Wikipedia talk:Featured article candidates/archive74

Source review review
I've recently done one basilar source review on Featured article candidates/James Wood Bush/archive1 and a few more thorough ones on Featured article candidates/Project E/archive1, Featured article candidates/Brachiosaurus/archive1 and Featured article candidates/Green Park tube station/archive1. Did I get the source review procedure right? Jo-Jo Eumerus (talk, contributions) 06:23, 25 October 2018 (UTC)
 * Er..what is a "basilar" source review? Brianboulton (talk) 13:47, 25 October 2018 (UTC)
 * "Basilar" means that I took a laundry list of all sources used and investigated them (publishers, authors etc.) to see if they were reliable and appropriate. Jo-Jo Eumerus (talk, contributions) 14:12, 25 October 2018 (UTC)


 * Source reviews have typically covered both reliability and formatting - it looks like you've covered the reliability but not the formatting. It also typically checks that everything that should have a source does, and on the tube station article there appear to be some unsourced statements (one is a summary of later paragraphs but two more don't seem to be). Nikkimaria (talk) 00:01, 26 October 2018 (UTC)
 * Seems like I got these now. Jo-Jo Eumerus (talk, contributions) 14:45, 26 October 2018 (UTC)

Mentorship request
Hello, Iv been trying to make Mahavira FA, but has not been successful so far. Recently I was informed about the mentorship program. I would love to get some good and elaborate guidance. Thanks in advance Capankajsmilyo(Talk 11:54, 28 October 2018 (UTC)

FAC mentorship request
Hi, I'm looking for help from an experienced FA editor to assist me in hopefully getting my first FA with Cardiff City F.C.. I nominated the article in September 2017 but it was closed due to a lack of attention and I was put off slightly in pursuing it. I've decided to have another crack and I'm hoping the involvement of someone who knows the process will make it smoother. Thanks. Kosack (talk) 10:08, 4 November 2018 (UTC)

Proposed close for the source review RfC
I propose to close the source review RfC tomorrow with what I believe is an uncontroversial outcome. For details, see here. If there are any objections please say so; and if you have not expressed an opinion yet and wish to, please go ahead. Mike Christie (talk - contribs - library) 00:41, 4 November 2018 (UTC)
 * Good. Tony (talk)  03:21, 4 November 2018 (UTC)
 * Ditto, with the proviso...no more process. I'm experienced with FAC but hardly as skilled with bits and bobs as the likes RexxS; for eg the PR noming process puts me off. Ceoil (talk) 07:07, 4 November 2018 (UTC)
 * An experienced closer has indicated they would close it with the "low-process" outcome, and we've just had one more support for that too. If I do close it myself that's how it will go, but I may leave it to the closer, though they're suggesting waiting the full 30 days (which would take another couple of weeks). Mike Christie (talk - contribs -  library) 10:35, 4 November 2018 (UTC)
 * . The "30 day rule" (guideline) is not merely an exercise in bureaucracy—I am as WP:NOTBURO / IAR as they come—but a way of enabling our less-active colleagues an opportunity to comment also. This RfC has, after all, been listed at WP:CENT, as well as two high-profile boards at the VP, so it is reasonable to expect that those editors who only edit at weekends, or even fortnightly, should be able to comment. That is the purpose of the month opening. As has noted, there is no deadline, and we are all volunteers. And since you wanted a watchlist notice—which could have brought hundreds of participants, few of them knowledgeable about the ways of FACs, but many experienced in Wikipedia's processes—one can only assume that you want as many participants as possible. Some of whom may be shut out by an early close. ——  SerialNumber  54129  10:55, 4 November 2018 (UTC)
 * OK, that's fair. Let's wait. Mike Christie (talk - contribs -  library) 11:02, 4 November 2018 (UTC)

Supports & opposes data
Per a request above, here is detailed support/oppose data, for the last month, and for all of 2018 to date. I've also included 2013, as the oldest year for which I have data, for comparison. These numbers only refer to content reviews, not to supports or opposes based on sourcing or images. Mike Christie (talk - contribs - library) 11:15, 4 November 2018 (UTC)

Discussion
The data above is so lengthy it slows down the editing window, so I'm splitting out a discussion section in case anyone wants to comment. I have now added all the years I have data for. One thing to notice is that the number of reviewers actually rose significantly in 2017, which is encouraging. Mike Christie (talk - contribs - library) 15:53, 4 November 2018 (UTC)

FAC reviewing statistics for October
Here are the FAC reviewing statistics for FACs ending in October.

There were 67 declarations of support, 13 opposes, two struck opposes, and one oppose that was converted to a support. Mike Christie (talk - contribs - library) 09:56, 3 November 2018 (UTC)
 * Thanks, as always. I must say I don't really like my reviews being described as "Type: Prose". They certainly aren't that in the way say Tony1's are. I always try to review the content. I think this is actually the area where FAC is weakest, and not having a "type" column for it perhaps demonstrates this. Johnbod (talk) 03:09, 4 November 2018 (UTC)
 * Mike, If its not too much bother, would like to see the opposes listed as a column. My strong opinion is that this is a minority activity, giving a result of lowering of the bar and clogging the process. FAC needs to be more aware of the gate-keeping and standard setting aspect - I often get get but x article passed thrown back, which is a race to the bottom. Separately, not sure how we can identify the type of expert reviews say Johnbod mentions and provides in his area, which of course are hens teeth. Frankly, I think one of these is worth 20 image reviews. Sorry Mike or Nikki for the angst; long day, and not at you; your energy here is badly needed. Ceoil (talk) 07:03, 4 November 2018 (UTC)
 * Will list the supports and opposes in a few minutes; it's not much trouble. Johnbod, I agree; I just use "Prose" to make it clear that it's not a specialized source or image review, but I like "Content" better and will try to remember to use that.  Have changed it above. By the way, I have now extended the historical data back to late 2012; I'm still planning on going back further but I have enough to do comparisons back that far. Mike Christie (talk - contribs -  library) 10:32, 4 November 2018 (UTC)
 * Thanks! Johnbod (talk) 15:40, 4 November 2018 (UTC)
 * , thank you for all your work pulling this together. I have another request, but I don't know how much extra work it involves. Some of the prose reviews are indeed just that, and the reviewers say so. Is it possible to make clear which were acknowledged to be prose reviews only and which content reviews? In other words, I'd like to see these categories: source review, image review, prose review, content review. (A content review ideally includes checking prose, sources, source-text relationship, structure, whether it's comprehensive etc). If that's too much work, please ignore. SarahSV (talk) 00:49, 6 November 2018 (UTC)
 * Sorry, I don't have that breakdown, and it can't be acquired with any reasonable amount of work. I've had to make some compromises just to get the data I do have, and broad-brush categories is one of those compromises.  I'll write up a more detailed list of caveats about the data at some point, but the short version is that any marginally helpful comment, however short, counts as a review, and the only distinctions are source review, image/media reviews, accessibility reviews (very rare), and everything else.  That makes the data of little use for the kind of question you're asking, I'm afraid. Mike Christie (talk - contribs -  library) 02:27, 6 November 2018 (UTC)
 * Is there a specific purpose behind all of the requests you get, and for collecting this data? I suppose this does not fall under big data, but when I was in charge of many systems I made sure that every request for data came with what the person was going to do with the data. Just want to make sure you are not doing a lot of extra work for low benefit!  Kees08  (Talk)   04:03, 6 November 2018 (UTC)
 * Okay, thanks, Mike. SarahSV (talk) 04:17, 6 November 2018 (UTC)
 * Kees08, there were several extended discussions in 2016 (see this WT:FAC archive) which prompted me to start capturing the data going forward, partly in order to do the monthly reviewing statistics. I've been involved in many discussions about what to do about FAC over the last twelve years, and I've often thought that those discussions were full of assertions I wanted evidence for -- "citation needed", if you will.  Some assertions were clearly true, such as "There are fewer FAC reviewers than there used to be"; or "FACs take longer now"; some were less clear, such as "reviewers these days are less inclined to oppose".  In both cases I wanted to put numbers to the assertions, so in my mind I'm building a reliable source for discussions at FAC about what works and what doesn't.  Other uses spring to mind, such as adding review counts after people's names, as is done at GAN (this was suggested in 2016 but not agreed to), or building a toolserver page to produce a FAC history for an editor, rather like the edit count page.  If I ever get the dataset completed back to 2003, I could also do reports that show the correlation between FAC experience and success at FAC.  There's a rather depressing table in the archive linked above that shows that correlation, done by hand. Mike Christie (talk - contribs -  library) 10:39, 6 November 2018 (UTC)
 * This seems like a good opportunity to thank Mike for tirelessly generating and proving these stats. Its a very interesting and valuable set of contributions indeed. Thanks Mike. Ceoil (talk) 22:12, 9 November 2018 (UTC)
 * Thank you, Ceoil, that's much appreciated! I do it because I think the data is useful, but it's nice to know that others agree. Mike Christie (talk - contribs -  library) 23:06, 9 November 2018 (UTC)

