Wikipedia talk:Talk page guidelines

Request for comment: Do the guidelines in WP:TPO also apply to archived talk pages?

WP:TPO details several instances of comments that are appropriate to remove from talk pages, such as vandalism, spam, gibberish, and test edits. Does this apply to archived talk pages as well? I will post a more detailed statement and further context in the replies. Gnomingstuff (talk) 17:30, 16 October 2024 (UTC)

TO BE CLEAR/PLEASE READ: This is a yes or no question. For those who are having difficulty interpreting the yes or no questions:

• "Support" means "Yes, all of the appropriate edits listed in TPO are also appropriate edits to archived talk pages."
• "Partial support/oppose" means "Some, but not all, of the appropriate edits listed in TPO are appropriate to archived talk pages." If you !vote this, please specify which edits.
• "Oppose" means "No, none of the edits listed in TPO are approprate edits to archived talk pages."

This is going to be long, so apologies in advance. For context/rationale, see this protracted discussion on my talk page.

There is a large amount of vandalism to Wikipedia -- much more than one might think -- that has gone undetected for years, often since the early days of the project. I use the phrase "vandalism" here to encompass any unconstructive edit that would be reverted on sight, across the spectrum from oversightable edits to gibberish. I do not use it to encompass comments that are merely uncivil or waver off topic. Essentially, I'm using a slightly narrower version of the definition and precedent from WP:TPO.

My investment in this topic is that reverting undetected vandalism is most of what I do on Wikipedia. My priority was originally to remove this stuff from main article space, but I am no longer finding much low-hanging fruit there, so I am now working on talk page vandalism. I consider this a priority; these comments are not only readable on site but indexed by Google -- which is how I found the stuff in the first place. In addition, they are intended to serve as a readable record of what people actually said. Changing what people actually said, drive-by deleting constructive comments, and cluttering the discussion with nonsense all make it difficult for talk pages to serve their intended purpose and bloat the page for no good reason. As such, WP:TPO is pretty clear that this sort of thing can be removed.

When vandalism stands for 10+ years on busy talk pages, it frequently makes its way to page archives. Page archives have a banner stating "do not edit this page." However, I kept finding hundreds of instances of the stuff in my searching, and it felt wrong to just see them but do nothing. So, in March 2023, I asked a question on the help page for archiving talk pages whether the banner applied to removing undetected vandalism. At the time, I was asking about the most blatant cases of vandalism, since I expected the answer to be "only in rare cases of X," but the response I got from two people (one admin) was much broader: that the banner "doesn't apply at all" to "maintenance edits such as removing vandalism."

So, I went about removing such content for more than a year, generally in bursts, and received no negative feedback and some positive feedback. As before, I started with low-hanging fruit then moved on to the sort of disruptive edit mentioned in WP:TPO. To be clear, I do not intend to revert any edits not encompassed in those guidelines (if anything I think they are too liberal in what can be removed); there is no infinite slippery slope. The thing is just that there is so much undetected vandalism; thousands of instances reverted, probably thousands to come.

That being said, two people have complained about this in recent months, hence the RfC. The arguments against removing vandalism on archived talk pages, according to the complaints, seem to include:

- Reverting undetected vandalism on talk pages is not an improvement to the encyclopedia. I personally cannot think of a single place on the project where this is true, and WP:TPO seems to state that it's appropriate.

- There is no urgency to removing vandalism that has gone undetected for years. I disagree. There is no deadline, etc., but I think removing vandalism of any kind is more urgent than many other tasks on the project.

- People have to check whether the edits are legitimate. I don't even know what to say to this one; these kind of edits, I would think, should speak for themselves. People frequently use rollback to remove similar content on talk pages, which is reserved per WP:ROLLBACK for edits where "the reason for reverting is absolutely clear."

