Talk:Foy–Breguet telegraph/GA1

GA Review

Article (edit | visual edit | history) · Article talk (edit | history) · Watch

Reviewer: David Eppstein (talk · contribs) 01:01, 2 January 2020 (UTC)

First reading

1. (Style and prose quality.) The title of the article should use an en-dash, not a hyphen. The same is true for every concatenation of surnames of multiple people within the article. The lead does not appear to summarize the text of the article. Instead, it is the only place that contains a clear description of this system, the only place with a clear statement of who invented it when, the only place that states (without a source) that its optical predecessor existed elsewhere but that this electrical version did not. And "nobody else besides France" is unnecessarily colloquial; France is not a person. The text of the article switches around confusingly between tenses (usually simple past, but occasionally present and occasionally past perfect).

2. (Verifiability and original research.) The article is sparsely sourced. Some parts of the lead are unsourced, which would be appropriate for a lead that summarizes information provided in more detail later in the article, but not for material presented only there. I was unable to get easy online access to the Haigh and Holzmann sources, so I am taking their content on good faith, and only checked the others. In particular, I am assuming on good faith that the entire "Development" section is properly sourced to Holzmann; it is certainly plausible that all of the information in that section comes from the source. However, most of the "Connection to England" section, claimed to be sourced to Shaffner, failed verification. Only the single sentence "The Foy-Breguet operator could instantly see the letter being transmitted from the visual pattern, whereas the Cooke-Wheatstone operator had to count the left and right deviations of the single needle" matches the content of the source; the rest of the section (except for the one unchecked but sourced sentence at its start) appears to be unsourced. And much of the "Withdrawal" section is sourced to page 63 of Coe, which contains nothing of relevance (it appears to be mostly about using dog sled teams to construct a particular telegraph line in Alaska). The line about the 1851 adoption of the competing Gerke code is properly sourced, but the information in the same line claiming that this union "represented many central European countries" is not sourced and appears to be actually misleading, unless one treats the separate parts of what is now Germany as being the many countries in question.

3. (Breadth of coverage.) The article is lacking much description of how the instrument was designed, what its signals looked like, how it encoded and decoded them, and under what principles it operated. There are pictures but that's not enough to understand what the operators actually did or what the machines they used actually did. The sentences "Many central European countries were members of this union and they used a code known as the Hamburg code or Gerke code. This was a heavily modified version of the original American Morse code and later formed the basis of International Morse code." go into far too much depth about material that is off-topic for this article, especially for the lead. Similarly, the later sentences "There were strong arguments put forward for the superiority of optical telegraphs over electrical telegraphs. Chief amongst these reasons was that electrical systems were vulnerable to attack by saboteurs. In an optical system, only the telegraph stations needed to be defended. An electrical system was impossible to defend over its many hundreds of miles of exposed wires." are completely off-topic, as is the information about how the inventor's grandfather became a watchmaker.

4. (Neutrality.) No issues found.

5. (Stability.) The article was created last March and significantly expanded last November by the nominator; other than that it has few edits. I don't think there is any problem with stability.

6. (Images) The images are appropriate and appropriately captioned. Two of the three are on the English Wikipedia rather than (as would be more appropriate) Wikimedia Commons, and even more mysteriously have been tagged to be kept that way, but that is not an issue for the GA rules. The England–France image is likely public domain, but its provenance is dubious (pointing only to a modern web site). If it is possible to find out where it originally came from, that would be an improvement. The image of the display appears to be properly sourced and public domain, and the image of the code symbols is likely uncopyrightable and in any case is adequately licensed.

Overall, this requires significant revision to make all text sourced, to move all claims into the body of the article making the lead be purely a summary of body content, and to remove unnecessary off-topic detail. However, I think if these things are done, this can eventually become GA; it is not so far from the criteria to be an immediate quick fail.