Templates in FACs
The FAC instructions say not to include templates; I recall this was an issue years ago. I've been guilty of using tq extensively in my reviews as I'd forgotten this was still in the instructions. Does it need to be? Are computers now fast enough that we can ignore this? Personally I'm not keen on the done template, but that's just because I don't think it adds anything useful to the word "Done". Coloured quoting, as provided by tq, xt, and !xt, seems fine to me, if the speed problem is no longer an issue. Mike Christie (talk - contribs - library) 14:14, 28 October 2018 (UTC)
 * It's not a speed issue - it's a page limit issue. There is a limit on the number of templates that can be displayed on a page - it doesn't come up on the actual page itself, but it used to impact the archives of nominations - because of the number of FACs that were transcluded there - the pages would run into the template limit. Ealdgyth - Talk 14:17, 28 October 2018 (UTC)
 * If I recall correctly, Geometry Guy once pointed out in a conversation like this that it was nested templates that cause the problem. Didn't we once have a conversation that demonstrated that non-nested templates were not going to be an issue?  I just searched the WT:FAC archives and can't find it; I'll dig around some more. Mike Christie (talk - contribs -  library) 14:20, 28 October 2018 (UTC)
 * No clue, I was just told it was a page limit thing with the archives. Personally I detest that gods awful geeen quote thing and thing it makes thing harder to read and makes pages look like they have leprosy. Ealdgyth - Talk 14:25, 28 October 2018 (UTC)
 * It was introduced in January 2008 because the template transclusion limit was cutting off the oldest noms: Wikipedia talk:Featured article candidates/archive26 (and it was slow to load). DrKay (talk) 14:34, 28 October 2018 (UTC)
 * WP:DYK encountered a similar problem a few years back and it was rather disruptive leading to the creation of the approved page and associated bot. I would rather keep the general prohibition on templates and avoid having to think about the transclusion limit, though would suggest substing if you really need to use one and it's not particularly complicated. Wugapodes [thɑk] [ˈkan.ˌʧɹɪbz] 18:08, 31 October 2018 (UTC)
 * Since a substantial part of all material in a given FAC is quoted text, I support a way to differentiate quoted material. I find it easier to read in the mess of bullet points, indented reply bullet points, confirmations of "doneness", and so on. I'll take even subtle differentiation. Is there a simple solution other than templates or subst'd templates? I can't think of one. A bot could subst certain templates once they hit the archives, for example, if there is still a technical problem. Outriggr (talk) 00:31, 1 November 2018 (UTC)
 * Yep. "Leprosy" aside, having a way to distinguish quotes from the article or sources from other text is extremely helpful to me. Der Wohltemperierte Fuchs (talk) 19:21, 1 November 2018 (UTC)
 * What he said. I kind of agree about the leprosy, and wouldn't be heartbroken if they changed the template such that it highlighted the text some other way, but in terms of function it's invaluable, given how often it's necessary to make it clear exactly which point is being commented on. As I understand it, it tends to be citation templates that push pages over the transclusion limit (because each citation template in turn transcludes multiple other templates); RexxS will know the technicalities. &#8209; Iridescent 20:49, 1 November 2018 (UTC)
 * If you can come up with the styling you'd like to highlight the text in some other way, I could easily make that into a template for folks to use. You're right that templates that transclude other templates are the ones most likely to hit a limit (maximum for "Highest expansion depth" is 40), although using Lua modules to implement citation templates removes much of the problem these days. The other template limits that might have once been a consideration currently allow up to 2 million characters passed to or produced by templates per page, so that's quite unlikely to be an issue. I doubt that using tq or xt or anything similar is going to hit the present limits. --RexxS (talk) 23:13, 1 November 2018 (UTC)
 * It sounds like we can update the instructions to remove the stricture on templates. If nobody objects (or beats me to it) I'll make that change this weekend.  (Ealdgyth, I promise not to use them on your FACs.) Mike Christie (talk - contribs -  library) 08:52, 3 November 2018 (UTC)
 * My advice would be to leave 's discouragement of the "done" and "not done" graphics that seems to have enjoyed consensus for the last 10 years. No amount of technical fixes will alter the aesthetic arguments against graphics like those. --RexxS (talk) 10:42, 3 November 2018 (UTC)
 * I agree re. maintaining a general prohibition on the done/not done templates. Whereas highlighted text (incidentally,    without leprosy or template!) does help distinguish text elements, I feel the icons themselves actually do the opposite: they are bigger than the text that follows them, and so inconsistently increase line spacing, and generally make text harder to read through interruption.Incidentally, apart from apart ✅ and ❌, you'll all be glad to hear we've also got , ,  and ...  ——  SerialNumber  54129  11:04, 3 November 2018 (UTC)
 * I'm not fond of the done/not done graphics either, but can we justify a proscription on aesthetic grounds? I can't think of a parallel to that sort of instruction elsewhere in non-article space. Mike Christie (talk - contribs -  library) 11:17, 3 November 2018 (UTC)
 * Probably not! But I don't think I (at least) did emphasise aesthetic grounds (certainly unintentionally): as I say, they actively interfere with the text, and thus they could easily interfere with the reviewing itself. I'd also suggest that a page covered in loads of multi-coloured blobs is likely to deter, rather than encourage, potential new reviewers to join a discussion—and that, surely, is the prize :) ——  SerialNumber  54129  11:23, 3 November 2018 (UTC)
 * Definitely! On Wikipedia we don't have prescriptive instructions. We have instructions that document our agreed best practices. The prohibition on gaudy graphics has enjoyed consensus for 10 years. The grounds of aesthetics are supplemented by grounds of functionality (poorer readability through distraction and increased line-spacing) and grounds of unnecessarily increased bandwidth usage for those with poorer or metered internet connections – a single tick mark costs more than 750 characters of text. --RexxS (talk) 11:59, 3 November 2018 (UTC)
 * Definitely. The purpose of those templates was for multilingual projects like Commons, where readers won't necessarily know what "abgeschlossen" or "heb ei wneud" means without a visual aid; they don't really have a useful purpose at a monolingual project like Wikipedia other than as a "look at my post, it's more important than yours" tool. &#8209; Iridescent 23:59, 3 November 2018 (UTC)
 * It was, historically, mostly a page limit problem, but it was not that exclusively. Reading through cluttered FAC pages is made unnecessarily difficult by the addition of templates, sub-headings, and the like.  Done, not done, etc mean little to the delegate/coordinator who should be deciding such things on their own, rather than looking for checkmarks.  Plain text is easier to sort; sub-headings, templates, extraneous colorful junk are sometimes used to mislead other reviewers, while only making the page more difficult for the one who has to look at what is actually done, and decide consensus. And today's FAC pages are much longer and more cluttered than they once were. (Thanks for the ping, as I rarely peek in here anymore ... still working on recovery from the tree that fell on me in a hammock-- curiously, I promoted that FA and it came back to whack me!) Sandy Georgia  (Talk)  15:48, 5 November 2018 (UTC)

Now done, with language that I hope reflects the discussion here. Mike Christie (talk - contribs - library) 20:06, 11 November 2018 (UTC)

Half a shekel for an FAC coordinator?
I'm just dropped in to beg anyone who's willing to take a look and help clear the image/source check requests above. One has been waiting for a spot check for more than a month. I can't offer much but I could perform select numbers from Camelot over Skype for the amusement of your family. -- Laser brain  (talk)  11:56, 16 November 2018 (UTC)
 * I might try one today if I get Salar de Punta Negra finished today. Jo-Jo Eumerus (talk, contributions) 07:09, 19 November 2018 (UTC)

Bot report
FACBot notes that Featured article candidates/List of Kolkata Derby Matches/archive1 has not been transcluded. However, I think that it should go to FLC, not FAC. Hawkeye7  (discuss)  00:40, 13 November 2018 (UTC)
 * Bot still reports this one every day.  Hawkeye7   (discuss)  02:27, 28 November 2018 (UTC)
 * Pinging, the creator. SabyaC, I suggest you request speedy deletion for this and recreate it as a featured list nomination, rather than a featured article nomination, following the instructions at WP:FLC. Mike Christie (talk - contribs -  library) 11:38, 28 November 2018 (UTC)
 * Featured article candidates/Meghan Trainor/archive1 has not been transcluded on the nomination page.  Hawkeye7   (discuss)  00:47, 19 November 2018 (UTC)

Source review workshop: status and next steps
See above for the background to this; for brevity I won't repeat much of that here.