- Removing vandalism makes it more likely for other vandalism to fall through the cracks because it adds entries to watched pages. I find this argument, frankly, ridiculous. It can be applied to literally any of the millions of edits made to pages that might show up in a watchlist; should we stop doing those too? Given the breadth of subject matter of the vandalized pages, I also find it hard to believe that any one person would be watching enough of them for this kind of edit to make much of a difference.

- I make a lot of edits. This is true, and I have tried to take WP:MEATBOT into account. (I do realize that I tend to get locked in on tasks that require going through long lists.) I don't use any bots or tools more advanced than wildcard search, however. (i.e., no regex, per the searching guidelines; I tried regex a handful of times and found it not very useful for this). This is less a policy complaint than a personal complaint, but I am mentioning it for completeness' sake.

- More people might start editing vandalism on talk pages, exacerbating any of the above. That sounds great to me! More people should be doing counter-vandalism (to the extent that anyone "should" be doing anything here).

I am happy to address comments and discussion by other editors. Gnomingstuff (talk) 18:06, 16 October 2024 (UTC)

Addendum to the last point: I've seen "more people might start reverting vandalism on archived pages" come up repeatedly during this discussion, and well, the best argument against that is that no one did much of it for over 15 years, so it's hard to imagine many people starting now.

There's also a finite amount of undetected vandalism on current archive pages (even if it keeps revealing itself as more than anyone thought), so 15 people doing it is no different than 1 person doing it, it'll just get done faster. Gnomingstuff (talk) 07:12, 17 October 2024 (UTC)

Further addendum: In the past few days, I have identified 150 instances and counting of undetected bad edits (vandalism, nonsense, test edits, etc.) that now persist indefinitely in archived pages. Each contains the offending diff(s), the majority of which originated earlier than 2010. I am comfortable saying that none of these 150 instances are legitimate or constructive parts of the discussion, and many of them are especially egregious. There is no way of knowing how much is out there, but if I have already found 150 cases, that does not bode well. I have not made any changes to the archived pages themselves. Gnomingstuff (talk) 20:40, 15 November 2024 (UTC)

• Update 11/20: Over 500
• Update 11/29: Over 1000, 134 really bad

Survey re TPO Guidelines

• Oppose. And see AN discussion Striking comments from banned sockpuppets and modifying archived comments. Perhaps participants there should be informed that for some reason this RfC about the WP:TPO guideline appeared. Peter Gulutzan (talk) 18:08, 16 October 2024 (UTC)

  Thanks for the link, I was unaware of it. Gnomingstuff (talk) 20:41, 16 October 2024 (UTC)

• Oppose. Reverting archived vandalism wastes editor time (of the person searching for it, of the person editing the page to remove it, from watchers of the page, and from those looking through contributions) and draws attention to things that are best just ignored. The alleged benefits are at best trivial and in many cases incorrect. Thryduulf (talk) 18:42, 16 October 2024 (UTC)
  • For clarity, I opposed changes to the status quo which clearly does not apply to archived talk pages. Archived talk pages should be edited only when there is some active harm being caused, which is almost never the case. Thryduulf (talk) 19:03, 16 October 2024 (UTC)
• To the question Does this apply to archived talk pages as well?, no. Do not edit archives. (And seriously, what value would that work contribute? Surely there are more constructive edits to be made.) (edited to add) Tryptofish's comment made me think of an exception: removing vandalism/disruptive edits that were made after the content was archived. Schazjmd (talk) 19:01, 16 October 2024 (UTC)
• Generally oppose the editing of archived Talk pages, with possible exceptions for libel and copy violations. Mathglot (talk) 19:05, 16 October 2024 (UTC)
• General oppose I could see in exceptional circumstances instances where this was appropriate (as mentioned by Mathglot), but in general this seems like a bad practice. -- LCU ActivelyDisinterested «@» °∆t° 19:12, 16 October 2024 (UTC)
• My inclination is to oppose the editing of archived talk pages. The benefit to the encyclopedia is minuscule in these cases, and I think the risk of confusion or annoyance to other editors outweighs that benefit. There are a handful of exceptions to this general case—for instance, I believe that material that merits revdel or oversighting should be removed, even if it's on an archived page. However, non-constructive yet comparatively innocuous comments (such as test edits or gibberish) are probably not worth the effort to revert. ModernDayTrilobite (talk • contribs) 19:13, 16 October 2024 (UTC)

