This is an archive of past discussions about Manchester Baby. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
M. Williams, in the reference, gives the information about first running first a simple division program, then the relatively prime program, and then the factor of an integer program. Every other source I've seen (including the website in the external link, computer 50) just mentions the third one, calling it the first stored program to run. It seems logical to me that they would first test just the division routine, wo Williams' description seems reasonable. If that is true, then you could argue that the division routine was actually first. If that is considered too trivial, an arguement could be made that the relatively prime program was about as sophisticated as the largest factor program.
Also, Williams says that the relatively prime program used the division routine. Given the instruction set (mainly a subtraction operation, data moves, branch, and stop), it might make more sense to use the original verison of Euclid's algorithm, which uses subtraction only (instead of division). I wonder if that was done.
Can anyone comment on this? --Bubba73 15:33, 5 Jun 2005 (UTC)
If the division routine ran first, then the division routine was the first program. You can't exclude it on grounds of triviality - if it's a sequence of instructions, it's a program, and if those instructions were executed, then the program ran. You might argue that the coprimeness or factoring programs were the first applications to run, but that's a whole other kettle of fish!
As for the implementation of the coprimality test: a subtraction-based Euclid's algorithm would be simpler to write from scratch, but since they'd written the division routine by that point, i suppose it wasn't that much harder to write a division-based one, and a division-based one will run much faster. That makes the second program another landmark - the first reuse of a software component!
"Then this routine was used in a program to show that 314,159,265 and 217,828,183 are relatively prime".
Hm. Looking at those numbers, the first is int(rnd(pi * 10^7)), and the second would be int(rnd(e * 10^7)) except that the leftmost "1" and the "7" have been transposed. I'm unsure whether the numbers given are correctly quoted, or a typist has fumbled somewhere along the line... would someone mind checking? Kay Dekker01:28, 20 May 2007 (UTC)
"The first small machine worked in June 1948. It's interesting why we built that small machine. We had a cathode ray tube which would store patterns on the CRT store over long periods but it wasn't really proof that the cathode ray tube system would work in a computer, because if at very high speed you write noughts over ones, or ones over noughts, which is what you are doing in a computer constantly, the signals you get from the screen are not balanced and the base line starts heaving up and down. We'd never seen this heaving up and down because we'd never fed it quickly enough but we surmised that it would be there and indeed it was.
So I decided to design some gear which would test this, but after a few weeks (actually I was travelling into Yorkshire at the time in that awful winter of 1947 and I did a lot of design on the train) one of the conclusions I came to was that the only way to test whether the cathode ray tube system would work in a computer was, in fact, to build a computer. So I designed the smallest computer which was a true computer (that is a stored program computer) which I could devise, and we ended up with a one-tube, 32-lines, 8-digit machine. The signals did in fact heave up and down and the design of the amplifier and the clamping system to deal with that was quite an interesting exercise."
I've been asked to do a peer review, of sorts, as a means to aid the article towards WP:GAC. I'm completely green to the content of the article and my experience of computing is Adobe Photoshop and Wikipedia! In this capacity, I've found the following issues:
Why was SSEM nicknamed "Baby", and who by?
"Alan Turing provided the definition of a universal computer". What was it? Is this still accepted as a standard definition? Did anything exist prior to the definition?
Could we mention a little more about what computing existed prior to SSEM, or Turing's definition? Just a sentence or two would help the casual reader (like me), IMHO.
Without needing to navigate away from the article, who was John von Neumann?
The second paragraph of the Background section is unsourced. For example, whose POV is it that EDVAC had drawbacks?
Could the Development section have a little more context for the techno-jargon challenged? For example, explain how much computing power a "bit" had in terms familliar to casual readers, or say something like "the SSEM had a single operand architecture.... meaning that XYZ..."
Citation is missing for the "seven instructions".
First program is a single paragraph section, which is generally discouraged by MOS.
In Later developments, we have the line "The SSEM developed into the Manchester Mark I, which led to the Ferranti Mark I". Did it develop sentiently? If not (!), who developed it and why?
In Later developments, I think the table needs re-adjusting so that it reads as a comparison with SSEM, and makes that clearer.
Other than that, not a bad article. Quite some way off a "good" GA I think though. The pictures are good, but in short, what is lacking is context; explain what the jargon means and who the people are at every opportunity, it'll make for a much stronger article. Hope that helps. --Jza84 | Talk 21:33, 2 November 2008 (UTC)
F&f's critique
Background
"During the Second World War, researchers working on the problem of removing the clutter from radar signals had developed a form of delay line memory, the first practical application of which was the mercury delay line, developed by J. Presper Eckert. The idea was to eliminate the radar reflection from static objects by delaying each returned pulse in a delay line for the time between transmitted pulses and then to compare the returned pulse with the stored pulse, leaving a residual signal which contained only the images of any moving objects."
The explanation should be improved. "The idea was" is vague; it is better to be more explicit.
No mention is made anywhere that mercury is the medium through which the acoustic waves (corresponding to the returned radar pulses) travel. What is the "stored pulse?" It is not explained.
"delaying each returned pulse" seems to be incorrect in light of what is written in delay line memory. It is only "half of each returned pulse;" the other half is sent directly to the receiver.
The inversion of the delayed pulse to (algebraically) eliminate the direct pulses from static objects is not explained. ("... compare the returned pulse with the stored pulse ..." is confusing.) More explanation may add more text, but it is imperative that the reader be on board at this early stage, otherwise, s/he will stop reading. In any case, it is not already a lengthy article, so a little extra will not hurt it.
Reply: Mercury was not always the medium used in delay lines, it was just one of the materials used. I've added a very brief overview of how radar works so as to explain the idea of "pulses" and that each received pulse was compared with its predecessor, which would have been stored in the delay line. The material in the delay line memory article is unsourced, so I can't speak for its accuracy. The explanation I've given in this article comes from Brown (1999), A Radar History of World War II: Technical and Military Imperatives. --MalleusFatuorum02:02, 19 March 2009 (UTC)
"In 1945, scientists within the Ministry of Supply decided that Britain needed a National Mathematical Laboratory to coordinate machine-aided computation,[1] and so a Mathematics Division was set up at the National Physical Laboratory (NPL)."
"decided" is probably not the best word here. Perhaps "were persuaded" or even "concluded?"
We need a transition sentence before the sentence to tell us how this is connected to the Manchester machine, otherwise the reader begins to wonder why all this is being described.
Reply: I've swapped "decided" for "concluded". Turing is partly the common thread through the section, leading up to Williams being allocated his circuit technicians and joining Manchester University. I've had a go at adding a linking sentence though, to make the change of emphasis between the first and second halves of the section seem less abrupt. The whole thing is really just trying to paint a picture of the context in which the SSEM was developed though. --MalleusFatuorum03:28, 19 March 2009 (UTC)
"On 19 February 1946, Alan Turing presented the design for an electronic stored-program computer, to be known as the Automatic Computing Engine (ACE)."
Doesn't say where. Was it at the math division at NPL? Or was it in a paper?
"The NPL did not have the expertise to build such a machine however, and so they contacted Tommy Flowers at the General Post Office's (GPO) Dollis Hill Research Laboratory."
Needs to be tighter. How about, "The NPL, which did not have the expertise to build such a machine, in turn contacted ... ?"
Do we need the abbreviation, GPO? It is used only once later. Too many abbreviations can be confusing to the reader. I feel that unless a full name is extremely long (as that for ENIAC), an abbreviation should be used only if it will be needed at least twice later.
Reply: Tightened up the sentence much as you suggested. I see no harm in keeping the GPO abbreviation, it's how most people here in the UK would know the precursor to BT and the Post Office. "General Post Office racks" doesn't seem like a very attractive alternative. --MalleusFatuorum03:28, 19 March 2009 (UTC)
"Flowers had designed Colossus Colossus, the world's first programmable electronic computer, but other commitments meant that he was unable to take part in the project, although his team did build some mercury delay lines for ACE."
This too needs to be tighter. How about, "Flowers, who had designed Colossus Colossus, the world's first programmable computer, had other commitments and was unable to take part in the project; his team, however, did build some mercury delay lines for ACE."
The reader now becomes aware that ACE had mercury delay lines. So, either this fact should be first mentioned in the sentence about ACE above, or not be mentioned at all here.
The number of different names ENIAC, Colossus Colossus, Manchester Mark I, Ferranti Mark I, ... are now beginning to add up, and each is the world's first something or other. It might be a good to have a diagram or table, which has the year of first operation of each computer, its name and "domicile," and its claim to fame. It would certainly help the reader.
Reply: There's such a table in History of computing hardware. I'd be very reluctant to duplicate it here, especially in the Background section. I've tightened up the first sentence, much as you suggest. I don't think it's particularly important that the ACE used mercury delay lines, most computers of the period did; the fact is only mentioned in passing to demonstrate that Flowers' team did do some work on ACE even though he himself didn't. --MalleusFatuorum03:28, 19 March 2009 (UTC)
Here the syntax is not the best; the reader has to make sense of the punctuation to clarify the sentence (i.e. that MW was not at both places). Better to say, "MF at UCML was approached, as was the TRE."
If MW will be playing a future role in the Manchester machine, then it should be indicated here ("MF, who was later to play ..., was also approached ..."); otherwise, it seems like a random comment to the reader. Same for TRE.
Reply: I've switched the order of this sentence as well as rewriting it. The important player here is TRE, not Wilkes, as Williams was at that time working for the TRE. Wilkes played no part in the SSEM's development, he's only included here for completeness. --MalleusFatuorum03:51, 19 March 2009 (UTC)
The sentence is confusing because we have been told that TRE was only approached. It seems now that they had already begun to work.
Why do we need to know about DSIR? As far as I can tell, it never appears again in the article. If it needs to be mentioned, why not simply say, "NPL's parent organization" without adding another abbreviation? ("ran the NPL" is vague)
Reply:' The sentence says that the TRE was already doing work for DSIR, not that that work had anything to do with ACE, which it didn't. I've rewritten the offending sentence to say: "The government department responsible for the NPL ...". --MalleusFatuorum03:28, 19 March 2009 (UTC)
"This led to the superintendent of TRE's Physics Division visiting NPL on 22 November 1946, along with Frederic C. Williams, and A. M. Uttley."
What is "This?" How about something like, "NPL's decision to make ACE a major focus of its collaboration with the TRE led to a visit by ...."
Were Williams and Uttley also at TRE? Their affiliation should be mentioned.
"Most of Williams' circuit technicians were being transferred to the Department of Atomic Energy, and he was leaving six weeks later to take up a professorship at the University of Manchester."