The source review workshop has effectively concluded. The workshop page is here. Five articles were submitted for review; two passed, two failed, and one has had no review yet. Some comments on the reviews:
 * Roger B. Chaffee. Nominated by .  I reviewed this; I'm not particularly experienced as a source reviewer but I was as thorough as I knew how.  This passed; the review was short and painless.
 * Lion-class battleship. Nominated by .  No review.  I think/hope that this is just because this was a workshop, not a real process, though we do have plenty of nominations that hang about waiting for source reviews here at FAC.
 * Me, too! Although maybe it's a function of my mad sourcing skillz! ;-) Sturmvogel 66 (talk) 22:48, 8 October 2018 (UTC)
 * Parliament of 1327. Nominated by . A long and extremely thorough review by  -- far more thorough than most FAC reviews.  This passed and the review was copied to the FAC as evidence of an existing source review.
 * Saving Light. Nominated by .  Reviewed by both  and .  Failed; Laser brain noted "Concerns include issues with WP:RS, verification, and close paraphrasing that require further work."
 * Shannen Says. Nominated by .  Reviewed by .  Failed, without any particular problems noted; the review is incomplete.  Laser brain noted in his closing that this was a case that would need discussion if we go ahead with a separate source review process.

The goals of the workshop were to determine if a source review process could be:
 * Well-defined: what is reviewed, how are reviews done, and what are the pass/fail criteria?
 * Useful: does it eliminate work elsewhere, rather than duplicating work?

I believe the answer is yes to both questions. Source reviews are already taking place at FAC so the question of definition just means identifying the criteria at issue. The workshop took 1c and 2c from WP:FACR as the requirements:


 * (1c): well-researched: it is a thorough and representative survey of the relevant literature; claims are verifiable against high-quality reliable sources and are supported by inline citations where appropriate;
 * (2c): consistent citations: where required by criterion 1c, consistently formatted inline citations using either footnotes or Harvard referencing (Smith 2007, p. 1)

The reviews are useful: one of the passed articles was cited in that article's FAC, and I expect the other passed article will do the same. If Saving Light should come to FAC, I think a reviewer would be justified in saying "what about the sourcing issues that were identified?" and suggesting withdrawal if they had not been addressed. If we have a separate source review process we can go further and insist that the article pass the source review before absorbing reviewer time at FAC. , and others have repeatedly pointed out that polishing prose based on unreliable sources is like painting over rotted wood. Let's make sure the wood is sound before we bring the article to the paintshop.

Some points that would need to be hammered out if we adopt this process:
 * Do we allow multiple simultaneous nominations? Or immediate renominations of fails?  I'd suggest we start with the same rules FAC has, and tweak if necessary.
 * How long can an article wait before it is failed as incomplete? I think we should leave this to the coordinators, who can choose to nudge the reviewers to try to get a declaration, if they think fit.  See the review for Shannen Says, listed above, as an example.

The next step is to get feedback from the community in the discussion section below. If there's enough support, I'll draft an RfC page for further discussion. At this point it might also be good to ping in others who might be interested. For example, the FLC coordinators,, , and , may not have seen this discussion, and FLC might be interested in using a separate source review process such as this. (Note: this section was deleted and re-added to make the pings happen, because I screwed up the pings first time round.) Mike Christie (talk - contribs - library) 23:19, 8 October 2018 (UTC)

Discussion

 * I'm a bit confused by the result re: Shannen Says. I did the same source review I would have done at FAC. Is the review considered incomplete simply because I didn't add Support/Oppose at the end? Or was I meant to do a spotcheck? Some other reason? Nikkimaria (talk) 00:51, 9 October 2018 (UTC)
 * I was looking for feedback on that one. I know your reviewing style and I would of course accept your review as comprehensive at FAC, but if this is going to potentially live outside FAC, do we need to ask reviewers to make some kind of definitive statement of support? -- Laser brain  (talk)  02:51, 9 October 2018 (UTC)
 * Would it need to be a "support"? Would some statement of closure, "Passed source review" suffice? Image and source reviews are the only reviews which can individually sink a nomination, in contrast with a single prose oppose; I always think of them as absolute reviews, which need to be passed, not just changed to get support. – SchroCat (talk) 06:10, 9 October 2018 (UTC)
 * That would be a reasonable move. I do not generally say "support" after an image review but that does not mean that the images aren't fine. Jo-Jo Eumerus (talk, contributions) 06:14, 9 October 2018 (UTC)
 * Yes, in the context of a normal all-encompassing FAC, I assume that if all points raised by a source or image reviewer have been acknowledged/actioned and it's evident that the reviewer is satisfied with the responses then we're good to go. I don't think we need "support"s as we do for other aspects of review, but I'd be happy to see a clear "pass" or some such comment to confirm the source or image review is okay -- that would be helpful whether we go for separate source and image review stages, or stick with our current all-through reviewing system. Cheers, Ian Rose (talk) 06:21, 9 October 2018 (UTC)

I would prefer source review to occur first. I have John Glenn going through the FAC process, which has almost 300 citations. While it has gone through a prose review and a partial image review, since it has not received a source review it could be a long ways from FAC.  Kees08  (Talk)   06:24, 9 October 2018 (UTC)
 * On the larger question of whether we should separate source and image reviews from the rest of the FAC process, I can see some advantages and disadvantages so I'm keeping an open mind and await further community input. If we were to adopt this concept then, yes, I think it would make sense to do the source review first. On the question Mike raises of whether we would permit multiple noms for the source/image review stages, I tend to agree we should stick to the principles we have at FAC now, i.e. one nom per person at a time, two if one is a co-nom. I don't think it makes much sense for people to build up a cache of source/imaged-reviewed articles if we continue to only permit one at a time at the general FAC stage, because the 'parked' articles could change between the source/image review stage and the general FAC stage. Similarly, the two-week pause after an unsuccessful nom seems to make sense for the source review stage (it may not be quite so big a deal for the image review stage). Cheers, Ian Rose (talk) 07:38, 9 October 2018 (UTC)