  In the "taxonomy of non-constructive edits" section below, I gave some more detailed opinions on which types of non-constructive comment should or shouldn't be removed. In addition, Rhododendrites raised a good point below that edits to archives can also include fixing syntax errors, which in my opinion has definite value. I still don't think comments such as pure gibberish are necessarily useful to remove, but there are enough categories of material that merit removal that I no longer find it appropriate to consider my vote an oppose per se; instead, I'd say my position is support if under specific circumstances. ModernDayTrilobite (talk • contribs) 15:27, 17 October 2024 (UTC)

• Partial support, at least. Following Mathglot's link, I checked four links (example), all of which were what I'd call "graffiti". I see no reason to oppose edits like my example; they're worthless, and the text gets in the way for anyone who's consulting the archive. This is a constructive edit, so the guideline shouldn't restrict such edits, and if Gnomingstuff wants to do it, we shouldn't say "do something more constructive". I say "partial" because I haven't yet noticed any edits other than anti-graffiti. Nyttend (talk) 20:17, 16 October 2024 (UTC)

  Can I strike this comment to make things easier on the closer? – Closed Limelike Curves (talk) 19:56, 9 November 2024 (UTC)

  You can strike your own comment of 19:56, 9 November, however per WP:TPO do not strike anyone else's. Also, please read WP:THREAD to learn how and where to add reply comments; you are replying to an October comment that has already has responses, and this is not the right place for your reply. Please do not attempt to fix the placement now, as it will likely just make things even worse, just spend some time at WP:THREAD. Thanks, Mathglot (talk) 20:12, 9 November 2024 (UTC)

  To clarify, this question was directed at @Nyttend specifically (I was asking him for permission, not asking if it's allowed—I know not to strike a comment without asking first). – Closed Limelike Curves (talk) 21:10, 9 November 2024 (UTC)

• Never mind, full support, now that I've found edits like [1] and [2] and [3]. The encyclopedia definitely benefits from the removal of outright vandalism like this, so the guidelines shouldn't stand in the way. It's a tiny benefit, but if Gnomingstuff wants to do it, "are probably not worth the effort to revert" is irrelevant; we're not talking about a bot that's using limited resources. Nyttend (talk) 20:20, 16 October 2024 (UTC)

  This is a representative sample, I think; from what I have found it's skewed toward the gibberish side of the spectrum (there is a huge spike starting 2022, probably from ChatGPT), but it also extends far enough to the other end that I've emailed oversight multiple times. Gnomingstuff (talk) 20:39, 16 October 2024 (UTC)

  I clicked link 2 just to see...why would anyone oppose removing outright nonsense like that which made its way into a Talk page before it was archived? -αβοοδ (talk) 04:46, 19 October 2024 (UTC)