Here again the cause comes after the effect. Better to say, "William had already accepted a professorship at WUM and most of his circuit technicians were in the process of being transferred ..." Don't really need to know "six weeks later."
At this point the reader becomes aware that Williams is some kind of big wig in all this. Something about his future role should have been intimated before we went into the history (to provide some direction for the reader).
Reply: I've added an explanation that Williams led a group at TRE developing CRT stores as an alternative to delay lines, and removed the six-week timescale. Williams was introduced in the lead as one of the main people behind the SSEM, and the following section on the Williams tube makes clear exactly what his primary contribution was. --MalleusFatuorum15:56, 19 March 2009 (UTC)
"The TRE therefore agreed to second a small number of staff to the university to work under Williams' direction, and that others would work with Uttley at TRE."
"second?" This usage of "second" (as tr. verb) is not that widespread (at least in my experience); would "support" work?
Reply: "Second" is quite correct and appropriate here, and it is widely used in this sense. "Support" doesn't have the same meaning. Someone who is "seconded" is working at one establishment like any other employee, but is actually an employee of another establishment, to which they will return at the end of their secondment. --MalleusFatuorum02:02, 19 March 2009 (UTC)
"small number of staff?" why not simply "support a small staff of circuit technicians," since those are the ones that were staying behind?
Which others? The ones that were to be transferred to DAE? Are we really saying that the technicians were not transferred to DAE, that some went with Williams to Manchester and others stayed back at TRE to work with Utley? If so, all this should be clearly stated.
Reply: I've made a few changes along the lines suggested to clarify that of the the technicians scheduled to be transferred to the DAE, some were transferred, some were were seconded to Manchester University, and some stayed behind to work with Uttley. --MalleusFatuorum14:35, 19 March 2009 (UTC)
"The NPL also sent a proposal and draft contract to both Williams and Wilkes to work on ACE."
Is this really needed? If it is important to the development of the project, then something should be mentioned here about what role either the proposal or draft contract would play.
I am surprised that the table 'Defining characteristics of some early digital computers of the 1940s' from the History of computing hardware has been removed. It is transcluded in the articles for almost all the other machines that it cross-references. Am I alone in thinking that it provides useful contextual information for those who want to use the encyclopedia to clarify what is meant by the first computer, a word that is most justifiably used in the lead of this article? --TedColes (talk) 07:23, 26 May 2009 (UTC)
I don't see that the table adds much, if anything at all of value to this article, which is about the SSEM, not the history of computing. Anyone who wants a wider historical overview can very easily click on the link to the main article. The lead does not in any case describe the SSEM as the first computer, which it most certainly wasn't, it describes it as the first stored-program computer. --MalleusFatuorum13:47, 26 May 2009 (UTC)
"Although early computers such as CSIRAC made successful use of mercury delay line memory,the technology had several drawbacks: the amount of data it could store was limited, it was heavy, and it was expensive."
The "it" becomes a little ambiguous in the second and third independent clauses. I would probably favor: "it could store limited data, it was heavy, and it was expensive."
Doesn't all memory ultimately store limited data? Can we be a little more specific?
Also, how heavy and expensive was it? Are any numbers available? If they are they would certainly help in highlighting (by comparison) the CRT memory.
Reply:Mercury memory of UNIVAC I (1951) You're right, all memory is ultimately limited, so I've tightened that bit up. I don'thave any figures for price or weight, but this pic to the right should give some idea of the scale of the beast. I would have included that in the article if I'd also been able to get a decent image of a Williams tube, but the licensing on the pic in its article looks decidely dodgey, and wouldn't get past FAC reviewers. Next time I'm at the museum I may try to get a shot of one of the replica tubes. This article would have been so much easier if the supporting articles had been in better shape. Still, I suppose we have to start somewhere. --MalleusFatuorum22:05, 20 March 2009 (UTC)
"In addition, because data was stored as a sequence of acoustic waves, and because the velocity of sound through a medium varies with temperature, the mercury had to be maintained at a constant temperature."
Do you think it might be clearer if we said, "In addition, since data was stored as a sequence of acoustic waves whose velocity, in turn, varied with the temperature of the medium of propagation, the medium, mercury, had to be maintained at a constant temperature."?
Why was it such a problem to maintain the medium at a constant temperature? Air-conditioners were around in 1946.
Reply: I've somewhat reworded that sentence along the lines you suggest. The mercury had to be maintained at a constant temperature of 40° C, a bit more than most AC units would be able to achieve even today. --MalleusFatuorum22:05, 20 March 2009 (UTC)
"The Manchester Small-Scale Experimental Machine (SSEM) was designed to prove that the Williams tube, the first random-access digital storage device, could operate at computer speeds."
Could "computer speeds" be clarified or wikilinked? (This, of course, does become clearer in later sections when we see the evidence of the speed of the computation.)
I wonder if this also might be the place to add one sentence clarifying the difference between William tube's RAM and the previous serial memory. A reader might wonder that since the speed of sound in mercury is quite fast (1450 m/s), it would make it seem almost RAM-like. So, how was the new memory that much better (or something to that effect).
Development and design
You say that by August 1947, the memory capacity had increased to 2,048 bits or a 64x32 bit array. Wouldn't this correspond to two 32x32 bit words? Later, you say that the memory was able to hold 32 32x32 bit words. When did the memory capacity increase from 2,048 bits to 32,768 bits (32x32x32)? Or did I misinterpret this?
First programs
Am I correct in assuming that starting with and in descending order, each divisor was repeatedly subtracted until the remainder either became 0 or a negative number?
Reply: Not quite. To find the highest factor of the program tried every integer from downwards, in the way that you describe. was the largest number that the SSEM's memory could store, but obviously a number larger than the one we're looking for the highest factor of couldn't possibly be a factor. So the program tried every integer between and , at which point, as you say, the remainder from the repeated subtractions was zero and the program stopped. --MalleusFatuorum22:18, 20 March 2009 (UTC)
Thanks for your very detailed review, which has undoubtedly improved the article in some important areas, and hopefully gone at least some way to making it more accessible to the general reader. --MalleusFatuorum22:23, 20 March 2009 (UTC)
References
^Lavington (1980), Chapter 5: "ACE, the 'British National Computer'".
ABC?
Since this page is discussing electronic digital computers (and which one was the first), I thought it prudent to provide the following link.
Since this article deals with a computer built in the UK, I don't know if this link is 100% relevant, but it maybe the basis for a future edit if this becomes a topic of interest.
I want to let the article and secondary links speak for themselves so I won't quote anything here, so read it and decide for yourself.
That's obviously interesting stuff, but I'm not certain it's relevant here. This article isn't discussing which was the first electronic digital computer, it's specifically describing the SSEM, which is undoubtedly the world's first stored-program electronic digital computer. --MalleusFatuorum12:52, 30 May 2009 (UTC)
Wrong information about the Z3
"Programmable computers such as the Z3 and Colossus were subsequently developed during the 1940s, but their programs were generally represented by patch cables and switches."
This is definitively wrong regarding the Z3. The Z3 was the world's first working programmable, fully automatic computer. Its programs were stored externally on tape, so the Manchester Small-Scale Experimental Machine can claim the improvement of internal program storage. - I'll try to correct this in the article. Cheers, MikeZ (talk) 09:36, 30 May 2009 (UTC)
Ah, I see from your comment below that you're using the term "stored-program" in a rather specialised sense in distinguishing between internally and externally stored programs. Forget I spoke. :-) --MalleusFatuorum12:56, 30 May 2009 (UTC)
first stored-program computer… considered "small and primitive" compared to its contemporaries
(cur) (prev) 16:14, 30 May 2009 Malleus Fatuorum (talk | contribs) (26,375 bytes) (its contemporaries were under construction; the SSEM was the first to be completed) (undo)
(cur) (prev) 16:28, 30 May 2009 Crunchy Numbers (talk | contribs) (26,343 bytes) (so its future contemporaries were under construction... at the time who knew what would happen?) (undo)
Malleus Fatuorum># everyone knew, please read the text about ACE and EDSAC and so on)
I did read the referenced web page at this section.
It said that from June 1948 to October 1949 (when EDSAC was fully functional) this was the only working stored-program computer. As a matter of logic and language usage doesn't it follow that something can't be both "the only" and have contemporaries at the same time?
We could say that it was small and primitive compared to contemporary designs, to competing designs. But this was really a test bed right? Not intended to be the desired design.
Also when I asked who knew what would happen I meant there could have been budget cuts or a crisis and perhaps non of the "contemporaries" would have been completed.
I don't claim to be an expert on this topic but I am a good editor. I tend to work on articles where contributions are welcomed. I also do some vandalism reverts from time to time. You seem to have things under control here and there are other articles where I can work. It was just a shame to see a mistake, even if minor, on a featured article. -Crunchy Numbers (talk) 17:20, 30 May 2009 (UTC)
Well, other stored-program electronic computers were certainly "coexistent in time" with the SSEM, which just happened to the first one to be successfully demonstrated working. Indeed as the SSEM was incrementally morphed into the Manchester Mark 1 it would be difficult to say when one became the other. Nevertheless the present wording appears to be open to different interpretations, so I've rewritten the offending sentence: "Although considered "small and primitive" by the standards of its time, it was the first working machine to contain all of the elements essential to a modern electronic computer." --MalleusFatuorum18:45, 30 May 2009 (UTC)
Pic removed
An artistic representation of a Turing machine
I removed this picture from the article because I found it did not add anything to the reader's understanding of the Manchester Small-Scale Experimental Machine, or of Turing machines. The parts are not labeled. I don't know why there's an aerodynamic mousetrap in the middle of it. It's likely to make some readers think that it's not a metaphor and that it's literal. Tempshill (talk) 06:53, 23 May 2009 (UTC)
I agree but possibly for different reasons.
This picture is not directly related to the SSEM. The coverage of Turing machines is background to this article.
It is a poor picture of a Turing machine - it is purely abstract and does not illustrate the concepts involved even for a Turing machine.
It does not aid understanding of the SSEM at all.
As such it is merely decoration. An "artistic" representation has no place here. This is not an article on the arts and this is an encyclopaedia, not an art site.
If there is no good reason for a picture there then why do we have a picture? It does not back up the points made in the article or introduce new information. It is just clutter that serves to distract from the main focus of the article. CrispMuncher (talk) 21:26, 23 May 2009 (UTC).
