Robert K S 03:39, 11 November 2007 (UTC)

An uncommon perspective on origins of the ENIAC's architecture

From reading in past years, I rather vividly recall how lots of Friden four-function desktop calculators (Perhaps model STWs, or possibly STs) were operated by women. Despite their best efforts (and I do believe they were quite conscientious), the whole operation went awfully slowly. Apparently there were many stages of computation, and the results read from the calculator carriage dials had to be transcribed and manually entered into other calculators designated to handle a later stage of the problem. Earlier in my life, I learned about details of the Friden mechanical line, and have had a sustained (but waning!) interest in mechanical calculators. As well, I have been relatively well informed about electronics, having once seen schematics of the ENIAC decimal accumulators (probably in the IEEE Annals / Hist. / Computing). Most of my lifetime work was as an electronic technician.

What seems quite evident, although I have never read this idea explicitly, was that a major bottleneck in the slowness of the bank-of-calculators effort was the time required to transfer the contents of one calculator's result dials to another calculator's keyboard to be used for a calculation.

I strongly suspect that the basic concept of ENIAC was to "simulate" ten places of the Friden 20-place result (accumulator) dials, or the 10-place quotient dials. This was done by the 10-digit decimal accumulators. Cables interconnected the ENIAC accumulators, tremendously speeding up transfer of intermediate results. As well, Fridens were limited to 10 operations per second; even the wondrously ingenious and sophisticated Marchants operated at only 1300 cycles per minute. As well, entering numbers manually took seconds per calculation.

It seems to me that the ENIAC designers wanted to interconnect the decimal accumulators via pluggable cables, with the specific arrangement depending upon the problem to be solved. As well, calculation speed was hardly a trivial enhancement, in particular for mult., division, and square root. (Was the Friden SRW, which did square root, developed to assist in ballistics calculations? Would be interesting to know!) (The SRW used the division mechanism, with a quite-modest increase in complexity of the total machine, perhaps 10% or so. Think of it as a division operation, but with a constantly-increasing number in the keyboard.)

After WW II, (almost certain), Friden sold Computypers, business/accounting machines with Flexowriter input and output. Those were via paper tape or edge-punched cards, the latter using paper tape standards, as well as keyboard and typed documents. Inside the machine was an STW mechanism with solenoids operating the keystems and rotary switches on the shafts of the result dials -- it had electrical input, output, and commands. It's just possible that the electrical I/O had its origins in trying to speed up the earlier partly-human effort. (Of course, the actual computations were no faster in the Computyper than in a desktop calculator.)

Regards, Nikevich 10:08, 6 December 2012 (UTC)

What a fascinting speculation. But are there any souces that can be cited to provide evidence?--TedColes (talk) 10:25, 6 December 2012 (UTC)

Programming (first 2 paragraphs)

Hello GroveGuy, If you feel that my edits are changing the meaning of a sentence or are not constructive, please explain why. I've carefully made these edits to make the meaning of the paragraphs more clear by improving the flow of the sentences. I'm guessing you wrote these paragraphs and that's why you have such an interest in maintaining them, but this is a public encyclopedia, not a personal project, so all work is subject to revision and change. If anyone else has opinions, here is the diff, and please do share. C0h3n (talk) 13:31, 1 February 2014 (UTC)

C0h3n - the common belief is that they were running programs on ENIAC for artillery tables. I believe Goldstein says that the project got taken over early on by the Atomic Bomb people. Your change takes away from the conflict of the table people vs the bomb people. In the second change single stepping the machine is not the same as executing the program step by step. It is a level lower. I, too, invite others to give their opinions. Don't forget to sign your comments on talk pages. GroveGuy (talk) 03:48, 1 February 2014 (UTC)

I'm not sure what single stepping the machine is in that case, but perhaps a brief explanation would be in order. I'll add the artillery tables part back in. C0h3n (talk) 13:31, 1 February 2014 (UTC)