• Mostly support. I've looked at the links to previous discussions and past edits, posted here by other editors, before coming to this conclusion. I'm not very sympathetic to the argument that it wastes editor time when it shows up on watchlists, because you don't have to watchlist archives unless you want to see if archives change. And I'm not very sympathetic to the argument that there's a template at the top of archive pages, saying not to edit them, because the intention of that template is to indicate that the discussions are closed, not that the page should be treated as if full-protected. Now the reason that I say "mostly" is that it seems to me that the real goal here should be that editors who might later look back at an archived discussion should be able to see, without being misled, what the discussion was, at the time that it took place. For that reason, if a sockpuppet commented at the time, but the sock comment was not struck at the time, then the comment should be left as is, because that's what the discussion consisted of at the time. But a lot of the vandalism being discussed here has the effect of altering the discussion, as it took place at the time. And that's appropriate to revert. If some vandal comes along and ignores the template saying not to alter the archive page, and vandalizes it, it's silly to scold the editor who undoes that vandalism. Let's say that, long ago, I took part in a discussion and said whatever I said then, and it's long since been archived. Now a vandal comes along and changes what I wrote to something stupid. What's the purpose of preserving the vandalism? What is it being preserved for? All it accomplishes is making my long-ago comment sound stupid, in a way that misleads editors who come along later to find out what happened in that discussion. --Tryptofish (talk) 20:54, 16 October 2024 (UTC)
• Comment. Nyttend's diffs make a good case that "vandalism that modified another editor's comments" should be reverted in archives, if current policy discourages that it should be changed. I think that the threshold for "cleaning" archives should be higher than "would revert on an active talk page", reversions like Special:Diff/1251394873 feel unnecessary. Walsh90210 (talk) 23:16, 16 October 2024 (UTC)
• I agree with Tryptofish that the goal of an archive is to preserve the past conversations. This doesn't mean every literal wikitext source character has to be kept the same, or that the output has to be a pixel-perfect match with the past. (Fixing unclosed elements is a common edit that is done to preserve the original appearance of the discussion, after the MediaWiki software started rendering the output HTML differently.) But it should be possible to look at the archive and experience the discussion as it occurred at that time. So if a banned editor made comments without being detected at the time, their comments shouldn't be deleted from the archive, as that wouldn't reflect what the participants read and responded to. If someone vandalizes an archive, the change should be reverted, in order to restore the discussion to its original archived state.
• (On a side note, template transclusions are a problem with this goal, since they always transclude the current version. Anyone concerned about this should subst: the template, or find a way not to use it.) isaacl (talk) 23:25, 16 October 2024 (UTC)

  You raise a good point, and I could see an interesting proposal coming out of it to the effect that archive bots could have a subtask that substed templates at archive time. Ping me, if you get involved with a proposal like that. Mathglot (talk) 02:53, 17 October 2024 (UTC)

• Support, as one of the commenters on the original discussion on the archiving help page (side note: I don't think a user's adminship or the lack thereof has any bearing on the worth of the comments there). As I said at the above-linked discussion, I sometimes make such vandalism removals myself, such as this edit to Talk:Chewbacca/Archive 1. I'm interested in such vandalism removals from the angle of preserving the first good-faith comments made on a talk page, like this edit to Talk:Dylan Thomas/Archive 1. Graham87 (talk) 02:57, 17 October 2024 (UTC)
• Oppose You don't need permission to remove BLP stupidity such as that shown in Nyttend's diffs, see WP:IAR. However, disturbing an archive just to remove fluff (diff), even if it met WP:VAND, is a bad idea because it makes examining archives much harder because now you have to also examine history to see if the record has been altered. Also, gnoming archives sets a bad precedent which would encourage enthusiasts to make other "fixes". Removing junk before it is archived would be great (I do that). Johnuniq (talk) 06:10, 17 October 2024 (UTC)

  A bit confused on the "oppose" here - "BLP stupidity" is part of the WP:TPO guidelines, so it seems like you're saying that some of those guidelines but not all should apply to archived pages, e.g. a partial oppose/support? Gnomingstuff (talk) 07:00, 17 October 2024 (UTC)

  This is a badly worded RfC. The issue is clearly whether someone should "fix" archives. The answer to that question is no. I believe archives should be a record of what occurred on a talk page and should not have adjustments made unless for compelling reasons (such as linter errors, BLP violations, serious copyvios). Johnuniq (talk) 07:58, 17 October 2024 (UTC)