 * I think this would take FAC in the wrong direction., your review seemed (to me) unfinished in part because you didn't say whether you'd checked the text against any of the sources, and because you didn't reply when the nominator asked whether you had any further comments. I assumed that you were planning to return to it. If this does go ahead, it would make sense for reviewers to say pass or fail. , this two-step process would work if we had enough reviewers willing and able to offer thorough reviews. But we don't, so you're going to end up with inadequate reviews and/or a bottleneck as nominators wait for the attention of the tiny number willing to do it. (And note that there might be more source reviewers at the start of a trial because of the novelty factor; you have to imagine how many there will be in two years' time.) There would be two other consequences of this separate process. (1) It would make FAC more GAN-like, in that moving an article to the second stage would depend on a single editor's opinion, so that other editors may not even get the chance to review it; and (2) it will make other reviewers even less likely than they are now to check the sources, because they'll assume that it's been done. We should be trying to get everyone involved in checking sources. SarahSV (talk) 20:28, 9 October 2018 (UTC)
 * That goes to my question above - is this process meant to include what has been called a "spotcheck" at FAC, or not. Nikkimaria (talk) 00:20, 10 October 2018 (UTC)
 * 1c says: "claims are verifiable against high-quality reliable sources". Mike also didn't do a spot check. I can't see the point in creating a new source-review process that doesn't ask reviewers at least to try to make a spot check. SarahSV (talk) 01:00, 10 October 2018 (UTC)
 * (ec) Nikki, I think it should include that, yes, but I think spotchecks at FAC should still be encouraged -- editing prose can cause it to drift away from correctly representing a source. Currently we require spotchecks at FAC for new nominators, and I think we should continue to do so. (Post ec: Sarah, I did a spotcheck but did not note it.) Sarah, I have no crystal ball, but here are some thoughts. We already have that bottleneck, and having it inside FAC does not seem to be driving many people to source reviewing. I think placing the bottleneck before FAC will make it more likely that nominators will review. Here at FAC we can tell ourselves we don't need to do source reviews because we've done lots of prose reviews, but a separate source review process would make it clearer that that's a fallacy. Your two numbered points would be serious problems if they turn out to be true, but I suggest (a) any article nominated at FAC should disclose any sources added since the source review, and those should be reviewed at FAC (and if the new sources are too numerous the coordinators can require a return to the source review process); and more importantly (b) any editor can do a source review at FAC if they wish to. I've seen experienced source editors here add second source reviews when they thought they were necessary, and that practice can continue. I don't think our best reviewers will hesitate to challenge a source if they think it's unreliable or has been misused, just because another reviewer passed it. It's a minimum bar: you may say that's too low a minimum, but it's no lower than it is now.
 * The main benefit I hope will accrue is that we will no longer waste prose/MoS reviewers' time on reviewing material that is poorly sourced. I have skimmed through over a thousand FAC nominations in the last month or so, harvesting support/oppose data, and I can tell you that many, many hours have been spent by reviewers on reviewing articles that would have not made it to FAC had they undergone a source review first. Mike Christie (talk - contribs - library) 01:15, 10 October 2018 (UTC)
 * (ec) As things stand, the FAC standard for source review may be lower than the GAN and DYK standards.
 * Reviewing good articles: "At a bare minimum, check that the sources used are reliable ... and that those you can access support the content of the article ... and are not plagiarized ..."
 * Did you know/Reviewing guide: "Check that the article does not contain plagiarism or close paraphrasing" and "If the article includes information about living individuals, make sure it does not violate Wikipedia's policy on biographies of living people."
 * Guidance on source reviewing at FAC: "Reviewers should carry out spot checks to ensure that sources have been used appropriately, that the sources do indeed support the text, and that the article contains no plagiarism, including close paraphrasing without in-text attribution. The extent to which spot checks are pursued is a matter for each reviewer. ... The FAC coordinators will usually require spot-checking for first-time nominations.
 * I think we should remove the final sentence and stop requiring them only for first-time nominations. SarahSV (talk) 01:20, 10 October 2018 (UTC)
 * I might support that (I'd want to see some discussion) but I think there's no linkage between that and having a separate source review process. Either change could be implemented without the other. Mike Christie (talk - contribs -  library) 01:31, 10 October 2018 (UTC)
 * , I agree, but I think we need to remove it either way. We could say something like: "The extent to which spot checks are pursued is a matter for each reviewer, but the FAC coordinators will not promote an article unless they are satisfied that spot-checking has been done." Either that or we should remove from 1c: "claims are verifiable against high-quality reliable sources", and make clear that FAs should be well written with consistently formatted citations to reliable sources, but that we don't care whether the sources actually support the text. SarahSV (talk) 15:53, 10 October 2018 (UTC)
 * I don't disagree, but I'd rather the discussion of altering 1c get its own heading at some point. I've found startling errors when I conduct spot-checks of even experienced nominators. -- Laser brain  (talk)  16:01, 10 October 2018 (UTC)
 * Spotchecks of sources have never been required of every FAC. I would prefer they were, but the community, several of whom have remarked on the time it takes to get articles through the process, will need to decide if it wants that. At the moment the coords require a spotcheck for newbies, and for old hands who've been out of the game for a while, the rationale being that lessons will be learnt from those checks. It's true though that even the most experienced nominators can make errors in this regard, and several regular reviewers do some spotchecking as a matter of course (I do myself occasionally as part of my copyedits) but I would not object if we made it mandatory for every review, the caveat being that by their nature, spotchecks can't be guaranteed to eliminate every error of accuracy or close paraphrasing, they can simply give a feel for how good the article is in that respect. Cheers, Ian Rose (talk) 22:06, 10 October 2018 (UTC)
 * Proper spotchecks are not required by DYK and GAN either (and to claim "FAC standard for source review may be lower than the GAN and DYK standards" just isn't true). If you actually read the requirements, they ask for no close paraphrasing: that's not whether the text reflects the source claimed (except for a BLP), but a check for plagiarism. In other words to run the article through Earwig's Copyvio Detector and accept or reject on that basis. I think we already pass this (fairly) low hurdle as it stands. In other words, 1c. is still a higher hurdle than any other review process we have. - SchroCat (talk) 22:24, 10 October 2018 (UTC)

Here are a couple of examples of FACs that received multiple supports prior to a source review that ultimately sank the nomination. I can provide more examples if needed. -- Mike Christie (talk - contribs - library) 02:10, 11 October 2018 (UTC)
 * Bentworth, 2nd nomination. Six supporters, including some very experienced and capable nominators, prior to an oppose on sources by Brianboulton.
 * Of Human Feelings, 1st nomination. Several reviews, including a long prose review, and a source review that did not include spotchecks, prior to an oppose on sources by Quadell.
 * Do you think it is time to put this up to a vote? How do we want to make this final decision?  Kees08  (Talk)   00:52, 14 October 2018 (UTC)
 * I'll create an RfC page and post a link here when it's ready for review, before we launch it. Mike Christie (talk - contribs - library) 12:40, 14 October 2018 (UTC)
 * Sorry, I've been lazy about keeping up with these threads. It seems important to spot-check for close paraphrasing and straight plagiarism—and that would-be nominators know that will happen. It's clearly important for first-time nominators' work, but also important that the risk of a spot-check not be absent for subsequent noms by editors, even it spot-checks occur less often for them. Spot-checkers need online access to books and academic journals, right? That means most of us can't do it properly. Tony (talk)  05:02, 15 October 2018 (UTC)
 * Sometimes WP:RX can help with non-online sources. Jo-Jo Eumerus (talk, contributions) 06:08, 15 October 2018 (UTC)
 * I've sent through copies of articles and scans of relevant pages that have been requested by reviewers before. It's not too much extra work to send through a selection of requested material. - SchroCat (talk) 06:38, 15 October 2018 (UTC)

"far more thorough than most FAC reviews"
Was my review too thorough? Factotem (talk) 20:04, 9 October 2018 (UTC)
 * I don't think so; I said "more thorough than most", but there are plenty of equally thorough reviews in the archives. Brian has done some, and further back in time Squeamish Ossifrage did very detailed source reviews. I think any article would benefit from that level of detail; few reviewers can spare that much time on a regular basis, though. Mike Christie (talk - contribs - library) 01:15, 10 October 2018 (UTC)

RfC draft is up; please comment
I have drafted an RfC page here; please comment, or jump in and edit. Once it appears to be stable I will make it live and post notices at the usual places, perhaps including a site notice if we all agree it's sufficiently significant. Mike Christie (talk - contribs - library) 01:39, 16 October 2018 (UTC) Mike, thanks heaps for this. A few comments (and I've tiddled with a bit of wording on the page, withough substantively changing the meaning, I hope). Generally it would be nice to reduce the size of the text (by 20%?).
 * Have FLC people been warned of this? Are they being made to feel part of the process? The lead doesn't immediately clarify whether the new sourcing check would audit both FACs and FLCs, but readers will automatically think it does. Later we find that it's just a teaser for FLC. But in my view this complicates matters. It's easy enough for FLC to take on the process later if they see benefits (this could be mentioned toward the bottom).
 * There's no mention of the current perceived deficiencies at the top (but maybe my expansion of the "in a nutshell" banner does this?).
 * Generally I think it's all a bit long.
 * FQSR sounds like SPQR from ancient Rome: a mouthful. Any reason not just FSR, dropping ''quality"?
 * We going with "would" or "will"? I see both.
 * "if such an article is nominated without asking for leave, a coordinator will decide whether to remove it"—In my view this needs to be fixed for FAC itself. It invites rule-breaking. "if such an article is nominated without asking for leave, a coordinator will decide whether to remove it".
 * This bullet: " "Spotchecking" is FAC shorthand for checking sources against the article text to ensure the text is supported by and does not plagiarize or too closely paraphrase the sources. Spotchecks are only required at FAC for first-time nominators or nominators who have been away from FAC for several years, though any reviewer can choose to spotcheck at any time.  The same will be true for FQSR."  I thought spotchecking meant a sample-check, at random; does it imply a full check? Unclear to me. And does this mean there's no source checking for second and subsequent nominations by an editor? I'd like to keep the option open, as a practical motivation through risk exposure for serial nominators. And what's the rule when one of two nominators is a first-timer?