I have no objection to this image being replaced with a better one, but I strongly object to its removal. The image was taken from here in response to comments made at this article's FAC. Let's also all remind ourselves that there never was and never will be a Turing machine, so any image is inevitably an artist's impression. --MalleusFatuorum22:08, 23 May 2009 (UTC)
Do you believe that this illustration aids the reader's understanding of a Turing machine, let alone the SSEM? I am surprised if so. If it doesn't, it doesn't belong in the article. Tempshill (talk) 05:30, 1 June 2009 (UTC)
That can't be right. What is the original? —Preceding unsigned comment added by 66.72.196.26 (talk) 04:43, 30 May 2009 (UTC)
--MalleusFatuorum 13:23, 22 June 2009 (UTC)
I realize it's a replica, but that blue-screen Hewlett-Packard oscilloscope sitting on the storage rack is an anachronism. --Mensanator (talk) 16:34, 21 June 2009 (UTC)
So are you suggesting that the replica machine does not incorporate an oscilloscope, as is clearly shown that it does on the photograph? An oscilloscope belongs to the time the replica was built, wouldn't you agree?. --MalleusFatuorum21:19, 21 June 2009 (UTC)
Be that as it may, it still looks out of place, like a jet amongst a display of replica WWI biplanes. Or, as in the Museum of the Air Force at Wright Patterson Field, a biplane amongst the jets (they were re-arranging the displays and hadn't gotten around to unhooking the biplane from the ceiling). Looks amusing. --Mensanator (talk) 06:09, 22 June 2009 (UTC)
I don't think that's a good analogy at all; it's more like a modern spanner being used in the maintenance of a WWI biplane. Oscilloscope's predate the SSEM of course, but insisting that a replica built in 1998 and being maintained in 2009 ought to be worked on only with a period oscilloscope seems a little unrealistic. --MalleusFatuorum13:24, 22 June 2009 (UTC)
Yeah, I suppose you're right. Still, a period scope would have imparted a nice "this is what we had to work with back then" feel, but that's probably beyond the scope (no pun intended) of the project. And I, for one, wouldn't want to have to work with one. And who else even notices things like this? It was just a first impression, so never mind. Wouldn't mind visiting it someday, although I picture my sister gnawing her own leg off as I try to explain the significance. --Mensanator (talk) 00:19, 23 June 2009 (UTC)
Arithmetic operations
I do not agree that unary operations such as negation are arithmetic operations; the essence of arithmetic operations is that they are binary. However it seems that I am outnumbered in that view, so I shall simply have to bite my tongue, grin and bear it, and all that jazz. "Subtraction from zero", for instance, is a binary operation, not unary, but that's not what negation means, especially in the context of the SSEM, or indeed any other computer, where it simply involves bit flipping. --MalleusFatuorum22:20, 16 June 2009 (UTC)
Let me restate my case more clearly; which of the SSEM's seven instructions allows for any arithmetic operation other than subtraction? --MalleusFatuorum23:16, 16 June 2009 (UTC)
The technical explanation is that the term "operator" in the context of computer programming is a subset of the more broad mathematical concept of "operator", which is essentially an n-ary function that maps one or more operands to a single value. The number of operands that a given operator takes is its arity. So yes, unary operators are bona fide arithmetic operators. To answer your question about the SSEM, the LDN and SUB instructions are therefore the arithmetic instructions. And to be especially pedantic, all CPU operations are in reality nothing more than "bit flipping". — Loadmaster (talk) 00:09, 24 June 2009 (UTC)
I don't recall asking any questions about the SSEM, but have it your own way. Make whatever changes you like. I've done what I can with this article and I'll leave it to you now. --MalleusFatuorum00:43, 24 June 2009 (UTC)
The First Program
An earlier discussion here reflected on what was truly the first program. I have the benefit of a copy of Geoff Tootill's notebook of the time. The first program, written by Tom Kilburn, found the highest factor of a number by successive subtractions. Geoff Tootill had written a program to find the HCF of two numbers, and it was chance that it was Kilburn's program which was first to run. About 3 weeks later Alan Turing submitted a long division routine which itself occupied all 32 words of the store. See the paper in the Annals of the History of Computing, "Early Programs on the Manchester Mark I Prototype", Jul-Sep 1998, Vol 20, No 3, p.4. (The term "Mark I Prototype" in the above title is wrong with hindsight, we should have said "Small-Scale Experimental Machine".)
I've modified the implicit register used in the definition of JRP:
JRP: Jump to the instruction at the specified memory address plus the accumulator (relative unconditional jump)
become
JRP: Jump to the instruction at the specified memory address plus the program counter (relative unconditional jump)
which seems to be more realistic and does corespond to Napper 2000, p. 367. harvnb error: no target: CITEREFNapper2000 (help).
(program counter can be changed into instruction adress)
But, seeing the 'original program', I've got a doubt on the formulation of JMP/JRP. I think it should be:
JMP: Jump to the instruction at the adress obtained from the specified memory address (absolute unconditional jump)
JRP: Jump to the instruction at the program counter plus the relative value obtained from the specified memory address (relative unconditional jump)
(a double indirection system due to the fact that the SSEM always get its value from the memory, cf. LDN/SUB with direct adress)
And more, in the 'original program', there is a shift from the value stored at the adress and the target (+1). Ex: Jump back 2 lines is done with JRP 21, and Mem[21] coresponds to -3. This is probably due to some automatic PC<-PC+1, so JMP/JRP could/should be:
JMP: Jump to the instruction at the adress obtained from the specified memory address plus 1 (absolute unconditional jump)
JRP: Jump to the instruction at the program counter plus 1 plus the relative value obtained from the specified memory address (relative unconditional jump).
Reading p.70 Barbacci, M. R. 1979. Instruction set processor specifications for simulation, evaluation, and synthesis. In Proceedings of the 16th Design Automation Conference (San Diego, CA, United States, June 25 - 27, 1979). Annual ACM IEEE Design Automation Conference. IEEE Press, Piscataway, NJ, 64-72. I think the 1st change was ok, and 2nd is necessary, the 3d needs to be formulated (maybe just a footnote for JMP and JRP to say that the program counter was increment after. As a consequence, the adress stored in memory was the target adress -1) BDenis.--129.88.133.117 (talk) 09:52, 17 May 2010 (UTC)
First program source code
Is there a record of what the actual program code for the first program(s) was? A 17-instruction program is short enough to be shown in its totality in the article, and would be very interesting from a historical computer science perspective. Or has the original code been lost to the sands of time? — Loadmaster (talk) 18:18, 16 June 2009 (UTC)
The original program has been lost, but it has been reconstructed with the help of the original team from a modified version written shortly afterwards found in, IIRC, Tootill's notebook. Lavington includes a copy in an appendix called "The 'real' first program". It's actually 25 lines long, as it includes the working storage area as well. It looks pretty strange to a modern eye though, as of course it doesn't use assembler mnemonics, as there were none, so it would need quite a bit of explanation. To give you a flavour, here are the first four lines:
-18,C // Clear accumulator
-19,C // load +a
Sub 20 // trial substraction
Test // is difference -ve?
To be perfectly honest I'm not certain this would be a valuable addition to this article, but now you've raised the issue in my mind I'm wondering whether the First computer program wouldn't be a worthwhile article in its own right. What do you think? --MalleusFatuorum20:54, 16 June 2009 (UTC)
Perhaps that's a better idea ... Computer programming sounds like a good potential candidate, but there's quite a bit that could be said about the first computer program, not least it was the only program Kilburn ever wrote ... have to think about it more. --MalleusFatuorum22:02, 16 June 2009 (UTC)
The first program is missing. Could we, please, have -here- the code of the original program with the asm syntax used in this article and some explanation (n.b. the syntax used in this article seems clear but quite different from the 'original' syntax [which is not as clean]. Which syntax have to be chosen? Is there any real 'original' syntax? Did compiler/assembler already exist in 1948? Was the 'original' program' written with this syntax (i.e. with mnemonic) or only with binary code?). BDenis--86.206.30.123 (talk) 11:59, 14 May 2010 (UTC)
The four lines above could be (in the cleaner syntax used in the article):
LDN 18 // Clear accumulator
LDN 19 // load +a
SUB 20 // trial substraction
CMP // is difference -ve?
what about a modern version of the program (from the resurection paper, with the syntax used in the article, and symbolic adress):
1 LDN Zero \ (re)set init value for Acc
2 Div: LDN MinusA / with some integer A
3 SUB B \
4 CMP | loop Acc -= B until Acc <0 (search for rest of Acc/B)
5 JRP Back2 / will exit with Acc = (A % B) - B
6 SUB MinusB \
7 STO Remainder / compute real Remainder = A % B
8 LDN MinusB \
9 SUB One | prepare next division : B--
10 STO B | (set B and MinusB)
11 LDN B |
12 STO MinusB /
13 LDN Remainder \
14 CMP | test if remainder==0 (if remainder<0, and -remainder<0 then remainder==0)
15 JMP ptrEnd | if yes, then it's finished (Remainder is the greatest proper divisor (-1) of A)
16 JMP ptrDiv | else restart a new division
17 End: STP /
18 Zero: 0
19 MinusA: -262 144 | initial value -A (A any integer)
20 B: 262 143 | initial value : A-1, will change, decreasing
21 Back2: -3
22 MinusB: -262 143 | initial value : -(A-1), will change, increasing
23 ptrDiv: One: 1 | Div-1
24 Remainder: 0 | any initial value, , will change with the remainder A % B
25 ptrEnd: 16 | End-1
It's interesting, the 7 instructions of the instruction set are used, even if JMP/JRP could have been expressed with only one of the two forms.
I would need some explanation about line 1, and line 18 (I've seen some explanation, but it didnt convince me), maybe the reason is below (about line 23? but I'm not sure, maybe it's just some kind of a separator between program and data)
A propos, I've been surprised about line 23 (value 1): it is referenced twice with different meanings
as adress line 16 (for begin of division)
as value line 9 (for decrement)
As a conclusion line 23 is difficult to translate. It could have been:
23 ptrDiv,One: Div-1 | 1
24 Remainder: 0 | any initial value, , will change with the remainder A % B
25 ptrEnd: End-1 | 16
(see the coherent change line 25 also)
Maybe, the 'original' should be put side by side to compare and understand the logic.
