Wikipedia:Good article reassessment/Manchester Mark 1/1

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Manchester Mark 1[edit]

Article (edit | visual edit | history) · Article talk (edit | history) · WatchWatch article reassessment pageGAN review
Result: List as GA, but without prejudice concerning the optimal structure or emphasis for the article, per discussion below. Geometry guy 19:08, 17 February 2009 (UTC)

I believe that this article more than meets the GA criteria and that it was failed because of an ill-informed and idiosyncratic view of what such an article ought to look like, and the material that it ought or ought not to contain. --Malleus Fatuorum 00:05, 11 February 2009 (UTC)

The reasons for failing it are given at Talk:Manchester_Mark_1/GA1#Conclusion_of_review. --Philcha (talk) 00:20, 11 February 2009 (UTC)

For the sake of clarification I will outline my objections to the reasons for failure:

  • It is in my view perverse to suggest that a description of how data was entered into the machine and represented internally is "unnecessary detail". It is certainly true that Zuse had no contact with the Manchester team, but neither did Eckart and Mauchly, who the reviewer kept insisting be included because of their imagined input into the Mark 1's design. Zuse is mentioned only in the context of placing the development of stored-program computers into its historical context in any event, a section of the Background that the reviewer insisted be expanded, against my better judgement.
  • Williams tubes did not quite rapidly displace mercury delay lines as "the standard memory technology" as the reviewer claims, and indeed were never "the standard memory technology". Both were quite quickly replaced by core-storage, because both had very significant disadvantages.
  • I fail to see the relevance of a long list of features that the Mark 1 did not have, such as hardware interrupts and an operating system, which made it more difficult to program. Especially when that observation is considered against the objection to the article including features that it did have which made it more difficult to program, such as the need to break each 40-bit word into eight 5-bit chunks encoded by a modified version of the IAT2 teleprinter code, and then reverse the binary representation because of the machine's "endianness".
  • The Mark 1 had two purposes, as I believe the article makes perfectly clear. Initially to provide a realistic computing facility for the university, but very quickly after work began to also provide a prototype on which the design of the Ferranti Mark 1 could be based.

--Malleus Fatuorum 15:29, 11 February 2009 (UTC)

  • Comment. The GA process is not a good vehicle for resolving disagreements over content and structure. Such issues are better handled when all views are represented by editors on an equal footing, as contributors to the article, on the article talk page. My initial impression is that the differences of opinion on the content to include here are not major enough to be a GA issue, per criterion 3. There are many ways to write a broad article.
There's also nothing in the criteria about the target audience. All Wikipedia articles should be as accessible as possible, but it seems to me that this one is of greatest interest to readers who are expert in or fascinated by the history of computing, not necessarily modern programmers. It is not the role of every article on every early computer to emphasise just how different programming was then. They should certainly describe it and provide wikilinks, but I think the article now does a reasonable job, in part thanks to improvements made during the GAN review.
One issue I would question is the "Cultural impact" section: if this is notably about the Manchester Mark 1, rather than the Manchester project in general, then it is worthy of mention in the lead; if not, it could be cut. If kept then "unexpected" needs to be explained (who didn't expect it?).
Another small issue I noticed is the paragraph about index registers, beginning "Perhaps the machine's most significant innovation was its incorporation of index registers,..." The novice reader may find this disorienting, as the first additions mentioned are the multiplication registers, rather than the index registers. Geometry guy 20:46, 11 February 2009 (UTC)
  • Reply
You make a good point about the "Cultural impact" section. The chronology suggests that the debate was sparked by the publicity surrounding the Manchester Mark 1 (and/or its prototype, the SSEM), and that's also what the source suggests, so it should be included in the lead, as you say. My excuse is that I only added that section yesterday, and hadn't thought about its implications for the lead. Thin, I know. --Malleus Fatuorum 21:00, 11 February 2009 (UTC)
The reaction was unexpected by the developers, which I've clarified. --Malleus Fatuorum 22:37, 11 February 2009 (UTC)
  • Comment As a non-expert, I found the article informative and interesting, and think you've got the balance about right. I thought the passing mention you made of the lack of features we see today was enough, and with the descriptions of the advantages of the Manchester Mk 1 over its predecessors, was sufficient to set it in its historical context without going into unnecessary detail. The usual GA technicalities; references, layout, lead, images etc look fine. Personally I'd be happy to pass this as a GA. EyeSerenetalk 11:35, 13 February 2009 (UTC)
  • Comment I would also pass this as a GA without any of the drastic changes called for during the previous review.Pyrotec (talk) 21:39, 13 February 2009 (UTC)
  • Intention to close as list, but without prejudice concerning the optimal structure or emphasis for the article. The original reviewer has indicated on my talk page that, despite disagreements, the article is probably acceptable for GA anyway. Any objections to such a close? Geometry guy 23:48, 15 February 2009 (UTC)
  • No objections to Listing as GA.Pyrotec (talk) 09:49, 16 February 2009 (UTC)
Resolved comments hidden. Noble Story (talkcontributions) 01:13, 17 February 2009 (UTC)

Sorry to rain on the party, but...Oppose Listing. I find this article undercited.

...but it was mathematician John von Neumann who became widely credited with defining that computer architecture, still used in almost all computers.  Done --Malleus Fatuorum 17:01, 16 February 2009 (UTC)

The University of Manchester's Small-Scale Experimental Machine (SSEM), the world's first stored-program computer, had successfully demonstrated the practicality of the stored-program approach and of the Williams tube, an ea rly form of computer memory based on a standard cathode ray tube (CRT), by running its first program in June 1948.  Done --Malleus Fatuorum 17:01, 16 February 2009 (UTC)

Stored-program computers were also being developed by other researchers, notably the National Physical Laboratory's Pilot ACE, Cambridge University's EDSAC, and US Army's EDVAC. The SSEM and the Mark 1 differed primarily in their use of Williams tubes as memory devices, instead of mercury delay lines.  Done --Malleus Fatuorum 17:01, 16 February 2009 (UTC)

On the the Intermediary Version programs were input by key switches, and the output was displayed as a series of dots and dashes on a cathode ray tube known as the output device, just as on the SSEM from which the Mark 1 had been developed. However, the Final Specification machine, completed in October 1949, benefitted from the addition of a teleprinter with a 5-hole paper-tape reader and punch.  Done --Malleus Fatuorum 17:01, 16 February 2009 (UTC)

All without a reference. Also, just wondering, how reliable a source the site is. I mean, the article (here) attributes it to the University of Manchester, but the actual site only says it's done by a former faculty member. Seems like a self-published site to me. Noble Story (talkcontributions) 14:20, 16 February 2009 (UTC)

Napper is a clearly reliable academic source, and he built the web site for Manchester University's Department of Computer Science; the university claims the copyright for that material, so it's hardly "self-published".[1]. That von Neumann is widely credited with defining the computer architecture which bears his name is hardly contentious, but if a citation is demanded so be it. --Malleus Fatuorum 15:31, 16 February 2009 (UTC)