Tony (talk)  03:54, 16 October 2018 (UTC)

PS Possible to remove lots of procedural text by simply stating that it's basically the same as for FAC (with a link to the process)? Tony (talk)  06:28, 16 October 2018 (UTC)
 * Your edits look fine to me. Serial Number 54129 just cut the length significantly by eliminating the bureaucratic paragraph about rules; I think that's probably a good move.  Some replies to your points above:
 * See this conversation. I did ping the FLC coords at an earlier stage, and will go ahead and leave a note at WT:FLC shortly.  I'm not at all sure they'll be interested, but I felt it was necessary to be explicit in the RfC statement about FAC because FSR is only likely to be used if it's a prerequisite for at least one of FLC and FAC.
 * I think your expanded nutshell does indeed take care of mentioning the deficiencies early.
 * I agree on length and have trimmed a bit more; please keep cutting wherever you see an opportunity.
 * FSR is a good acronym, though my OCD twitches at the loss of the Q: I think we need to retain "quality" in "featured quality source review", because the outcome of FSR will be featured quality sources, not featured sources.
 * "Would" is probably better; I've edited in that direction but please scan again for consistency.
 * Spotchecking does refer to a sample, not a complete check; I've added "some" to clarify this.
 * -- Mike Christie (talk - contribs - library) 08:20, 16 October 2018 (UTC)
 * Adding a note, since I see I didn't respond to your final point: typically no spotchecks are required if one or more nominators are experienced, though any reviewer may choose to spotcheck any nomination. In the discussion section above  proposed requiring spotchecks for all nominations. Mike Christie (talk - contribs -  library) 09:04, 16 October 2018 (UTC)
 * Apologies for my terse edit summary in that edit—I hit return to early and got distracted elsewhere. The reason I removed it was that, in replicating a piece of general FAC guidance, it seemed only tangentially relevant to the actual source review element of the page. On the FSR/FSQR thing; I noticed the Roman thing a while ago, and...actually quite liked it in a tongue-in-cheek kind of way. As in, there was the Empire—and then there was Rome. On put it another way: There are articles—and then there are Featured articles :D  geddit? There's no harm, I think, in a touch of light-hearted, good-humoured self-indulgence when it is also accurate and self-explanatory...my 2p anyway.  ——  SerialNumber  54129  10:08, 16 October 2018 (UTC)
 * I also saw the SPQR connection and didn't really object -- but I think shorter is better. However, I did just notice that WP:FSR is taken (though not by a high-traffic page) so we may end up with WP:FQSR. No worries on the edit; it was an improvement. Mike Christie (talk - contribs -  library) 10:18, 16 October 2018 (UTC)

Reviewers are expected to make it clear that they have fully evaluated the article on both criteria. Why force reviewers to take on the whole review which, if done thoroughly, represents a significant amount of time and effort for one person? Wouldn't allowing partial reviews spread the load and maybe encourage more reviewers?

Is it also worth adding the actual steps of a source review in the process section, i.e. Breaking it down in this way might provide a framework for consistent source reviewing and again encourage wider participation. Factotem (talk) 09:53, 16 October 2018 (UTC)
 * Technical checks for formatting (inline refs are consistent, p/pp, ISBN format, publisher locations, etc.);
 * External link checks (the most important of which I think, and the most overlooked, is checking that ISBN/OCLC/etc. links lead to the correct editions);
 * Verification that sources are both high quality and reliable;
 * Survey of sources to ensure comprehensiveness;
 * Spotcheck verification of sources for accuracy and plagiarism.
 * I agree with your first point and have edited the RfC to suit. I'm not convinced we need to add the steps of a review to the RfC, though I like the idea of adding them to the instructions for the nominations page, if the RfC passes.  I think adding them to the RfC would make it longer without helping contribute to commenters' decisions to support or oppose. Mike Christie (talk - contribs -  library) 10:23, 16 October 2018 (UTC)
 * ...so long as both criteria are fully evaluated by at least one reviewer By my reading, this still mandates that at least one reviewer must complete a full review for the review to be valid. Suggest removing "by at least one reviewer". Factotem (talk) 10:48, 17 October 2018 (UTC)
 * Yes, thought I'd fixed that already but I guess I was careless. How does it look now? Mike Christie (talk - contribs -  library) 11:03, 17 October 2018 (UTC)
 * Good. Factotem (talk) 11:06, 17 October 2018 (UTC)

I will probably start the RfC tomorrow night unless there are further edits that require discussion. Mike Christie (talk - contribs - library) 23:40, 16 October 2018 (UTC)

Hi Mike. Thanks very much for your work on this. One thing: My impression above was there's some good consensus between SarahSV, Laser Brain, and Ian Rose that it could well be a good idea to make spotchecks more mandatory for more than just newbies and people away from the game a while. I was surprised there was no reflection of this in your draft (or did I miss consensus to leave this question for later?). I've always thought it was a no brainer that these should be required for everyone. As has been said above, even experienced editors can make mistakes. I'd go further to say that if experienced editors know they're not going to be spotchecked, it's only human nature for them to sometimes be lax about putting in the extra effort before nomination to double-check the integrity of all cited information. Moisejp (talk) 01:58, 17 October 2018 (UTC)
 * In view of the scarcity of reviewers, perhaps mandatory for first-time noms, and possible for subsequent noms? Tony (talk)  03:01, 17 October 2018 (UTC)
 * That's more or less the current state. Moisejp, I omitted it because I didn't want to suggest introducing two changes at the same time -- there's no linkage between splitting source reviews from FAC, and increasing the spotcheck requirement.  If you think we should do that I'd suggest waiting till the RfC is over, and then starting that conversation here.  I agree with Tony that a shortage of reviewers is likely to be the main (or only) objection. Mike Christie (talk - contribs -  library) 10:01, 17 October 2018 (UTC)

...and since it just acquired a !vote and the text seems stable, I went ahead and made the RfC live. Please add your !votes and comments there. I'll add notifications in various places this evening, but have to head off to work now; if someone else wants to add some notifications that would be great. I would suggest WT:FLC, WT:RS, the village pump policy page, the central notifications page, and a request for a site notice, at least. Mike Christie (talk - contribs - library) 10:44, 17 October 2018 (UTC)
 * I listed at WP:CENT and posted notices to WT:RS and WP:VPR ("proposals" seemed more appropriate than "policy" to me, but feel free to add a notification there also if you like). It was already mentioned at WT:FLC. --RL0919 (talk) 18:09, 17 October 2018 (UTC)
 * Thanks. I've requested a watchlist notice. Mike Christie (talk - contribs -  library) 02:28, 18 October 2018 (UTC)