I have two concerns. The first is that, as I'm sure you're aware, nobody knows what the first program really looked like, as nobody bothered to keep a copy. All we have is the best guesses of some of those who were involved in the project, many years later. I'm not sure it's even known for certain whether this version has ever been run. Secondly, if we make a modern mnemonic version then we're slap bang into original research territory, which is strictly verbotten here in wikipedia. All we can do is to find a modern version someone else (reliable) has published and use that. It's often frustrating, I know, but those are the rules. MalleusFatuorum16:52, 14 May 2010 (UTC)
PS. There were no compilers, assembers, or mnemonics in 1948; all the programming was done by setting binary values into each memory location in turn, using switches. MalleusFatuorum16:55, 14 May 2010 (UTC)
Second try
maybe something closer to the original (with mnemo from the wikipedia article) without my comments and symbolic adress.
there is no original work problem (it's just a translation of the 'original program' with the langage used in the wikipedia article, maybe this exists already in some paper)
the program is really missing in the article
the question is which form is the best (the first try, or the second, or another [or binary code, or the text from the resurection article]?)
because, it's not my program, not my language, not my values, what is my contribution? what is the original part? I'm sure, it's not research work, revolutionary ideas, personal position. It's clearly easily verifyable. BDenis.--86.206.30.123 (talk) 14:29, 17 May 2010 (UTC)
Yes, really. You are translating the original program into a form using mnemonics that did not exist when the program was written. That's analagous to a translation from, say, Polish to English. If such a translation were to be included in a featured article, the reliability of the translation would need to be established. Have you reason to believe that your translation would be accepted just because you claim it to be an accurate representation? I don't. MalleusFatuorum14:49, 17 May 2010 (UTC)
Do you compare Polish with the SSEM's instruction set (only 7 words), or English? I've added the original notation in the table describing the SSEM's instruction set.
Jump to the instruction at the adress obtained from the specified memory address S[t 1] (absolute unconditional jump)
100
Add S, Cl
JRP S
Jump to the instruction at the program counter plus (+) the relative value obtained from the specified memory address S[t 2] (relative unconditional jump)
010
-S, C
LDN S
Take the number from the specified memory address S, negate it, and load it into the accumulator
110
c, S
STO S
Store the number in the accumulator to the specified memory address S
Subtract the number at the specified memory address S from the value in accumulator, and store the result in the accumulator
011
Test
CMP
Skip next instruction if the accumulator contains a negative value
111
Stop
STP
Stop
^As the program counter was incremented at the end of the decoding process, the stored adress had to be the target adress-1.
^As the program counter was incremented at the end of the decoding process, the stored adress had to be the target adress-1.
^The function bits were only partially decoded, to save on logic elements.[1]
As you can see, the translation (second try) is straightforward. (I admit, the 'first try' is a bit different)
How could I convince you that the second try is not an original work? Take the 17 lines original program (from the resurrection article), for each line find in the table the corespondance between original notation and modern notation (from Lavington, ref 27), and you will obtained the same result (second try).
Unfortunatly, I didn't find the article from Lavington (ref 27), but I will not be surprised if the program is in it.
About the form, should we put back the original comment (I think so)
You mention the possibility of writing an article First Computer Program, I did'nt find any entry about it. Was the idea abandonned? I'm not a specialist of the history of computer/algorithm, so I don't know much about other first computer program, but the idea is very appealing (and, for the ssem first program, I may give some help ...) Do you think it would be better place to give the ssem first program? (I think, The first program is sadly missing from the article.[computer scientist point of view])
What about the two other programs for the SSEM? (long division, higher common factor) I didn't find the code. Are they available? (A propos, in the section 'the first program' of the discussion, it is said that Tootill's program was HCF, but in the article it is said that it was an amended version of Kilburn's program. Where is the truth? Idem about the timeline, it's not completely clear, in the article it is said 'following month' between first and second program, and 'sept' is mentioned about Turing, in the discussion it is said '3 weeks' between first and third program. Does 'following month' mean the first days of july? Can we be more precise? About 'Sept', it is confusing, can it be improved?)
It's original work because it's you that's doing it. You need to find a reliable source that's published a version using modern mnemonics. BTW, I like your addition of the original notation to the table, nice. I've never seen listings of the other two programs written for the SSEM, and I suspect that they haven't survived. From what I know of Turing, I'd also suspect that he just wrote his program down in binary. I've got no objection at all if you want to add the source code for the first program to the article, in fact so long as it's properly sourced I think that's a good idea. I took my copy of the Lavington book, where a lot of this material came from, back to the library, so I can't help with the chronology you're asking about. Do you have access to a copy? MalleusFatuorum15:23, 18 May 2010 (UTC)
PS. As you can see, there was some discussion about whether the first program should be included here, or whether there would be enough material to justify it having its own article. My (slight) preference was to have a separate article for it. The idea wasn't so much abandoned as stalled. MalleusFatuorum15:27, 18 May 2010 (UTC)
third try
with original comments
1 LDN 18 clear accumulator
2 LDN 19 load +a
3 SUB 20 Trial subtraction
4 CMP is difference negative?
5 JRP 21 still positive, Jump back two lines
6 SUB 22 overshot, so add back bn
7 STO 24 store +r n
8 LDN 22 load bn
9 SUB 23 form b(n+1) = bn - 1
10 STO 20 store b(n+1)
11 LDN 20 load -b(n+1)
12 STO 22 store it
13 LDN 24 load -r n
14 CMP is remainder zero?
15 JMP 25 yes, Jump to line 17
16 JMP 23 no, Jump to line 2
17 STP
18 0
19 -262 144 -a
20 262 143 b
21 -3
22 -262 143 -b
23 1
24 0 r
25 16
If Lavington's book does not contain an authoratative copy of the program, might it be worth writing a letter to the Editor of 'Resurrection: the Bulletin of the Computer Conservation Society' [1] asking readers whether they think that the program (as laid out in 'third try', and with a copy of the instruction set) is likely to be a reasonable representation of the actual program? If it was then published, it could be cited, and any endoresment from readers would add weight. --TedColes (talk) 06:01, 19 May 2010 (UTC)
That might work. Lavington does, of course, have an "authoratitive" copy of the program, but not one using modern mnemonics. MalleusFatuorum11:42, 19 May 2010 (UTC)
Fourth try
To be sure the third try is the best one, here is a fourth try (and I'll work on a C-Like version, not to compare, but to propose something easier to understand)
I don't think a binary version would be better (even if it would be closer to what have really exist)
the program:
Original Modern Original
notation notation comment
1 -18, C LDN 18 clear accumulator
2 -19, C LDN 19 load +a
3 Sub 20 SUB 20 Trial subtraction
4 Test CMP is difference negative?
5 Add 21, Cl JRP 21 still positive, Jump back two lines
6 SUB 22 SUB 22 overshot, so add back bn
7 c, 24 STO 24 store +r n
8 -22, C LDN 22 load bn
9 SUB 23 SUB 23 form b(n+1) = bn - 1
10 c, 20 STO 20 store b(n+1)
11 -20, C LDN 20 load -b(n+1)
12 c, 22 STO 22 store it
13 -24, C LDN 24 load -r n
14 Test CMP is remainder zero?
15 25, Cl JMP 25 yes, Jump to line 17
16 23, Cl JMP 23 no, Jump to line 2
17 Stop STP load -b(n+1)
and the data:
18 0
19 -262 144 -a
20 262 143 b
21 -3
22 -262 143 -b
23 1
24 0 r
25 16
Maybe better ... (even better than asm for modern readers)
(the only bad thing : minusA/minusB ... I would like to write -A/-B but it is not legal C idf)
int minusA = -262144;
int B = 262143;
int minusB = -262143;
int Rem;
int Acc;
Acc = - 0;
division: Acc = - minusA ;
substraction: Acc = Acc - B;
if (Acc>=0)
goto substraction;
Acc = Acc - minusB;
Rem = Acc;
Acc = - minusB;
Acc = Acc - 1;
B = Acc;
Acc = - B;
minusB = Acc;
Acc = - Rem;
if (Acc>=0)
goto end;
goto division;
end: /* end */
For the record, Bdenis has been doing the same thing on fr:, has been told the same things as here, i.e. that copying the program is a copyright violation and that adapting it is original research. His reaction there is the same as here, namely ignoring whatever he is being told. Now he's back on fr:, and citing the changes here as a reason why they should be done there. Oyp (talk) 19:59, 26 May 2010 (UTC)
I agree, and I've removed it. The citation suggested that this modern version came from the Resurrection, but it didn't. I had thought that we'd sorted this out already, above. MalleusFatuorum01:08, 27 May 2010 (UTC)
Sorry, I do not understand, you said 'That might work', so I thought it was ok.
About 'copyright violation', you said 'Lavington does, of course, have an "authoratitive" copy of the program, but not one using modern mnemonics.'
About orginal work, I've thougth you have understood that it was not an original one (and in consequence, 'That might work').
About the citation, maybe my formulation is not correct, I do not want to say that the modern version came from the Resurrection. it does not. It is the orginal program and the comments from the authors which come from the Resurrection. The modern version is just a technical modern transcription of that orginal program with the modern notation from Lavington. (I'm not sure, the 'fr' admin have seen that it was not the original program which was added, but a modern one...)
So what is the conclusion?
ps: about 'en'/'fr' as the source is here, I thought it was better to discuss here.
When I said "that might work", it was in response to the suggestion made by TedColes above: " ... might it be worth writing a letter to the Editor of 'Resurrection: the Bulletin of the Computer Conservation Society' asking readers whether they think that the program (as laid out in 'third try', and with a copy of the instruction set) is likely to be a reasonable representation of the actual program? If it was then published, it could be cited, and any endoresment from readers would add weight." The very clear conclusion is that your version using modern mnemonics is original research. MalleusFatuorum14:15, 27 May 2010 (UTC)
One of my first lectures @ Manchester (1996) was by ( as I recall) Tom Kilburn about the Baby, Mark I and early computers, in it he said that one of the problems they had was that due to the departments location on Oxford Road in manchester and the CRT memory units, whenever a badly maintained moped drove past it scrammbled the memory and crashed the computer. If anyone wants to put this in and/or comfirm/fill in blanks be my guest.--ElvisThePrince11:15, 15 April 2006 (UTC)
I'm unhappy about this link, although my change was reverted.
There are two issues, firstly what's a "stored-program" [computer]? This is obviously contentious, because the precise definition is influential on whether the SSEM was the "first" of anything amongst other machines, such as Colossus / ENIAC / Z3 / Harvard MkI. I'd agree that SSEM certainly was significant, novel and the first, but we could perhaps be clearer and more specific as to what it was doing that the others hadn't already demonstrated (the notion of an internally-stored and changeable program, beyond a merely "stored" program or even an internally-stored program).
Perhaps the link is "appropriate", but it's far from "clear". It's hard to describe the link as "wrong", but nor do I see it as helpful in terms of writing an accessible article that leaves the reader knowing more afterwards than we required them to know beforehand in order to understand it.
Why is SSEM significant? It's not the first "stored program" computer, it's the first of a narrow (and significant) subset of stored program computers.
What is a VN architecture? It's almost always presented in contrast to a Harvard architecture. While this is a correct distinction even for the SSEM, it's not obvious as to why it's important. The crucial aspect of SSEM, one of the two reasons it was important (along with being the Williams testbed) and the reason why it has a claim to primacy is from its implicit ability (of VN compared to H) to write self-modifying programs programs that can generate other programs (and the whole world of self-hosted compilers and programming). This important ability isn't entirely dependent on VN, as a Modified Harvard would have counted as being just as significant.