Colossus computer a few months earlier than ENIAC

If you read and compare the Colossus computer and ENIAC articles in detail, it appears that the Colossus computer was earlier. Thus the claim of ENIAC being the first computer is debatable. But I suppose you could add lots of qualifications to your definition of computer to make ENIAC earlier. 78.144.207.126 (talk) 22:37, 15 December 2009 (UTC)

There is no doubt that Colossus was working and serving its end users in February 1944, at least two years before the same could be said to be true of ENIAC. The problem lies in defining 'computer'. ENIAC had more of the characteristics of today's computers than Colossus did, but did not have its program stored in volatile memory as was the case with the SSEM. Colossus ENIAC was, however a general-purpose machine whereas Colossus was designed and implemented for a narrow range of special purposes. The phrase 'the first computer' is not very helpful, and the apparent suggestion that the definition of 'computer' should be qualified to ensure that ENIAC comes out as the first, does not do justice to the various parallel developments in different places that happened in the 1940s.--TedColes (talk) 23:11, 15 December 2009 (UTC)

"Colossus was, however a general-purpose machine whereas Colossus was designed and implemented for a narrow range of special purposes." Which was it? General-purpose or special purpose? Jpaulm (talk) 15:44, 12 June 2010 (UTC)

I'm sure this was a typo. Ted must have meant to say "ENIAC was, however a general-purpose machine whereas Colossus was designed and implemented for a narrow range of special purposes." GroveGuy (talk) 20:09, 12 June 2010 (UTC)

Yes, that is what I meant to say. My apologies. --TedColes (talk) 22:37, 12 June 2010 (UTC)

Both the Colossus and the Zuse Z3 were earlier computers under different definitions of computers. I think it's misleading to describe ENIAC as the first computer using a highly qualified definition. ~~Andrew Keenan Richardson~~ 22:42, 29 July 2010 (UTC)

ENIAC should not be described as the first computer. Colossus most certainly came before it and it provided a significantly more remarkable service to humanity than ENIAC. Andrew Keenan Richardson is right, using a highly qualified definition to allow it to be called the first is misleading. JamesGeddes (talk) 20:21, 20 August 2014 (UTC)

Error in caption?

It seems that the close-up photos of the "tubes" is not actually of vacuum tubes at all . . . But rather they are metal-can electrolytic capacitors. The mounting clamp is clearly a common VR style that's been traditionally used for this purpose, and the gooey paper stuff between them and the capacitor is most likely an insulation material to keep the negative side of the capacitors insulated from the chassis (i.e. for a negative supply rail). — Preceding unsigned comment added by 75.81.31.29 (talk) 06:52, 29 July 2012 (UTC)

I'm really puzzled by this comment; it must refer to a photo that has been replaced by the current one, which shows the orange glow from the heaters and cathodes of the tubes. I'm 76, and when I was young, octal-based tubes (such as 6L6s in some of today's guitar amplifiers) were the most common. The non-circular details in the orange glows are created by the mica supports inside the tubes.

It's likely that not all tubes look alike, because several types were used, and perhaps because those of a given type were made by several companies.

My first impression when I saw the photo was that the tubes are so closely spaced that cooling must have been of real concern. Then, I thought of total heater power...

Regards, Nikevich 09:31, 6 December 2012 (UTC)

These are definitely not tubes, but electrolytic capacitors. The orignal rack is in the HNF Computermuseum in Paderborn, which I have just today checked there. The original picture was 90° rotated, as in reality. Thus, I will remove picture, at least here.