Contesting archive of FA-nominee by Laser Brain
Laser Brain wants to archive this FAC. It has two extensive reviews supporting the nomination, and two other reviews in from this past week, one of which has final actionable objections posted just two days ago, and I have since resolved them/responded to them. That would potentially make it three supports. Archiving it now would discard months of progress/work, and is rash, premature, as a review is concluding. Dan56 (talk) 16:42, 30 November 2018 (UTC)
 * FYI, the archiving has already taken place, sorry. ——  SerialNumber  54129  16:45, 30 November 2018 (UTC)
 * This candidate was open for a month. The amount of significant commentary that was added after it received two supports sheds some light on the weight the closer should give to those supports.  When a candidate has received two supports, additional reviewers should find it very close to promoteable, but it seems that subsequent reviewers found other issues.  A close after two weeks, when there is no consensus, is approriate .... waiting a month to archive, when the FAC looks like a peer review, is generous.  An archived FAC is often the fastest route to promotion, as it allows the nominator time to clean up issues off-FAC, and bring back a better prepared article.  Also, the idea that a FAC is promoteable with three supports should be dismissed; there is no such guarantee.  Further, the work that has gone into this article is not "discarded", nor is archiving "rash" or "premature"-- each review results in improvements, and all of this work will be shown worthwhile if the candidate comes back to FAC better prepared.  17:13, 30 November 2018 (UTC) (Fixing my incomplete sig.  Sandy Georgia  (Talk)  20:21, 30 November 2018 (UTC))
 * I have had FAs nominated with far less extensive reviews before than this one. And Laser Brain said "I will be archiving it shortly", but it appears it was done immediately. The candidate page says the bot will close it formally within 24 hours, no? So Laser_Brain can surely undo their own edit. The reviewer in question of the most recent activity there,, echoes my sentiment for an extension and that he was leaning to support the article. This, along with a successful image review, doesn't make a good case for an extension? Or at least an exemption from the two-week wait? Dan56 (talk) 19:59, 30 November 2018 (UTC)
 * As far as the progress being discarded: the progress of attracting reviewers was difficult, let alone the progress of getting the article nominated. Now, finally, once an engaged reviewer in Brandt gets involved to get us closer to a consensus, and is close to finishing their review, the entire thing gets closed without even a heads-up or warning of how much time there is left? Honestly, someone should have said we were working against a fucking clock and time was expiring. Dan56 (talk) 20:09, 30 November 2018 (UTC)
 * Nothing may come of this fuss I'm making. But I don't appreciate the tactlessness of this closure. So I'm making it. Dan56 (talk) 20:13, 30 November 2018 (UTC)
 * "archiving it shortly" in this context usually means that as soon as I finish a) writing this comment and b) reviewing the FAC candidates, when I make my next archive pass, where I formally transclude this FAC to the archived page, this one will be included. In this context, there are three separate steps:  a) finish reviewing the current FAC and the FAC page, b) transclude all archived FACs to the archive, and c) the bot then finishes the technicalities of closing the FAC.  Once a FAC is moved to archive, it is an archived FAC, even if there is a delay in bot processing of the final technicalities.  Just explaining the process, but adding that I don't see any lack of tact in the closure.  Even if another reviewer had supported, three supports does not equal promotion (for that matter, I once archived a FAC that had 28 supports.  One solid review finding problems offset 28 fan supports in that case).  Sandy Georgia  (Talk)  20:21, 30 November 2018 (UTC)
 * What do "fan supports" have to do with this FAC? Dan56 (talk) 20:49, 30 November 2018 (UTC)
 * I did not mean to imply that they did; I was offering an example of why this notion of three supports to promotion needs to be discarded. Sandy Georgia  (Talk)  21:47, 30 November 2018 (UTC)
 * Dan, There is little point railing against what is a 'good' close. Take a couple of weeks, go through the comments of BLZ, and see if anyone else will take a spin over it pre another run at FAC. There's no rush to get anything to FAC, and a couple more week's polishing can't hurt. Cheers - SchroCat (talk) 20:24, 30 November 2018 (UTC)
 * I already went thru the comments; did you not notice BLZ said he was "close to closing his review and offering his support"? There is really nothing left to do but wait and go thru the formality of another nomination. Assuming I will have time! I would like exemption from the two-week wait. If the problem was "minimal feedback", then that fits WP:FAC's mention of being exempt. If there was enough feedback from BLZ's review, then he should have been allowed to conclude it! Then a judgement of "not enough to determine consensus" would have been fair if the coordinator still felt that way, which is doubtful. Dan56 (talk) 20:49, 30 November 2018 (UTC)
 * Dan, its not easy and we have all been there with reviews being slow coming in, but I think Schro's advice for now is for the best. One the plus side, BLZ is one of the most astute music reviewers we have and addressing any of his points is sure to get the article sailing through on the following nom. Ceoil  (talk) 21:13, 30 November 2018 (UTC)

SandyGeorgia, I think you were right to say that the closure did not suffer from a "lack of tact" or anything like that. The closure itself was not problematic or improper; the article's FAC has been open for a long time, so I understand the archiving as a matter of course. That said, I empathize with Dan56's feeling of surprise. The candidacy suffered from long periods of stagnation without new reviewers. Even after I got on board, I reviewed section by section in detail and was sometimes delayed in my responses to Dan56. I've turned over just about every stone in the process and Dan56 has been diligent in revising the article and responding to my comments in a timely manner—again, usually more timely than I was able to be. I can understand Dan56's feeling that the rug was pulled out from under just as he was getting somewhere with a very challenging and extensive review, even if the coordinators had not acted improperly. Despite the nomination closing, I've expressed my intention to complete my review and communicate with Dan56 about what issues from my review, if any, remain to be resolved (I still have to go back and look over his latest responses before looking over the article one more time, but I was very near the end of the review).

Is it possible for the two-week waiting period to be shortened, perhaps to one week? I'm not sure if that's something that has ever been done before, so I'm inquiring from a position of genuine naiveté, but if it's possible I think it could be helpful here. Shortening to one week may help Dan56 sustain the interest he has managed to attract, while also giving him time to rustle up reviewers to ensure the new nomination won't suffer from the same purgatorial stagnation. Regardless of the length of the waiting period, Dan56, I'd recommend that you follow up with past reviewers to see if they'd be able to turn around new comments quickly, whether to express support per their past reviews or to rereview the article in light of changes. You could also reach out to other frequent FAC reviewers on musical topics, as you had with me, to see if they would be willing/able to turn around a review whenever the article's nomination reemerges. If you can get reviewers onboard, it might make a better case for shortening the waiting period. Like Ceoil said, review stagnation is no fun; I'm bummed things turned out this way, but I'm willing to help you out and I'm sure others will be too. —BLZ · talk 22:06, 30 November 2018 (UTC)
 * Maybe we could reproduce BLZ's ongoing review to the article talk page; resolve there and then Dan re-noms, referring to the talk page work in the nom blurb? I would certainly be interested in reviewing the 2nd nom. Ceoil  (talk) 22:17, 30 November 2018 (UTC)
 * Yea for shortened waiting period, resolving remaining concerns at talk page, and soliciting previous reviewers. Dan56 (talk) 23:00, 30 November 2018 (UTC)
 * Thanks Dan56, and to note the work the volunteer delegates do here is tedious, hard, sometimes unpopular but absolutely necessary to keep the overall machine ticking along; and I'm seeing a satisfactory outcome on the horizon here; given BLZ is vouching for your speedy and diligent responses, I have no doubt that you are equal to closing this out to FA level. Ceoil  (talk) 23:08, 30 November 2018 (UTC)
 * Tks everyone for their input -- having been busy in RL this past week (and more really) I've only had a brief time to familiarise myself with this one so would prefer to reserve judgement on whether we shorten the waiting period. Andy (Laser brain) will be away for about a week, I'll keep an eye on the article talk page in the meantime. Cheers, Ian Rose (talk) 01:02, 1 December 2018 (UTC)
 * To be clear, I'm not advocating shortening the two week period in this instance per se; this is very a short time in FAC cosmology (and was established for very good and material reasons, as people who know their history will remember), where well deserving noms often drift for months for lack of attention; and at the end of the day we seem to be heading towards article quality consensus here, so, you know this week its back up, or next week; whats the difference if the article will randomly go for a few weeks again after that depending on reviewer bite rate. We are dealing here as best we can with the demand on resources we have here. To be blunt, experienced nominators should realise this.  Ceoil  (talk) 01:40, 1 December 2018 (UTC)

Source review RfC discussion link
A new discussion has begun at the sourcing RfC about alternatives to the approach put forward by the RfC. If you're involved with FAC you may find it interesting; please comment there. Mike Christie (talk - contribs - library) 23:12, 5 December 2018 (UTC)