Maybe the link target is right and it's the text in the article that needs work?
As it is though, the article contains a term "stored-program" which isn't precise enough not to raise uncertainty (SSEM wasn't the first "stored-program computer") and then following the link dives off into a discussion of bus architectures. Neither of these clearly communicate the crucial factor, the opportunity for manipulating its own programs. Andy Dingley (talk) 16:07, 24 March 2010 (UTC)
The crucial factor is most certainly not the ability for a stored-program computer to modify its own programs, it's the easier and faster programmability. What, in your view, was the first stored-program computer if you don't agree it was the SSEM? --MalleusFatuorum16:11, 24 March 2010 (UTC)
What's the difference (back in the '40s or '50s) between "modify its own programs" and "easier and faster programmability"? These days I can develop for a pure Harvard microcontroller using a cross-compiler hosted on something more powerful, such as my fridge. Back then, it was either cross-compile on paper by hand or else a machine with a self-modifying capability. If the machine couldn't create or modify a program for itself, there wasn't anything else that could do it automatically. I don't mean, of course, "self-modifying programs" in the sense of an obfuscated Perl contest, but rather a simple compiler or assembler that uses one program and some input source data to create another program.
As for first "stored", "program", "computer" I'd go with Colossus, but that depends on your definition of each of the three terms. I'm happy enough with plugboard-programming, but do think there's a distinction between computers and calculators that rules out the Z3. (This being Wikipedia, I'd also point out that Colossus predated ENIAC, which would otherwise qualify) Andy Dingley (talk) 16:51, 24 March 2010 (UTC)
The difference is in who's doing it. The SSEM did not create or modify programs, and it had no assembler anyway, if that's your definition of "create or modify programs". I think it's stretching credibility just a little too far to consider a network of cables and plugs as a "stored program", in any sense that term would be understood today. All reliable sources agree that the SSEM was the world's first electronic stored-program computer. I also don't see the relevance of your Harvard vs von Neumann architecture distinction in this context, unless you're arguing that the SSEM used a Harvard architecture, which it most certainly did not. --MalleusFatuorum19:50, 24 March 2010 (UTC)
SSEM might not have created its programs, but the very similar Manchester Mk1 did. This is a hugely significant step, and the Harvard Mk1 couldn't have done it - the limitations of Harvard over Von Neumann architectures. Importance of "electronically-stored programs" can be overplayed too, if that only meant Harvard rather than VN or Modified Harvard - toggling in manually isn't much better than plugging as a program loading technique. "Storing programs electronically" wasn't the breakthrough, storing them in a way that allowed their automatic manipulation and creation was the innovative part.
No, the Manchester Mark I didn't have an assembler either; you're thinking of the Ferranti Mark I, a quite different machine. You seem to be caught up in this automatic creation and manipulation of programs though, which is so far off base I'm beginning to wonder which planet you're calling from. --MalleusFatuorum00:32, 25 March 2010 (UTC)
Autocode first appeared on the Manchester Mk1. Before the Ferranti Mk1 (which is anyway closer to the Manchester Mk1 than Manchester Mk1 was to SSEM) and before the Ferranti Mk1*. Andy Dingley (talk) 00:50, 25 March 2010 (UTC)
I suggest that before you say anything else you check your facts. The Manchester Mark I was scrapped in 1950; Autocode appeared on the Ferranti Mark I in 1952. You may be being misled by wikipedia's rather dreadful article on Autocode, which perhaps you may like to improve? --MalleusFatuorum01:01, 25 March 2010 (UTC)
Hmmm, I see what you mean about the Autocode article. This is admittedly a bit before my time, but as I recall from the big 50th anniversary shindig a few years back there was work done on the Manchester by Turing & Glennie, then Brooker's work on the Ferranti later on. Turing (quite famously) wrote the manual for this in the 6(?) months or so between working machines, and that manual wasn't written in a vacuum, but rather based on their previous work (and some admitted vagueness about the new hardware architecture). Andy Dingley (talk) 01:21, 25 March 2010 (UTC)
There's a lot of confusion caused by the "Mark I" label. The Manchester Mark I had been scrapped by the end of 1950, but the first Ferranti Mark I was installed at the University just a few months later, hence some call that the "Manchester Mark I". Turing wrote the manual for the Ferranti machine installed at Manchester, not the Manchester Mark I. --MalleusFatuorum01:32, 25 March 2010 (UTC)
I can't see SSEM as the "first electronic stored-program computer" either, if that means "electronic computer" with "stored programs" - Colosssus and ENIAC both predate it. That only works when you qualify it as "electronically-stored" "stored program computer" and even then, that's not of itself so important unless (as above) that electronic storage is used to make the programs manipulatable. The question is moot as the earlier Harvard Mk1 was neither "electronic" (a relay machine, not valves), nor did it have a manipulable program store, even if we do grant that it was an early machine that stored its programs by better than plugboards. Andy Dingley (talk) 00:23, 25 March 2010 (UTC)
Perhaps after a year things might have clamed down and we can engage in a civil discussion? It looks like in November 2010 Stored program was changed to redirect to Stored-program computer, which alas, has zero references! Seriously, when there are different definitions, it makes sense to explain them (unless they are only held by a small fringe group) with citations. An additional comparison might be made to the IBM SSEC which I think stored its program in memory so by one definition might be a "stored program computer", but was not fully electronic and like others program memory was very different than data memory. The SSEC seems to have been a few months earlier than SSEM though. These are all ground-breaking machines, so we should focus on the unique contributions of each. W Nowicki (talk) 19:45, 22 April 2011 (UTC)
Looking at this link at Columbia University, it explains the controversy a bit. SSEC could in theory modify its program, but rarely did. In fact, IBM did not use word "computer" then. Interesting to me that at the time the people working on it did not seem to make such a big deal about it. Also see McPherson, John, "A Large-Scale, General-Purpose Electronic Digital Calculator--the SSEC" (1948), IEEE Annals of the History of Computing, Vol.4 No.4 (Oct 1984), pp.313-326. Bashe, C.J., "The SSEC in Historical Perspective", IEEE Annals of the History of Computing, Vol.4 No.4, pp.296-312 (1982) W Nowicki (talk) 20:01, 22 April 2011 (UTC)
The SSEC wasn't really a computer in the modern sense, more a programmable electromechanical calculator, but interesting nevertheless. Moreover, despite what our wikipedia article says, it wasn't able to store programs: "The SSEC was not a computer in the modern sense. It had no memory for software storage, and programs had to be entered on punched paper tape."[2] There's no credible doubt that the SSEM was the world's first stored-program computer. The condition of that article has absolutely nothing to do with this one, but it could certainly do with some work. Why not expand it and add a few references to it yourself? MalleusFatuorum20:14, 22 April 2011 (UTC)
John Backus's comment is also rather revealing: "I think it's an extreme stretch to consider it the first 'stored program' computer--even though one of the programs I did used some specially prepared storage cells as the source of an instruction after some data was stored in it." MalleusFatuorum20:22, 22 April 2011 (UTC)
Yes, good ideas. I found some articles on SSEC but they are IEEE and my membership ran out awhile back. Not sure what is meant by the Columbia quote; perhaps it means is that there were no tapes nor disks nor core nor williams tubes etc. But "calculator" vs. "computer" is a bit of positioning. A good quote from the abstract of one of the Journal articles: "It [SSEC] executed programs stored exactly as data in any of the sections of memory, and could select, compute, or modify the program in a variety of ways." http://doi.ieeecomputersociety.org/10.1109/MAHC.1982.10041 Probably a reliable source. What is undeniable and needs to be stressed, is that the SSEC was not meant to break new ground, but to be used for actual calculations. "Baby" was the opposite, designed to experiment with concepts such as the stored-program. The very limited amount of electronic memory on the SSEC did not allow the stored program capability to be really used. Certainly the SSEC article glaringly omits any mention of the work on "the other side of the pond" and needs much more work than this one. W Nowicki (talk) 23:41, 22 April 2011 (UTC)
And with all due respect to the IEEE and its contributors, the statement that "it [the SSEC] could select, compute, or modify the program in a variety of ways" is nonsensical. MalleusFatuorum23:57, 22 April 2011 (UTC)
Had some time to work ont he IBM SSEC this week. Ready for another reviewer when someone has the time. Disclosure: I have over 35 years of programming experience; grew up thinking IBM was an evil empire but am related by marriage to one of the people briefly involved. I heard Wilkes talk but probably not the Machester people. Could you please clarify what is "nonsensical" about the statement? That is, do you believe it is false, or is it confusing what is being claimed? Here is what I got from my research: The January 1948 demo was very public, reported by New York Times, many witnesses etc. so all sources say SSEC was in operation first before SSEM. (Although academic labs are not as big on public displays as corporations, SSEC was working internally since autumn of 1947 it seems.) Also every source agrees with Bashe that the SSEC had the ability to do arithmetic on its instructions, and could take its instructions from any kind of memory. In fact, IBM was issued US patent 2,636,672 on the SSEC which included treating instructions as data as a claim. (Back then one generally had to implement something before patenting it) What I see as the issue is that at the time, the punched tapes were considered the "memory" inside the machine, since they were behind the slick glass cases. Nowadays these would clearly be classified as input/output devices since they were inherently serial (and even the loops had access times of about one second). None of the programmers seemed to have executed instructions out of the electronic memory, except perhaps rarely (not sure if it had a "conditional branch" or if arithmetic on instructions was the only way to change flow?). Another irony is that the 158 registers by about 77 bits in the SSEC was more than SSEM's memory, and Kilburn's famous program could have in theory fit into the SSEC registers. But theory is different in practice, and those would have been electromechanical (relay), not electronic. What seems clear is the SSEM was the first working electronic stored-program computer. That would be more clear anyway, since humans can be "computers", and humans can modify their instructions in memory, and store them in the same brain cells as numbers. Also surprising to me was the 1948 IBM paper already talking about the need for electronic memory. IBM must have licensed the WIlliams tube fairly soon after that, since it came out as a product on the IBM 701 soon after the Feranti Mark I. Have not found exactly the chronology yet, but still deserves a mention I would say. The real value of the SSEC seems to have been to get IBM good press, which it did. W Nowicki (talk) 23:48, 30 April 2011 (UTC)
It's probably just the way the IEEE's claim is written; in what sense could the SSEC select its program in any way, never mind a "variety" of ways? I don't think there's any credible doubt that that the SSEM was indeed the world's first electronic stored-program computer, and very little that it was the world's first stored-program computer of any type, in the sense that we would use that term today. I'm intrigued by the continual references to the ability to alter programs in memory, as if that's something any sane person would ever want to do. MalleusFatuorum00:21, 1 May 2011 (UTC)
Your first question: depending on the results of a calculation, the SSEC could, under program control (instead of human control), take its next instruction from any of the 158 registers or 66 tape readers depending on intermediate results. Modern computers generally use a conditional branch to do this; early programmable calculators like the Harvard Mark I did not have this capability, so it was a step forward. From the programmer's testimony (vs. the designers) it seems this capability was rarely used, since the applications like Ephemeris were not conditional but repetitive and loops could be made by physical loops of tape. As for altering programs by programs, any linker or compiler uses this. So it is also a notable ability that modern computers have; I would say necessary but not sufficient to be stored program. On every PC for example, hundreds of Dynamic Link Libraries (DLLS) get relocated using this capability. The irony was that the SSEC was programmed totally in numeric machine language, long before compilers or relocating linkers. It was so difficult that it inspired Backus et al to work on interpreters and Fortran for example. W Nowicki (talk) 17:38, 1 May 2011 (UTC)
Potential references
...in addition to those on article page
Papers
Functional Structure of the Baby. IEEE Annals of the History of Computing; Jul-Sep2005, Vol. 27 Issue 3, p47-47, 2/3p
Replicating the Manchester Baby: Motives, Methods, and Messages from the Past. Annals of the History of Computing; Jul-Sep2005, Vol. 27 Issue 3, p44-60, 17p
The Principle of CRT Storage. IEEE Annals of the History of Computing; Jul-Sep2005, Vol. 27 Issue 3, p45-46, 2p
The beginnings of the Manchester computer phenomenon: People and influences. IEEE Annals of the History of Computing; 1993, Vol. 15 Issue 3, p9, 8p, 2bw
'Baby's' Legacy -- The Early Manchester Mainframes; Burton, C.P. ICL Systems Journal; 1999 Vol. 13 Issue 2.