Rainglasz (talk) 19:59, 30 January 2015 (UTC)

Have reverted the removal, as it is used in many other articles, and will next week coordinate with the museum that has the original, and eventually file a renaming request in WikiMedia. Rainglasz (talk) 19:18, 31 January 2015 (UTC)

Verified with Museum (where the orignals are on exhibition) that these are not tubes. Picture finally removed. Rainglasz (talk) 16:49, 3 February 2015 (UTC)

Not A Computer

ENIAC was an electronic calculating machine, in no way was it a 'computer', the first actual computers were designed by Kilburn, Blackett and Newman at Manchester, on the Manchester SSEM. Why MUST American's try and grab every first irrespective of the facts??? Twobells (talk) 09:36, 31 August 2014 (UTC)

Someone keeps reverting 'electronic calculating machine' to 'computer', ENIAC was NOT a computer in any sense of the word.Twobells (talk) 09:53, 1 September 2014 (UTC)

That opinion is not the historical consensus, and it does not have any consensus here on Wikipedia. Even in its first iteration, ENIAC was programmable to run a full range of computing problems, and could run them at electronic speeds. ENIAC is referred to widely in the literature as a computer, and even has the word "computer" in the expansion of its acronymic name. From its very inception, it was referred to as a computer. Robert K S (talk) 20:11, 23 March 2015 (UTC)

Z3 part of war effort?

The article says: "The Z3, Colossus and ENIAC were developed independently and in secret as part of each country's war effort in World War II. The Z3 was destroyed by Allied bombing of Berlin in 1944.... For these reasons, histories of computing formerly mentioned only ENIAC and the Harvard Mark I from this period."

Certainly Colossus was part of the war effort but is this true for the Z3? The Z3 article is rather sparse and doesn't mention it. But the Konrad Zuse article says: "Zuse never received the official support that computer pioneers in Allied countries, such as Alan Turing, managed to get. The telephone relays used in his machines were largely collected from discarded stock. A request by his co-worker Helmut Schreyer to the war-time government for federal funding for an electronic successor to the Z3 was denied as 'strategically unimportant'."

Does anyone know which article is correct? Was the Z3 part of the war effort or not? As for "histories of computing (mentioning) only ENIAC and the Harvard Mark I", maybe this statement is just based on English language histories. Is it certain that German language histories of computing mentioned only ENIAC and the Harvard Mark I? Adrian Robson 09:46, 4 December 2006 (UTC)

Question answered on Talk:Z3. Adrian Robson 09:50, 26 January 2007 (UTC)

AFAIK, Zuse received his funding from developing special computers for the Henschel Hs 293, the precursor of today's cruise missiles. Schily (talk) 10:41, 18 August 2015 (UTC)

How to get access source?

How do I get access sources like Goldstine, p. 22? 2001:14ba:fc:d100::1 (talk) 22:59, 30 September 2015 (UTC)

Suggested addition to "Parts on Display"

Hi. I work for the Bradbury Science Museum that is part of Los Alamos National Laboratory here in (surprise) Los Alamos, NM.

The person who is working with our artifacts pointed out that we have a contribution to the "Parts on Display" section.

Here is the language she thought should be included:

“The Bradbury Science Museum in Los Alamos, New Mexico has a data register unit with vacuum tubes.”

Link=http://www.lanl.gov/museum/index.php

I'm including a couple of photos that show it is part of our computing display.

I did not want to edit the page directly and have trouble with COI.

I can be contacted at anderman@lanl.gov 505-665-9196 if there are any questions.

Laanderman (talk) 23:34, 6 November 2015 (UTC)

At the Bradbury Science

Turing completeness

I have a few questions (please give references when answering):

• Was the ENIAC designed to be Turing complete?
• If not, which was the first machine designed to be Turing complete?
• When was it first shown that the ENIAC was Turing complete?

Thanks, --Phrood 13:08, 14 September 2007 (UTC)

I doubt ENIAC designers Mauchly (a physicist) and Eckert (an engineer) even new about Turing's paper which introduced the notion of a universal machine, later called Turing machine. This theoretical paper was published in 1936 in a mathematical journal and it took quite a few years to be known outside of a very specialised community of logicians. On the other hand, it is known that John von Neumann was familiar with this work, but it is not clear whether the design of EDVAC in which he took part was influenced by Turing's ideas. Tsf (talk) 16:22, 14 February 2016 (UTC)

"Betty Holberton (née Snyder) continued on to invent the first sorting algorithm"

I'm not really well versed on the historical details surrounding algorithms, but this sentence immediately read as incorrect to me: "Betty Holberton (née Snyder) continued on to invent the first sorting algorithm".