• Support. Countervandalist editors are welcome and should be encouraged. I can't understand the opposers at all. Let our volunteers do what interests them, please.—S Marshall T/C 08:12, 17 October 2024 (UTC)
• Partial oppose - I'm against most edits to archives for the simple fact that none of the original participants will see the changes. That means there are only three good reasons to do so that I can think of: to fix syntax errors, to remove egregious attacks/vandalism/BLP issues, and to update a link to a separately archived thread for posterity. — Rhododendrites ^talk \\ 11:42, 17 October 2024 (UTC)
• I don't really see a need to do so except in rare cases where IAR could be applied, but I also don't see the point in prohibiting it. So supportish I guess? Alpha3031 (t • c) 13:18, 17 October 2024 (UTC)
• I support the removal of actually offensive content, anything that could be illegal or eligible for revdel/oversight (BLP/copyright vios), and changes made after a discussion was closed. I’m also pretty sure that those things are allowed under current policy. I think removing "spam" from archives is a waste of time, and I am opposed to sockstrikes in archives, as they likely influenced the outcome of the disucssion. Toadspike [Talk] 16:31, 17 October 2024 (UTC)

  It seems Gnomingstuff holds a similar opinion and expressed it in more detail in the section “Discussion on a taxonomy of nonconstructive edits” below. Toadspike [Talk] 16:33, 17 October 2024 (UTC)

• Support in at least some cases. False accusation personal attacks and other forms of bullying should be aggressively and systematically deleted from all pages on sight, including archives. We should treat such comments and behaviour as we would treat WP:BLP violations in the mainspace (if only because at least some false accusations against other editors, including pseudonymous editors, are equivalent to WP:BLP violations). We have had a serious problem with such behaviour in the past, and with the failure to stop such behaviour and delete such comments, and we have large chunks of archives (WP:ANI comes to mind) that need to be blasted out of existence to avoid perpetuating smears and bullying. James500 (talk) 02:42, 18 October 2024 (UTC)
• Oppose per Johnuniq and Alpha3031's responses. Cheers. DN (talk) 03:30, 18 October 2024 (UTC)
• Support after looking at several of the examples of vandalism cited here. There is no reason for nonsense like that to persist, even on archived talk pages, and edits to remove them essentially restore the record. -αβοοδ (talk) 04:51, 19 October 2024 (UTC)
• Support I find it extremely difficult to understand why someone would want to spend their time clearing vandalism, spam, gibberish, and test edits from talk page archives, though this RfC comes from a fitting username. However, it seems reasonable to interpret WP:TPO as superseding the "do not edit this page" archive banners in such cases. ViridianPenguin 🐧 ( 💬 ) 20:38, 19 October 2024 (UTC)
• support per my comments up there and down there on this page etc jp×g🗯️ 21:09, 19 October 2024 (UTC)
• Partially oppose, unless it is blatantly libel or copyvio as per Mathglot.--Takipoint123 (talk) 21:27, 19 October 2024 (UTC)
• Suppose Anything that meets WP:CRD {1,2,4,7} should be fair game (in agreement with Toadspike above). Philosophically, I'm in the same camp as Johnuniq and Tryptofish above, in that archives should be accurate records, even if those records contain worthless garbage. Post-archival vandalism should be removed for that same reason: inappropriate modification of the wikihistorical record.
  My main concern would be breaking search ordering by edit date, with rvv edits bumping decade-old archive pages to newer revision dates, but User:MalnadachBot already permanently ruined this everywhere two years ago. Folly Mox (talk) 14:43, 20 October 2024 (UTC)

  Sorting archive pages by last edit date was always a grotesque hack that nothing should have been built on, for exactly the reason that it was a house of cards that would be permanently ruined by the most perfunctory vandalism (or even by a bot fixing lint errors). jp×g🗯️ 05:21, 23 October 2024 (UTC)

  @JPxG: Where does this sorting occur? --Redrose64 🌹 (talk) 21:20, 23 October 2024 (UTC)

  @Redrose64: It's an option on the search page. Graham87 (talk) 11:07, 25 October 2024 (UTC)