A storage system for use with binary digital computing machines, Kilburn T, PhD Thesis, 1948,
Universal high speed digital computers: A small scale experimental machine, Proc IEE, Vol 98, Part 2, No 61, 12-28, Feb 1951.
Books
Early British Computers; Lavington, S.H.; 1980 : Manchester Univ. Press
Electrical Engineering at Manchester University -- 125 Years of Achievement; Broadbent, T.E.; 1998 : Manchester School of Eng., Univ. of Manchester. Document Type: Book Citation; (AN
There is no mention of this being in the UK or England. I think the first sentence should have the word "England" next to "Manchester". MetalFusion81 (talk) 21:36, 21 June 2015 (UTC)
It might be helpful to point out, at least in the article body, that for the particular choice 2^18, it is trivial to derive the correct answer, 2^17 with a single division. The point of the exercise was that the computer did it the hard way, by exhaustive search.JQ (talk) 22:25, 2 June 2009 (UTC)
Any suggestions you can make along those lines would be welcome. As you say, finding the proper divisor of any power of two is trivial, which is partly why 2^18 was chosen, as the expected result is obvious. The SSEM only had a subtractor though, so there was only one way it could perform the calculation, by repeated subtraction. --MalleusFatuorum22:41, 2 June 2009 (UTC)
Yet again, we have wiki-dogma imposed to de-capitalise an article, without any support from sources, or any prior discussion here: page move. This edit is particularly egregious, as it mis-represents linked sources to change their titles to the uncapitalised form. DickLyon, the editor concerned, has no evident prior involvement with this article or topic, but does have a long and acrimonious track record for such dogmatic and damaging changes elsewhere, such as the railway projects.
I'm in favour of reverting to the original capitalised version, as I don't think I have ever seen DickLyon's form in a source. Such a major change should surely be discussed here first. --TedColes (talk) 11:43, 7 March 2018 (UTC)
Rack, any particular reason for your preference? Take a look at the sources discussed below and let me know what you're thinking. Thanks. Dicklyon (talk) 03:00, 11 March 2018 (UTC)
Andy, your reactions seem be more based on your dislike of me than on what I actually did, which was source-based (see below). And if I accidentally changed/misrepresented the title of a reference or external link, please point it out; I don't think I did. Dicklyon (talk) 02:46, 8 March 2018 (UTC)
It seemed an unlikely form for a proper name (with hyphen between capped words), so I went looking at sources. The article was created in 2004, so lots of more recent sources have copied it from here, so I looked earlier. In books, I found a few like this hit with Small Scale Experimental Machine (no Manchester, no hyphen, but capped), and some with lowercase hyphenated like this one. Their 1951 paper was actually titled "UNIVERSAL HIGH-SPEED DIGITAL COMPUTERS: A SMALL-SCALE EXPERIMENTAL MACHINE" in all caps on the paper, but listed in title case at the IEEE; the wording clearly indicates that it is not a name but a description, and that text is not used inside the article at all. Independent sources such as this one treated it as descriptive, and added the hyphen. See more hits. If we want to cap it, where is the argument to do so, relative to guidelines like WP:NCCAPS and MOS:CAPS? A snippet from a 1998 article comments on what it's called: "… - one reason why the name 'Baby' has pre-dominated in the fiftieth anniversary literature over 'Mark I' or 'Small-Scale Experimental Machine'". Another says, "The 'Baby' (more formally called the small-scale experimental machine, SSEM) ...". Another says, "The University of Manchester's
Small-Scale Experimental Machine (SSEM), known as the Baby" (where the caps are for defining the acronym, not used as a name). So why aren't we just calling it the Manchester Baby? The 1948 Nature letter did not give it a name, and the replica in the museum (pictured) says "It was nicknamed 'Baby' because is was a small-scale experimental prototype." Dicklyon (talk) 02:08, 8 March 2018 (UTC)
Original capitalised version. For this topic I'd consider the Science Museum Group canonical, given that they quite literally write the history when it comes to the story of computing in Britain (and that their displays, publications and website will almost certainly be the main drivers of traffic to this article), and they unequivocally call it "The Manchester Small-Scale Experimental Machine". ‑ Iridescent09:12, 8 March 2018 (UTC)
Thanks for chipping in. I love visiting the Museum of Science and Industry, although I haven't been for a while and I've never seen the Baby actually running. EricCorbett09:33, 8 March 2018 (UTC)
It's less impressive than it used to be; the Ordsall Chord has severed their railway line, so they can no longer run their trains more than a couple of hundred yards out of the shed. I find the big blue shed full of vintage cars across the road more interesting than the main museum, and an order of magnitude less full of screaming schoolchildren. ‑ Iridescent09:43, 8 March 2018 (UTC)
The museum site you linked uses the term "Small-Scale Experimental Machine" only once, and Baby many times. The plaque by the replica machine says "It was nicknamed 'Baby' because is was a small-scale experimental prototype." Nowhere does anyone claim that somebody ever named it "Small-Scale Experimental Machine", unless I've missed that. Dicklyon (talk) 18:14, 8 March 2018 (UTC)
When you're in a hole stop digging. Lavington (2011) uses the name "Small-Scale Experimental Machine", as do Rojas & Hashagen (2002), Tatnall, Blyth, & Johnson (2013), Joubert, Leather, & Parsons (2016) ... EricCorbett19:03, 8 March 2018 (UTC)
Well, all but one of those are long after Wikipedia promulgated that name. The 2002 book uses "a Small Scale Experimental Machine" in the text; not clear why the caps; elsewhere they use "the"; no hyphen sometimes, with hyphen other times, indicating not much thought went into this. And the index for this says "see Manchester Baby". Dicklyon (talk) 04:52, 9 March 2018 (UTC)
Looks like the computer50.org page got rehosted here. This 1998 or 1999 page is the earliest I can find referring to "the Small Scale Experimental Machine", as opposed to a small-scale experimental machine. Did they thereby establish its "official name" by commemorating it? Seems odd, esp. since digital60.org has it with and without the hyphen (as well as "the small scale SSEM"), indicating little attention to detail. Dicklyon (talk) 05:04, 9 March 2018 (UTC)
Searching more, I find zero evidence of the Small Scale Experimental Machine before the 50th anniversary reconstruction of 1998, except for citations to the title with a. In adopting the name proposed by its reconstructors as if it's the machine's official proper name, we misrepresent reality, in my opinion. At the least, we should attribute the name to those reconstructors who made it up. Dicklyon (talk) 05:14, 9 March 2018 (UTC)
In the 1953 write-up of the conference "Bowden, B.V., ed. (1953), Faster Than Thought, London: Pitman, ISBN 0 273 41330 9" the machine was called "A Prototype Machine" (p118) and the second 1949 machine "A Large-scale Machine" (p123). So it seems that these early machines did not have names as such because they were intermediate steps in a continuous development towards what was then named M.A.D.A.M. (Manchester Automatic Digital Machine), but later the Manchester Mark 1. It is therefore incorrect to say that everyone called it "Baby", a nickname but not a formal name. I guess the formal name arose when the historical significance of this test bed for the Williams–Kilburn tube as the world's first stored-program computer was realised. --TedColes (talk) 08:23, 9 March 2018 (UTC)
OK, so not "everybody"; only those who felt it needed to be referred to by a name. So what do you mean by "the formal name arose". Arose where? Proposed by whom? Dicklyon (talk) 21:42, 9 March 2018 (UTC)
If you're suggesting that people such as Lavington copied the name from Wikipedia then I'm afraid that there's nothing more that can be said to you, other than that you're living in some kind of fantasy world. EricCorbett07:55, 9 March 2018 (UTC)
I'm not suggesting that, and I don't see why you want to make this about me. Let's focus on the evidence. Lavington 2011, in fact, usually uses lowercase, not caps. And take a look at Steve Furber's 60th anniversary talk/paper for instance. He says "The ‘Baby’ (more formally called the small-scale experimental machine, SSEM)..." So maybe that's what it's called formally, but still not a proper name, according to this particular expert, by 2008, even though wikipedia had that capped for 4 years by then. Dicklyon (talk) 21:42, 9 March 2018 (UTC)
Joubert, Leather, & Parsons (2016) use the name Small Scale Experimental Machine (without the hyphen) just once, and Baby 8 times. Maybe you can tell us what some of those other sources do. Dicklyon (talk) 22:03, 9 March 2018 (UTC)
The Bowden book says "A Prototype Machine -- The next development was to build a small prototype machine ... this prototype machine ... the prototype machine ...", so yes, no name there, and caps only in heading. Dicklyon (talk) 03:45, 10 March 2018 (UTC)
In this 1953 article, its creators called it "The Miniature Machine" (in a heading) and "the miniature machine" (in a sentence); no Baby or SSEM except in the reference to the earlier paper with subtitle "A Small Scale Experimental Machine". Dicklyon (talk) 19:29, 10 March 2018 (UTC)
And Lavington's 1978 paper on the Atlas refers to "the prototypes" and uses "a small-scale experimental machine", only in the references, in a context where only proper names are being capped (as in "An attempt to simplify coding for the Manchester electronic computer"). Dicklyon (talk) 21:21, 10 March 2018 (UTC)
This 1998 Shelburne & Burton paper on the original programs only uses Small Scale Experimental Machine when defining the acronym SSEM; it is common to cap non-proper names when introducing their acronyms, so this can't really be taken as evidence of treatment as a proper name. Dicklyon (talk) 21:49, 10 March 2018 (UTC)
Agar 1998 talks about names: "a rebuilt computer marked Manchester as 'home to the second industrial revolution', but also crucially signified a future, even a 'rebirth' - one reason why the name 'Baby' has pre-dominated in the fiftieth anniversary literature over 'Mark I' or 'Small-Scale Experimental Machine'" (that's all I can see in the search snippet; anyone have it?). Dicklyon (talk) 23:11, 10 March 2018 (UTC)
Your point is becoming even more obscure than it was, so perhaps you need to clarify what exactly it is. If it's that the SSEM wasn't known as such during its development, as it appears to be from what I can understand, then perhaps you can explain why it is that you have not similarly insisted on renaming the Wikipedia article on World War I to something else as well, as it was certainly not known as such at the time. EricCorbett18:39, 10 March 2018 (UTC)