Sorting algorithms date back to antiquity, so she certainly did not invent the first sorting algorithm. Additionally, the first non-trivial sorting algorithm described academically is Merge sort which is widely credited to Jon von Neumann. Holberton undoubtedly played some significant role related to the history of merge sort, but I don't see any evidence crediting her as the inventor of any particular algorithm.

Her own wiki article currently mentions that "she was one of those who wrote the first generative programming system (SORT/MERGE)" which is verifiable, unlike the unqualified claim currently in the article. I will update this article to replace the unqualified claim with this qualified claim from her page. SloppyG (talk) 15:46, 23 February 2018 (UTC)

First put to work in 1945 or 1946?

An anonymous editor has recently claimed that ENIAC was first put to use in 1945, with a reference that does not support that assertion. What is the correct date (preferably supported by a reliable reference)? --TedColes (talk) 18:40, 16 May 2018 (UTC)

The source claims that this is from a public trial; I haven't looked further into who claims the following in the trial:

"2. The ENIAC machine was constructed by mid-November, 1945.

1. The design for the ENIAC machine was frozen prior to the end of 1944 so that the construction of the machine could be completed as rapidly as possible to confirm the usefulness of electronic computation with such large machines.

2. By mid-1945, the construction of the various ENIAC units, was complete and testing of the completed units was commenced.

3. The ENIAC was placed in operation as a system in mid-November, 1945.

4. Moore School and Army Ordnance representatives considered that the ENIAC machine was being operated rather than tested after December 1, 1945." --Jhertel (talk) 19:21, 16 May 2018 (UTC)

Also, on the same page, which appears to be court findings:

"4. The Los Alamos calculations which commenced December 10, 1945, were the first problem placed on the ENIAC machine. When the first problem was put on the machine, it was the first time that the machine as a whole was being used. It was fully expected that the problem would be solved. It was.

1. The ENIAC machine, and hence any invention claimed in the ENIAC patent, was reduced to practice no later than the date of commencement of the use of the machine for the Los Alamos calculations, December 10, 1945." --Jhertel (talk) 19:28, 16 May 2018 (UTC)

Related with: Honeywell, Inc. v. Sperry Rand Corp.--89.25.210.104 (talk) 22:33, 17 May 2018 (UTC)

The Women who Programmed ENIAC

"Two women operating ENIAC" is all we have to say about the women who programmed ENIAC? The source site for the image (http://ftp.arl.mil/ftp/historic-computers/) captions it as follows:

Two women operating the ENIAC's main control panel while the machine was still located at the Moore School. "U.S. Army Photo" from the archives of the ARL Technical Library. Left: Betty Jennings (Mrs. Bryant) Right: Frances Bilas (Mrs. Spence)

We should at least mention the two best-known ENIAC programmers, Betty (Snyder) Holberton and Jean (Jennings) Bartik.

Betty Holberton http://www.livewirecom.com/columns/97.05.html

Jean Bartik http://inventors.about.com/library/inventors/blbartik.htm

The front lines: History of software begins with the work of some brainy women Petzinger, Thomas Jr; The Wall Street Journal; Nov 15, 1996; B1; http://www.haven.org/~dkap/writings/History_of_computers (wsj.com confirms that an article with this title and author was published, but they don't post archive full-text online)

--Rob* 06:05, 20 July 2005 (UTC)

Check out "The Women of ENIAC" in IEEE Annals of the History of Computing, 18:13-28 (1996), and also http://www.wired.com/news/culture/0,1284,3711,00.html and http://www.witi.com/center/witimuseum/halloffame/1997/eniac.php.