Are ISBN numbers required in the bibliography for FAC?
Is there any policy that says ISBNs are required at FAC? We're having some difficulties over at the FAC for George Washington with the bibliography. The nominators are very keen on providing GBook links for books used to source the article. The problem is that this has resulted in discrepancies with the rest of the bibliographical information provided in the article. Specifically, the GBook link relates to one edition of a book, but the ISBN provided in the article bibliography relates to a different edition. This becomes a problem when the two editions are paginated differently; we cannot guarantee that cited pages are the same across differently paginated editions. One of the noms is suggesting that we remove the ISBN references from the bibliography, and rely on GBook links. I can find only a content guideline that covers this, specifically WP:HOWCITE, which says that ISBNs are optional. Factotem (talk) 11:34, 4 December 2018 (UTC)
 * The problem with Googlebook links is that they are not universal. Readers in different countries can't access the same level of information as others. I always add GBook links (and either an ISBN or OCLC number), as an aid to readers, but I know others, such as, eschew the links entirely on more than one ground. If Google decide to paywall the books at some point to rake in even more profits, then those links are of really dubious use, whereas the ISBN is universally accepted and used by pretty much everyone. - SchroCat (talk) 11:44, 4 December 2018 (UTC)
 * I'm aware that GBook previews are not universal, but this is not at issue here. The noms are not linking to the previews, but simply to the book listing, which I believe is both consistent and consistently available across all territories. Factotem (talk) 11:56, 4 December 2018 (UTC)
 * There are other issues with them too, as I've noted. ISBNs are universal and uncontroversial in comparison. Combined with OCLC numbers for pre-1966 books gives two solid systems for use. - SchroCat (talk) 12:05, 4 December 2018 (UTC)
 * For a somewhat similar reason, I tend to use Worldcat links. I used to use GBooks, but an editor complained that I was thereby advantaging a commercial seller. I therefore only tend to use GBooks if the link is to an extract from the book, which one is quoting, rather than just general data. In answer to the specific question, I think you'll really struggle without ISBNs or OCLCs. KJP1 (talk) 11:49, 4 December 2018 (UTC)
 * Use ISBNs; do not link to GBooks. ——  SerialNumber  54129  12:01, 4 December 2018 (UTC)
 * I don't disagree that ISBNs should be the authoritative reference used in the bibliography, but is there any policy, guideline or previously established consensus that would prevent the noms from removing ISBN references and using GBook links instead? Factotem (talk) 12:32, 4 December 2018 (UTC)
 * I think there is a consensus that as WP aims to be an academic project, its best articles should try to reflect the very best practice of academia in reporting the sources used. Would a Google Books link be sufficient for a serious piece of academic work? Not a chance. If I was doing the source review for an FAC and there were no ISBNs/OCLCs, I'd not pass it, but that's just me. - SchroCat (talk) 12:41, 4 December 2018 (UTC)
 * I'm not aware of any policy or guideline that would prohibit that; but if I were doing the source review I would not have accepted GBooks links as a substitute for ISBNs. Frame the question like this: what is the standard and default way to identify a book? That way is what should definitely be there and definitely be correct. Anything else is there as an added convenience for the reader (whether it's an URL to some preview site, a DOI that leads to an online/ebook edition, or some other link or identifier).I can understand the urge to add GBooks links when they actually lead to a preview of the specific page cited (again, as an added convenience, not a substitute); but as identifiers GBooks URLs are entirely unsuited. Google does not, so far as I am aware, provide any stability guarantees for their book URLs, and their identification and differentiation of editions or volumes of a work is exceedingly sketchy at times (just yesterday I struggled with GBooks consistently identifying Vol. 1 of a work as Vol. 2, reusing metadata for the incorrect volume). Google Books is also a commercial service, the preferment of which over, say, Amazon (which provides the same functionality) is hard to defend. ISBNs are vendor-neutral and designed as stable unique identifiers (they are not reused, and I've yet to see duplicate ISBNs for different works, though I'm sure it does happen) and reliably differentiate between both editions and volumes within a work.In short, adding GBooks URLs that lead to page previews may sometimes be acceptable as a convenience for the reader, but they should never be used as a substitute for a proper identifier like the ISBN (or even OCLC numbers for older works even though these are far less reliable). --Xover (talk) 13:14, 4 December 2018 (UTC)
 * I never link to GBooks unless free full text is available. similarly I never link to journal abstracts, only full-text articles. In both cases we are just linking to sales sites, which should, in my view, be unacceptable Jimfbleak - talk to me?  13:27, 4 December 2018 (UTC)
 * ISBNs are not required. All that's required is that citations be enough to be able to locate the source. It's a good practice to make citations such that they are easy to locate, and ISBNs help with that - but they are not required, and editors need to remember that they do not exist for works before the late 1960s, and they did not become common until the 1980s. An ISBN is for a specific edition of a work, and as such, should not be retroactively applied to an edition that did NOT have an ISBN. Nor is it required that all ISBNs be 13 digit - you should cite the edition you use, and if it has a 10 digit ISBN, that is the ISBN that should be used. For works before the creation of ISBNs, use OCLC numbers if you want, but they aren't required either. And one further point - get your ISBNs from the published work - don't depend on Google Books for them because they can be wrong. If you have to get an ISBN from a source other than the work itself, WorldCat has a better record for getting ISBNs correct. Ealdgyth - Talk 14:15, 4 December 2018 (UTC)
 * I go back to a point a made a little earlier: would a Google Books link be sufficient for a serious piece of academic work? Not a chance. Would a piece of academic work which didn't have ISBNs as identifiers in the bibliography be considered complete or acceptable? Not by the standards I am aware of. - SchroCat (talk) 14:29, 4 December 2018 (UTC)
 * Actually, that's not true. Just pulling four history works from my shelves at random - Euan Cameron's The European Reformation (Second Edition, 2012, Oxford University Press) - neither the endnotes nor the extensive further reading sections have ISBNs. Ralph Turner's King John: England's Evil King? (2005, Tempus) does not include ISBNs in the extensive bibliography. Everett U. Crosby's The King's Bishops: The Politics of Patronage in England and Normandy 1066-1216 (2013, Palgrave Macmillan) does not include ISBNs in the (again) extensive bibliography. Adam Jones' Genocide: A Comprehensive Introduction (2006, Routledge) does not include ISBNs either. So, from that sampling, I can safely say that academic works (is anyone seriously going to argue that Oxford UP isn't academic?) do in fact not include ISBNs and are considered academic works. Ealdgyth - Talk 14:48, 4 December 2018 (UTC)
 * Ah, Chicago referencing, I guess. I wouldn't be so absolute as to say "academic works ... do in fact not include ISBNs": "some" don't. I've pulled a series of books from my shelf and 5 out of 6 have ISBNs in the biblio. I can't think of a good reason not to have them in there. - SchroCat (talk) 15:06, 4 December 2018 (UTC)
 * So much of Wikipedia isn't codified in policy or guideline, and in such cases, then we should tailor our actions to what improves the experience for the WP:READER: bottom line. I—personally speaking—can't see the benefit to them by omitting an ISBN, regardless of whether we also include other links. ——  SerialNumber  54129  14:40, 4 December 2018 (UTC)
 * I also don't understand why you wouldn't include ISBNs: as unique identifiers, they're the single best way of finding the specific edition of the book which was used as a reference (noting that pagination, etc, will often vary between editions, at a minimum). I would also oppose any FAC nomination which doesn't include ISBNs, as this means that the article does not meet WP:V. It's also an obvious discourtesy to readers. Nick-D (talk) 10:19, 5 December 2018 (UTC)