Now you're just being silly. My point is that small-scale experimental machine is not a proper name; it is not consistently capitalized in sources, so per MOS:CAPS we should not capitalize it. Baby, on the other hand, seems to be consistently capped, and was the name more commonly used, so if you want a proper name, that's a possible choice. As for whether it was known as the small-scale experimental machine during its development, yes, it was, as Tootill says in his oral history and as the subtitle of his MS thesis calls it; but they never treated that as a name. Dicklyon (talk) 19:29, 10 March 2018 (UTC)
I'm trying to discuss the most appropriate name for an article on this small-scale experimental machine called Baby, and you're talking about renaming World War I. Neither is likely to make anyone laugh, I expect, but thanks for playing. Dicklyon (talk) 21:09, 10 March 2018 (UTC)
I have absolutely no idea what it is that you're trying to do, unless it's to demonstrate how single-mindedly blind you can be when in the fevered grip of some misguided crusade or other. But perhaps we can leave the last word to Williams, Kilburn and Tootill? I assume you know who they are? The title of their 1951 paper published in the Proceedings of the I.E.E., published in 1951, is "Universal High-Speed Digital Computers: A Small-Scale Experimental Machine". The nickname of Baby does not appear anywhere in that article, which must surely be considered canonical. I shall not be further wasting my time in replying to anything else you have to say on this subject, and I dare say that you must have better and more productive ways to spend your time as well.EricCorbett01:15, 11 March 2018 (UTC)
But "a small-scale experimental machine", even if capped in a title (theirs was printed as all-caps by the way) doesn't suggest that "Small-Scale Experimental Machine" is a name, much less a proper name (and it's often lowercase in citations to that paper). So that's not a last word, just a word that led a few others years later to maybe treat that as a name; most did not. Dicklyon (talk) 03:42, 11 March 2018 (UTC)
So, the name has evolved over time. I don't see why you argue that "Small-Scale Experimental Machine" is not name or a a proper name. How do you classify "National Aeronautics and Space Administration"?--TedColes (talk) 10:05, 11 March 2018 (UTC)
I've compiled a list of sources, and what they call it, at User:Dicklyon/Baby. Please review and let me know if there are any remaining objections to moving to Manchester Baby, or whether the non-capped version of the present title would be preferred. And let me know if I missed any relevant source items. Dicklyon (talk) 03:45, 16 March 2018 (UTC)
Congratulations, Dicklyon on your list of sources. It seems to me that we have a number of different problems about the name of the article.
1. Should it be the descriptive name or the nickname?
2. If the descriptive name, which words to be in capitals?
3. If the descriptive name, should it include the hyphen?
4. If the nickname, Baby or 'Baby'?
I think that the list of of sources gives considerable support for continuing with the descriptive name. As capitals for all words seems to be the most common use, I favour that. As regards hypenation, I go along with the Oxford University Press who say that "If you take hyphens seriously, you will surely go mad”. So, in short, I do not think that there is sufficient evidence or consensus for a change. --TedColes (talk) 08:05, 16 March 2018 (UTC)
The evidence seems very clear that the present title is not a proper name, and is seldom treated as such in sources; there is no point in the chronology when reliable sources switched to treated it as such (that is, it hasn't "evolved" into a proper name, as the evidence shows). If we go with the descriptive name it would be Manchester small-scale experimental machine per WP:NCCAPS and MOS:CAPS; that is, we don't cap words in descriptive titles except when they are proper nouns. The "common name", some variant of which is used my almost all sources, seems to be Manchester Baby or Manchester Baby machine; there's really no reason to put the quote marks into it. Dicklyon (talk) 20:31, 16 March 2018 (UTC)
So the evidence for a change seems clear. We can judge consensus in an RM discussion if we need to. If you're saying you object to a move, then I guess we need to. Dicklyon (talk) 20:33, 16 March 2018 (UTC)
I really do not think there is anywhere near a good reason to change the title of this article. WP:TITLECHANGES says "Changing one controversial title to another without a discussion that leads to consensus is strongly discouraged. If an article title has been stable for a long time, and there is no good reason to change it, it should not be changed. Consensus among editors determines if there does exist a good reason to change the title." --TedColes (talk) 10:29, 17 March 2018 (UTC)
Well, there is a very good reason to change it: it's pretending to be a proper name when it's not; see WP:NCCAPS and MOS:CAPS. When I first boldly moved it, I had no reason to suspect a controversy might arise. I got reverted. Now we're discussing. I'm still a bit surprised at the extend to which some of you are willing to ignore consensus guidelines and evidence, so we'll need to keep discussing. I'll open a formal RM at some point, which will advertise the question to the larger community. Dicklyon (talk) 17:09, 17 March 2018 (UTC)
One more observation; see User:Dicklyon/Baby#Conclusion: Perusing the refs, one can see that the introduction of the capped/hyphenated name was in 1998 by C. P. Burton, the "Project Manager of the Small-Scale Experimental Machine Rebuild Project", and that most other occurrences in the literature are in fact by the same C. P. Burton. Very few other authors have picked it up. Furber did use "Small Scale Experimental Machine" once in 2016, without the hyphen, which at least has the right form for a proper name, but is pretty rare; he also used "small-scale experimental machine" in 2008, treating it as descriptive as the original authors did. It seems wrong to anoint this as a proper name based on just one author who had an obvious affinity for it. Dicklyon (talk) 04:00, 18 March 2018 (UTC)
Requested move 18 March 2018
The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review. No further edits should be made to this section.
Support as nom since the evidence at User:Dicklyon/Baby makes it clear to anyone who will look that this is not a proper name. The current title is the idiosyncratic name used by C. P. Burton, who led the rebuild project. I have no objection to him capping it, but in Wikipedia we have our own style, that relies on evidence in sources, per MOS:CAPS; we reserve caps for proper names. The editors who objected to the move previously (before the evidence was collected) have not even responded to questions or addressed the evidence. Dicklyon (talk) 22:09, 18 March 2018 (UTC)
Given your long history of such undiscussed and heavily badgered renames, are you still seriously asking this? Have you not yet noticed how much you have irritated and annoyed other editors? Andy Dingley (talk) 00:17, 19 March 2018 (UTC)
Like I requested in the discussion above, why not discuss the question rather than discuss me? I do understand you don't like me, but that's not a reason for me to stop trying to improve the Wikipedia. Dicklyon (talk) 00:21, 19 March 2018 (UTC)
Oppose. Having reviewed the discussion above, and the sources presented therein, it seems that the capitalised form is correct. Thryduulf (talk) 21:49, 19 March 2018 (UTC)
Oppose. Although the title for the original machine may have been descriptive, a half-century later it appears that the name has become capitalised–most certainly so for the replica. Useddenim (talk) 22:44, 19 March 2018 (UTC)
I agree the sourcing indicates the long phrase is a made-up description created by a particular writer.
WP choosing to use it is just us picking a WP:NTITLE / WP:DESCRIPTDIS descriptive article title (and it may actually fail NTITLE if it's not the name usually preferred by other writers on the topic, which seems to actually be true). Us using it as the article title doesn't magically transmogrify it into a proper-noun phrase (WP uses sentence case for article titles, remember).
That some people have taken to capitalizing it is irrelevant. Some people have taken to capitalizing almost anything you can think of (hell, we even had a debate a few years ago here about capitalizing things like "mountain lion" and "bottlenosed dolphin" because some people, including some academics, like to capitalize vernacular names of species).
What museums put on placards and in brochures is also irrelevant; they capitalize everything they put on display as a matter of house style. We do not follow their house style nor they ours.
"Small-Scale Experimental Machine" does not consistently appear in this capitalized form in reliable sources, as Dicklyon has clearly shown in his sourcing. That's what matters here.
The opposition above is pure WP:ILIKEIT that's directly defying both the sourcing and our guidelines.
The entire problem here is that people are mistaking a description for a proper name. Clear analogy: If you attend an institution called the McCandlish Academy of Basketweaving [Baby], and you descriptively refer to it as the basketweaving college [small-scale experimental machine], this is not to be overcapitalized as "the Basketweaving College" [Small-Scale Experimental Machine]. PS: Even if a capitalized form were kept, it would be "Small-scale Experimental Machine" (don't capitalize after a hyphen except when what follows the hyphen is a stand-alone proper name, as in "post-Soviet"). — SMcCandlish☏¢ 😼 20:42, 22 March 2018 (UTC)
Yes, this is probably the best way to avoid the whole "I like caps" problem. I had proposed maybe Manchester Baby, which already redirects here, is a more natural disambiguation, is used exactly so in some sources, and fits the scheme of other Manchester computers better; almost all sources use capped "Baby", in various contexts. We haven't had much response on that idea; might need another RM to elicit actual responses, since this one didn't get any useful responses. As Agar 1998 says "rebuilt computer ... crucially signified a future, even a 'rebirth' - one reason why the name 'Baby' has predominated in the fiftieth anniversary literature over 'Mark I' or 'Small-Scale Experimental Machine'." (even though that same fiftieth anniversary literature is where "Small-Scale Experimental Machine" was turned into a name by Burton). Dicklyon (talk) 05:40, 23 March 2018 (UTC)
But the sources that I surveyed make it clear that essentially all sources the use the "nickname" Baby, while the current name was made up by the 50th anniversary rebuild project and used only by them until after Wikipedia adopted it. So your "well sourced" comment sounds like you haven't reviewed the evidence presented. Dicklyon (talk) 18:09, 28 March 2018 (UTC)
More badgering
OK, if pointing out that arguments are vacuous is badgering, then more of it is needed.