The six women who did most of the actual programming of ENIAC were inducted in 1997 into the Women in Technology International Hall of Fame. They were Kathleen McNulty Mauchly Antonelli, Jean Jennings Bartik, Frances Snyder Holberton, Marlyn Wescoff Meltzer, Frances Bilas Spence and Ruth Lichterman Teitelbaum. (These are their married names, from the 1997 induction.)

-- B.Kell, 1 Dec 2005

I think the most helpful thing to do would be to refer to the ENIAC programmers by their contemporary names, and leave their married names to be elaborated in their individual articles. Otherwise we put ourselves in the position of referring to Kay McNulty as "Kathleen Rita McNulty Mauchly Antonelli," which is a mouthful, or Betty Jean Jennings (who now calls herself Jean Bartik) as something like "Elizabeth Jean Jennings Bartik," which is, in addition to being overlong, also incorrect, as Jean Bartik's birth name was "Betty Jean," not "Elizabeth Jean," contrary to what is printed in one popular source (McCartney). Robert K S 14:55, 2 August 2006 (UTC)

I wonder if the order of the people in the photo is correct. In the New York Times Obituary, Jean Bartik is identified as the person on the right, and Kay McNulty Antonelli as the woman on the left. Which is correct? Alpine Joy (talk) 16:00, 9 April 2011 (UTC)

The pictures in this article are captioned correctly. Your confusion may stem from an error in the NYT article that was subsequently corrected. [1] Robert K S (talk) 02:39, 10 April 2011 (UTC)

I think the women collectively should have its own article. This article, not surprisingly, downplays the significance of these women and their role in the history of technology and gender. JanderVK (talk) 01:09, 28 May 2015 (UTC)

One of the important points about Wikipedia is everyone's opportunity to write new articles and to edit and improve existing articles. You have brought up an issue that should be addressed. I hope that you address all of the issues that you just described by both writing the new article and editing any copy that you feel "...downplays the significance of these women and their role in the history of technology and gender..." Wa3frp (talk) 15:00, 28 May 2015 (UTC)

However it is edited, feminist "us against them" ideology and criticism should be left out. This article's main, if not only, purpose should be to describe a piece of technology; politics and ideology need not enter into it. Alialiac (talk) 02:53, 27 September 2018 (UTC)

...and Calculator, or ... and Computer

I am a Computer Science major from the 90's. I read many books on the history of computers and I dedicated their related acronyms to memory.

Memory serves that ENIAC was an acronym for "Electronic Numerical Integrator And Calculator"... not "and computer" as shown on Wiki. Perhaps an oversight? Perhaps I'm incorrect? —Preceding unsigned comment added by 141.153.57.114 (talk) 19:27, 5 August 2010 (UTC)

This link is from the University of Pennsylvania Archives: http://www.archives.upenn.edu/faids/upd/eniactrial/upd8_10.htmlWa3frp (talk) 19:54, 5 August 2010 (UTC)

Also see the discussion of this above. —Mark Dominus (talk) 22:58, 5 August 2010 (UTC)

Today's reverted edit is about the twentieth time someone wants to substitute "calculator" for "computer". I wonder why this tendency is so strong? GroveGuy (talk) 20:47, 13 September 2010 (UTC)

Feel-good revisionism? — Preceding unsigned comment added by 67.188.20.134 (talk) 01:13, 3 November 2018 (UTC)

Wrong information on tube count?

Under "Contents", the article states that ENIAC contained 20,000 vacuum tubes, but both of the sources that are sited in-line both claim lesser counts, both stating 17,468 or 18,000.

Any thoughts? — Preceding unsigned comment added by 5rockhopper4 (talk • contribs) 00:09, 3 December 2018 (UTC)

Sounds like people like to round up. Change it to "nearly 18000" or "over 17000" or something like that if you like. Dicklyon (talk) 00:47, 3 December 2018 (UTC)