—David Eppstein (talk) 01:37, 2 January 2020 (UTC)

@Spinningspark: You've edited since you were notified of this review, but not much and not since January 4. In the meantime I notice Talk:Porcupine (Cheyenne)/GA2 has failed. This one is likely to fail as well unless we get significant progress soon, or at least a good reason for delaying longer. —David Eppstein (talk) 08:05, 14 January 2020 (UTC)

David, sorry for the delay. As you noted, I have been away from Wikipedia for most of December and this month. I'm working on it now, starting with moving the page as you suggest. This review page will get moved too. SpinningSpark 10:52, 14 January 2020 (UTC) There may be a further delay if I need to retrieve the Holzmann & Pehrson source which I had on library loan, but is now returned. SpinningSpark 12:34, 14 January 2020 (UTC)

I'm trying to go through your points in order in my editing, but I'd like to discuss some of the things you have indicated are irrelevant outside of that;

1. I think you have misread Bregeut's grandfather became a watchmaker. Bregeut inherited the family watchmaking business. This is about Breguet, not the grandfather. I think Bregeut's technical background is relevant to his ability to design telegraph instruments.
2. The arguments in favour of the optical telegraph and against the electrical telegraph are highly relevant to this story, as is the widespread use of the optical telegraph in France. This explains the late development of the electrical telegraph in France, and the unique direction the design took when it was implemented.
3. The spread of the Morse system outside France is relevant to the demise of Foy-Breguet, but I'll look at cutting back on that material as it is adequately covered elsewhere.

SpinningSpark 12:34, 14 January 2020 (UTC)

David, I don't understand why you say the lead is the "only place that that contains a clear description of this system." It seems to me a fair summary of material in the final para of the Development section. I'm about to rework this material and move it further up the page for better flow, but will hold off until you reply to avoid causing confusion. I'm not saying I won't change anything, just that I need to understand the problem first. SpinningSpark 22:08, 14 January 2020 (UTC)

There is a vague description of the electromechanical operation of the system in the lead sentence "The system used two-needle instruments that presented a display". We at least learn that the components of the machine include two needles, and that those needles were somehow visible to the operators when decoding signals. We do not learn anything about what operators did to the machine to cause it to encode signals. We do not learn anything about how the machine transformed those operator actions (whatever they were) into electrical signals. We do not learn anything about how the signals were encoded electrically. We do not learn anything about how the machine decoded those signals back into the movement of its needles. We do not learn whether the needles moved, and their motion described the symbols in a message, or whether they moved to specific non-moving positions and those positions described the symbols. All we have to go on is that there were two needles, (from the figure, never described in text) that those two needles were somehow visible through an H-shaped set of grooves in the front panel of the machine, and that somehow the position of the needles could be transformed into shapes that were decoded into letters. Or maybe I'm misreading the figure and what I think are grooves are really the needles, but then why are there three of them rather than two? Even the shapes are quite mysterious: they are not merely a subset of the grooves in the panel. For instance, the A shape has a sharp angle at its left side that does not exist in the depicted grooves of the figure. So I think the "what it is" part of the article is seriously deficient. —David Eppstein (talk) 01:28, 15 January 2020 (UTC)

Ok, I'll get on to the deficiencies of the description of the operation later. What I am trying to address here are your comments about the lead containing information that is not a summary of the article body. I'm still not clear what you think is in the lead that is not in the article (on this one narrow point). SpinningSpark 00:56, 19 January 2020 (UTC)

Let me quote from the first line of my previous reply: "The system used two-needle instruments that presented a display". —David Eppstein (talk) 01:21, 19 January 2020 (UTC)

Sorry for being really thick here, but is your problem that the lead is the only place that states the device had two needles? That is not the case, in the body there is The Breguet design required only two signal wires, but at the expense of having only two moveable needles. These represented the indicators of the Chappe system. SpinningSpark 11:54, 19 January 2020 (UTC)

I've looked at the issue you raised on use of tense and can't really see any inappropriate uses. There are a couple of places where past-future tense is used such as Foy predicting how well the Morse system might work, but other than that it looks consistent to me. Perhaps you could give the specific examples that are troubling you. SpinningSpark 23:34, 29 January 2020 (UTC)

Now, more than a month after the initial review and more than a week after our last interaction, development on the article has stalled. It now looks really unbalanced, with an overlong "Development" section (relative to the rest of the article; if the rest were expanded that might not be problematic) and an empty "Operation" section. I don't think this is converging quickly enough to a passable state to keep this review on hold like this any longer. I'm going to fail it, in order to take away an artificial deadline for the revisions to be completed, but without prejudice against re-nomination once the article is back in a more completed state. —David Eppstein (talk) 08:25, 8 February 2020 (UTC)

Thanks David. That's a good decision. I'm barely finding time to keep up with my watchlist at the moment and can't devote the time to this that it really deserves. I'll ping you when I think I have it in a reasonable state. SpinningSpark 15:15, 9 February 2020 (UTC)

I don't think a ping is necessary; just nominate it again. —David Eppstein (talk) 20:11, 9 February 2020 (UTC)