 * is right. Bibliographic entries need the minimum information to find the book, i.e, title, author, date, publisher. Date & publisher will differentiate between editions (all of which have unique ISBN identifiers). Beyond that we need page numbers to find the cited material. is correct in this comment. And relying on a g-book version of a source is problematic for lots of reasons and definitely not best practice. If it's necessary to supply an ISBN from a g-book source, then use the ISBN on the copyright page (they almost always show copyright pages) for that edition which is the one digitized. URLs are useless and not best practice. If the book is in hand, and for a subject such as George Washington it should be, if the ISBN is required (they're not), then always use the one on the copyright page of that book. This is a pet peeve I have because more often than not I've seen an ISBN changed during a FAC to one that's different than in the book sitting in my lap that I've been using to write the article. Also, for a subject like George Washington I'd suggest looking on the Library of Congress Catalog  if it's necessary to sort out editions & ISBNs. For other subjects WorldCat does a much better job than g-books. Short answer: no, ISBNs are not required. Victoriaearle (tk) 14:43, 5 December 2018 (UTC) P.S - I did the PR for this article years ago; if I can, I'll take a look at the FAC. Victoriaearle (tk) 14:45, 5 December 2018 (UTC)
 * I think this is definitely a case where I for one will be doing what you—and you!—do, rather than say ;)   ——  SerialNumber  54129  14:34, 6 December 2018 (UTC)
 * At FAC we should aim for the highest standards of presentation, not the minimum we can get away with, and I'm surprised that with regard to ISBNs some experienced FAC contributors are apparently promoting the latter course. In the interests of professionalism, clarity and uniformity we should always be prepared to go beyond the bare minimum; it seems that around ninety-nine percent of nominators accept this. Brianboulton (talk) 11:01, 8 December 2018 (UTC)
 * While I agree entirely, in fairness the question raised was whether there is a policy or guideline that requires ISBNs, which there is not. CITEVAR (love it or hate it) provides for any style that fulfills WP:V and is consistent, and consistently omitting ISBNs fulfills that requirement while GBooks links can meet WP:V (even though they often do not). You can make a good and coherent argument that extraneous details, and especially technical ones like identifiers, are mere clutter and confusion once the essential bibliographic details (title, author, year, publisher, and place) have been provided. I do not, personally, find that argument persuasive, but it is an entirely valid one that many editors appear to adhere to (and which Victoriaearle and Ealdgyth have demonstrated is reflected in at least some relevant external citation style guides). My personal recommendation, based on far too many hours spent trying to track down sources cited with incomplete information, is to include as many identifiers and bibliographic details as you can: there's not a one of them that has not at one point or another been crucial for tracking down the intended source! --Xover (talk) 11:48, 8 December 2018 (UTC)
 * I answered the question that was presented - "Is there any policy that says ISBNs are required at FAC?" - and no, there is no policy of Wikipedia that requires ISBNs. And I think requiring them at FAC would be a bad move because then folks would demand them for works which were not issued with them. Yes, some places (google books, for example) appear to give ISBNs to works which did not have them originally, but this is suspect. ISBNs are issued by various means that differ by country - and they are issued for a specific edition (not a reprinting but an edition). We cannot be sure that what Google gives on its google books pages for a pre-1965 works (1968 in the US) are actually for that edition - it is likely that they are for editions later issued. Nor do I think we should be including ALL identifiers for works - I agree that linking to commercial identifiers (such as ASIN for Amazon) isn't a good idea. Nor do I think ISSNs are required. And I will point out that not all academic publishers use a format that includes ISBNs - I think I've shown that at least for history works, Oxford University Press and others do not include ISBNs in their bibliographic entries. All that said - I include ISBNs when they are included in the works I used, but I do not favor requiring them. I'd rather see the effort go to something more useful, quite honestly. Ealdgyth - Talk 12:58, 8 December 2018 (UTC)
 * I was addressing the issue in regard to the FAC in question, which I took a look at. The first sentence in George Washington is cited to a 1936 and a 1948 book. Trying to ascribe an ISBN to these is problematic - hence this discussion. That the article relies on these sources is equally problematic; my advice would be to rely on more recent scholarship, which almost certainly will have ISBNs. In general there have been cases when I'm working from a book, (some of Hemingway's stories come to mind), that was published pre-ISBN. In such cases, my preference is give the proper bibliographic details, (author, publisher, date), of the book at hand not what I've found on the internet. This is not at all meant as a shirking of excellence but rather done in the interests of precision. Victoriaearle (tk) 16:25, 8 December 2018 (UTC)
 * And that's why they invented OCLC numbers!--Sturmvogel 66 (talk) 16:35, 8 December 2018 (UTC)
 * I confess I'm a little troubled to hear that people would not pass source reviews when ISBNs are lacking. Someone earlier said "Would a piece of academic work which didn't have ISBNs as identifiers in the bibliography be considered complete or acceptable? Not by the standards I am aware of." I'm not sure I've ever seen an academic journal article or monograph include ISBNs in bibliographies, and I've certainly never seen it requested in style guides. Can I ask SchroCat what particular standards he's thinking of? I'm not denying that it happens elsewhere, but the pressure to include ISBNs in citations strikes me as more-or-less peculiar to Wikipedia. Josh Milburn (talk) 12:40, 8 December 2018 (UTC)
 * I was wondering the same thing. For example, see the bibliography section of the Cambridge history department style guide. No mention of ISBNs. Do any publishers require them? SarahSV (talk) 23:47, 8 December 2018 (UTC)
 * I am not aware of any major style guide that requires ISBNs for either references or bibliographies (APA, MLA, Chicago, OUP, CUP), nor even usually mention ISBNs (my biases there are humanities, anglophone, and UK). However, none of these publishers (I'm thinking of OUP and CUP primarily) have fully taken on board the changes in a digital environment. OUP's style guide mentions avoiding using endnotes because they might sell the ebook version of a chapter separately, and provides cautions related to footnote numbering (restarting per page etc.) that do not work well in an online context. They have even deprecated Op. cit. type references that do not work well online. But it is clear their primary focus and body of experience is still in paper publishing. For example, they suggest you do restart footnote numbering each chapter because otherwise it's hard to renumber all the following footnotes when one is inserted or removed, which is only an issue in paper publishing. I do not believe we should put too much weight on these style guides or practices for issues such as this.For our purposes, including ISBNs and other such identifiers (many books and chapters even have their own DOIs now) should be the gold standard. There is no quicker or more convenient way to identify and access a book than through our automatically linked identifiers; and if these are available they are the very best way to identify the correct edition. With the correct ISBN provided we do not even strictly need the name of the publisher and place of publication: the ISBN will resolve the US vs. UK edition automatically (meanwhile, OUP still requires explicitly specifying the edition number for everything except the first edition).My stance is that we should include as many bibliographic details and identifiers as possible for redundancy (including publisher and location even with an ISBN), having spent too much time trying to track down incomplete references, and that FAC is the appropriate place to set that standard (so long as CITEVAR pertains, this is an issue of quality and not style). But, absent appetite to modify the criteria, that's effectively up to those who actually do the source reviews.PS. Oddly enough, book reviews in journals published by OUP and CUP (et al) do usually, in my experience, specify the ISBN for the reviewed work. I have no idea why they consider it more critical there than in an actual reference, but there you have it. --Xover (talk) 09:17, 9 December 2018 (UTC)

FAC reviewing statistics for November
Here are the FAC reviewing statistics for November. Mike Christie (talk - contribs - library) 22:12, 8 December 2018 (UTC)
 * Thank you for all your hard work in maintaining these monthly stats, Mike. Moisejp (talk) 04:51, 12 December 2018 (UTC)
 * , ditto. Thanks for maintaining the stats and for opening the discussions about source reviewing. SarahSV (talk) 05:06, 12 December 2018 (UTC)
 * Thanks, both. The source reviewing RfC appears to have stalled, and will almost certainly be closed as no consensus.  Some more ideas have been raised at the discussion I linked to, but I think I should not be the one to start those conversations, so I'm hoping someone else will pick up the ball. Mike Christie (talk - contribs -  library) 12:03, 12 December 2018 (UTC)

Staten Island Railway
Due to serious issues with original research and sourcing, I've opposed the nomination and request that it be withdrawn. Catrìona (talk) 10:45, 21 December 2018 (UTC)
 * No need to bring it here, the coords will give your oppose the weight it is undoubtedly due :)   ——  SerialNumber  54129  10:53, 21 December 2018 (UTC)
 * Even so, on a more serious note, I am inclined to agree with your concerns; I also suspect a WP:FA?#2 failure, and will comment there as such. ——  SerialNumber  54129  11:05, 21 December 2018 (UTC)
 * I think it's a good idea to highlight reviews like this here. The extensive referencing problems identified early in the nomination mean that the nomination has no chance of succeeding, so the coords are likely to appreciate a note highlighting this. Nick-D (talk) 23:02, 21 December 2018 (UTC)
 * Yes, that's the note they'll look for...on the nomination page. I'm sure it is a good idea; just not one that has caught on in 15 years. ——  SerialNumber  54129  23:38, 21 December 2018 (UTC)

Exception to wait time policy
My previous FA nomination was just archived. I would like to nominate the article Escape of Viktor Pestek and Siegfried Lederer from Auschwitz, which is currently listed for closure at MILHIST ACR. It's a pretty short article and I'm not anticipating a lot of comments. My original goal was to get it at TFA for the seventy-fifth anniversary (5 April 2019), although I'm not sure if that is still possible. Thanks for consideration. Catrìona (talk) 07:10, 29 December 2018 (UTC)
 * I don't want to appear bureaucratic but while, per FAC instructions, we almost always waive the two-week rule when an archived nom has attracted little or no commentary, we usually don't in other cases. and I have just declined a waiver request in very similar circumstances, so while I know two perceived wrongs don't make a right, it would be a bit unfair to grant leave to one and not the other. If the new article is relatively uncomplicated then I wouldn't have thought the two weeks should adversely affect the proposed TFA, and it would allow time for any applicable lessons learnt from the previous FAC nom to be incorporated into the new one. Cheers, Ian Rose (talk) 03:42, 30 December 2018 (UTC)