An argument to keep the present title needs to be of one of two forms:
Ignore WP:NCCAPS and MOS:CAPS and keep the title capitalized even though it's not a proper name; or
It's a proper name, and the reason I think so, based on sources, is ...
But in the opposing arguments above, and in the previous discussion, it's hard to discern anything of that form. Let's look:
Andy Dingley: "Obvious consensus is obvious." and previously "we have wiki-dogma imposed to de-capitalise an article, without any support from sources" and other false and ad-hominem statements that can reviewed in the previous section. Yes, it was undiscussed at the time, but he has refused to engage in discussion based on sources.
Eric Corbett: "A consensus has already been established against this proposed move, above." Yet nobody had commented after I presented the source analysis; what kind of discussion/consensus can that be?
TedColes: "for the reasons I have previously given above" which must be referring to these gems:
"I don't think I have ever seen DickLyon's form in a source" which I refuted and he didn't acknowledge
"I guess the formal name arose when the historical significance of this test bed for the Williams–Kilburn tube as the world's first stored-program computer was realised." Not bad for a guess, but he would not engage when I queried where by whom the name arose; I did the research and showed it arose in the works of only one author.
"the name has evolved over time" which I analyzed in sources and showed that this "evolution" only involved one author, C. P. Burton, who likes to cap his own project and job titles.
"the list of of sources gives considerable support for continuing with the descriptive name. As capitals for all words seems to be the most common use, I favour that." Does he not see that descriptive means not proper, and that it is false that capitals is most common? And why not acknowledge that the name "Baby" is actually the one most common in sources?
"I really do not think there is anywhere near a good reason to change the title of this article." ignores policy and guidelines that provide the reason. No response when I pointed out WP:NCCAPS and MOS:CAPS.
Thryduulf: "Having reviewed the discussion above, and the sources presented therein, it seems that the capitalised form is correct." content-free opposition; there is nothing in the discussion above, and nothing in the sources, that would support this view; if he thinks there is, why can't he point to it? Then he throws in a gratuituous ad hominem assumption of bad faith "Quite simply people disagree that you are improving the encyclopaedia."
Useddenim: "Although the title for the original machine may have been descriptive, a half-century later it appears that the name has become capitalised–most certainly so for the replica." OK, yes, C. P. Burton caps it for his replica. But nobody else does, so what is this "it appears" referring to? I analyzed the sources; why not question my analysis a little more explicitly? And the article is not about his replica.
If you ignore all the opposition, then obviously the proposal is unopposed, but fortunately we do not work that way on Wikipedia. It appears that you are genuinely unable to see that people who disagree with you might have valid reasons for doing so. I strongly suggest you review WP:ARBATC, particularly those sections relevant to user:Born2cycle whose behaviour you are increasingly emulating, Wikipedia:Arbitration/Requests/Enforcement/Archive227#Born2cycle and the ANI threads linked there. If you do not stop badgering this and other discussions (and this very section is a perfect example of you doing just that) then I will be proposing you be topic banned from RM discussions, either completely or with something like that offered to B2c. Thryduulf (talk) 03:03, 20 March 2018 (UTC)
I'm happy to admit that people who disagree with me might have valid reasons. I'm just pointing out that if they do, they are declining to state them. Dicklyon (talk) 04:27, 20 March 2018 (UTC)
The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page or in a move review. No further edits should be made to this section.
Sources and naming
With more work, I've listed 74 sources now at User:Dicklyon/Baby, and tabulated what they call this computer. I can see that my downcasing suggestion was probably a bad idea, compared to Manchester Baby. Maybe we'll need to take that up at some point. If you see anything I missed or got wrong, please point it out, or go ahead and make the addition or correction. Thanks. Dicklyon (talk) 03:40, 16 April 2018 (UTC)
I added 4 more books, including a few that use exactly the Wikipedia title "Manchester Small-Scale Experimental Machine" and do not mention "Baby" at all. This could be taken either as diluting my point a bit, or as an indication that some authors just take the title from Wikipedia for their brief point, and ignore the fact that Baby is what it's called. Dicklyon (talk) 03:11, 17 April 2018 (UTC)
Requested move 24 May 2018
The following is a closed discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review. No further edits should be made to this section.
The result of the move request was: Moved – Support rationale is well constructed, whereas Oppose arguments are either based on personality of the OP, or on apparent citogenesis. — JFGtalk11:43, 31 May 2018 (UTC)
Several points have become clear from a study of nearly a hundred sources, and the article history, as collected at User:Dicklyon/Baby:
For its first fifty years, many called it the "baby" and nobody called it the "small-scale experimental machine", capped or not. The phrase "a small-scale experimental machine" appeared only as a subtitle of a frequently cited paper.
C. P. Burton, the Project Manager for Small-Scale Experimental Machine Rebuild Project, declared the "official" name to be "Small-Scale Experimental Machine" in 1997, even though no source had ever called it that.
Almost no source used Burton's name until after Wikipedia did in 2004.
Almost no source prefixed it with "Manchester" until after Wikipedia did in 2006.
Most sources before and after the Rebuild Project call it "baby" or "Baby". Half of those use "Manchester Baby", which is a reasonable parallel to "Manchester Mark 1", the machine that the Baby evolved into.
It is clear that Wikipedia has been unreasonably effective in inducing sources to use the name that our article is titled with. This is not a good reason to stick that that name, especially since what it was actually called, the "Baby", is still more common. Dicklyon (talk) 04:12, 24 May 2018 (UTC)
Oppose per long past discussion. I would also support a topic ban here, and more broadly, against DickLyon using these attrition tactics to push through page moves. Andy Dingley (talk) 09:09, 24 May 2018 (UTC)
So you still have no actual comment on the issue or the evidence, just on me? I suggest editors carefully review the "long past discussion" above to see whether anyone presented a reason to prefer the current wikipedia-invented name. Nobody did; certainly not you. My entire logic and process are as discussed above; I explained already at § Sources and naming that the previous RM was not the best choice and this title would be better; I gave you a month to review the evidence and think about it before I started the new RM. Just trying to drive things in the best possible direction. Dicklyon (talk) 14:49, 24 May 2018 (UTC)
Support, basically per previous discussions. We know for a fact that "Manchester Small-Scale Experimental Machine" is not the common name and is just a description made up by a single writer. This page has to move, and either "Manchester Baby" or "Baby (computer)" will work, since both are well attested in multiple sources. [Not that sources literally write "Baby (computer)"; rather, they write about a computer called "Baby", and some of them also prefix it with "Manchester".] — SMcCandlish☏¢ 😼 15:28, 24 May 2018 (UTC)
Oppose for the reasons I have given previously, particularly "Changing one controversial title to another without a discussion that leads to consensus is strongly discouraged. If an article title has been stable for a long time, and there is no good reason to change it, it should not be changed. Consensus among editors determines if there does exist a good reason to change the title.".--TedColes (talk) 15:33, 24 May 2018 (UTC)
The good reason to change it is well documented in the subsection immediately above. Also, I had summarized your reasons given previously in § More badgering. Please let me know if I missed any reasons or incorrectly summarized your position. I don't see any reason that the commonname should be considered controversial in this case, as the made-up name is. Dicklyon (talk) 17:17, 24 May 2018 (UTC)
And what TedColes wrote doesn't make any sense. "Manchester Baby" isn't controversial, but well attested. More importantly, it isn't possible for an RM concluding to move the page to be "without a discussion that leads to consensus"; that quoted material is about making manual moves without RM. RM is the consensus discussion. That is, discussing a requested move until we reach a consensus is the sole reason that RM exists. — SMcCandlish☏¢ 😼 00:11, 25 May 2018 (UTC)
Oppose: For reasons outlined above I think Manchester Small-Scale Experimental Machine as the title (ie use by Univ of Manchester and MOSI) with note nicknamed ‘Baby’ in the opening sentence woud seem fine. (Msrasnw (talk) 11:41, 25 May 2018 (UTC))
The only reasons I find "outlined above" are that some guys don't like me; can you please state what reasons you mean? The MOSI ([3]) mostly calls it Baby, but of course they also use Burton's name since they have his rebuild. Dicklyon (talk) 15:16, 25 May 2018 (UTC)
Hi sorry. By "outlined above" I meant "outlined by me above". My feeling is that both at the University of Manchester (its home) and at the Museum of Science and Industry - where the replica is, they seem to me to use the Manchester Small-Scale Experimental Machine as its proper name and Baby as a nickname. Both these seem authoritative. I think using nicknames for things might be one way to go but it would just seem less formal or less encyclopedic. (Msrasnw (talk) 17:29, 25 May 2018 (UTC)) PS: I am here via my having started a stub on Geoff Tootill one of its creators. PPS At IBM Don Estridge had his 'baby computer' too which might complicate things.
Thanks for the clarification of your "feeling". Note that the uses at the University of Manchester are primarily just in their rehosting of the computer50.org pages, where the name "Small-Scale Experimental Machine" originated. The Museum of Science and Industry is where that Rebuild lives, so also uses that term. But they use Baby even more, because the machine didn't really need a new name. Look at what Agar said of that 50th-anniversary literature (see User:Dicklyon/Baby: the name 'Baby' has pre-dominated in the fiftieth anniversary literature over 'Mark I' or 'Small-Scale Experimental Machine'. Neither site used "Manchester Small-Scale Experimental Machine" before Wikipedia adopted that title, so your observation is an example of the WP:CIRC that I'm trying to fix. Dicklyon (talk) 19:10, 25 May 2018 (UTC)
Support: Per nomination. Opposition arguments don't resonate. First one is a personal attack. Second one seems to indicate personal preference. Daniel.Cardenas (talk) 16:39, 25 May 2018 (UTC)
Support per nomination. Manchester Baby is concise, natural and consistent with most sources as indicated. kml (talk)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
70th anniversary lecture coming up this month at Manchester MoSI
See Manchester Literary and Philosophical Society news: "... the Manchester Lecture will take place at 7 p.m. on Thursday 21 June 2018 at the Manchester Museum of Science and Industry (MSI). This year it is in celebration of the 70th anniversary of the creation of the world's first stored-program computer on 21 June 1948 at the University of Manchester. The computer is known as 'The Baby' and a replica of it is located at MSI." Dicklyon (talk) 06:00, 5 June 2018 (UTC)