  If you mean Special:Search, I don't find that option, although I do observe that each result entry ends with a timestamp of the last edit to that page. --Redrose64 🌹 (talk) 19:10, 25 October 2024 (UTC)

  If you click "advanced search" there is a "sorting order" section at the very bottom. That gives three options: "Relevance" (the default), "Edit date - current on top" and "Creation date - current on top". Thryduulf (talk) 19:13, 25 October 2024 (UTC)

  No, not there either. Perhaps it's a Vector thing, or Mobile. I use MonoBook on desktop. --Redrose64 🌹 (talk) 21:19, 25 October 2024 (UTC)

  

  The advanced search options dialog on Monobook

  I also use Monobook on desktop, but it's in the same place in vector, timeless and on mobile. Thryduulf (talk) 23:17, 25 October 2024 (UTC)

  That is totally different from what I see. Is it a script or gadget that you have enabled? --Redrose64 🌹 (talk) 08:19, 26 October 2024 (UTC)

  It's only visible if scripting is enabled. Johnuniq (talk) 08:28, 26 October 2024 (UTC)

  Just to be sure, is "Suppose" a typo for "Support"? – Closed Limelike Curves (talk) 18:25, 9 November 2024 (UTC)

  Portmanteau of support and oppose. Mathglot (talk) 18:39, 9 November 2024 (UTC)

  Thanks! – Closed Limelike Curves (talk) 20:03, 9 November 2024 (UTC)

• mostly oppose Let the archive be archives, a historical record. Past. Done. Closed. If we find an error in a Mozart's own score, should we change it? No. If he wrote some nasty lyrics about someone (which he likely did... :-) ) should we remove them? Also no. Do not change the past. Some few edits should be OK, but considering they are edits to protect and improve the archive. Obviously reverting any undue changes to the archive is OK. Maybe subst'ing templates to the version as they were when used...? Maybe adding links (at the top or bottom) to related pages for context and indexing. The only content change I see as fit, would be content still harmful for living people, and even those changes should be somehow tagged. Note, that this is not at all about how {u|Gnomingstuff} uses their free time. This is about how do we want to keep our archives. Just as much as we have nothing to say to someone vandalizing pages on how to use their time, but we say we do not want those edits. - Nabla (talk) 15:58, 20 October 2024 (UTC)
  • ping Gnomingstuff, as I misspelt the template above - Nabla (talk) 16:01, 20 October 2024 (UTC)

  I don't think this is a good analogy. It would be more as if Mozart wrote a score, some guy broke into his house, ripped out one page and added a bunch of random notes, that fact was not discovered until centuries later, and musicologists cried "we can't change it, it's Mozart!" At the very least this would produce some kind of authorship controversy and the option of restoring Mozart's un-vandalized script would at least be on the table. Gnomingstuff (talk) 16:34, 20 October 2024 (UTC)

  Good point. Note that I am OK with improvements to the archive. The liost under discussing in another section might be a good start to indentify what IS improvements. Would restoring vandalism be an improvement?... I really am not sure, but in doubt, I'll stick to: don't change history. Otherwise we will not know what is historical. Note that my analogy used a work of art, which is indeed not a good analogy either way. We are keeping historical records, I think it is way too tricky to go changing records. - Nabla (talk) 19:01, 20 October 2024 (UTC)

  I feel like anyone who finds themself saying something like "Would reverting vandalism be an improvement? I am really not sure" (I assume "restoring" wasn't what you meant) should stop and ask themselves what we are even collectively doing here. They should also refresh themselves on Wikipedia policy, because WP:VANDALISM is policy, and the very first thing it says under the header "How to respond to vandalism" is "Upon discovering vandalism, revert such edits." It goes on to reiterate this: "If you see vandalism on a list of changes (such as your watchlist), then revert it immediately." "Repair all vandalism you can identify."

  Like, this isn't some obscure hidden policy that no one pays attention to anymore. It's common sense. Or at least I thought it was. Gnomingstuff (talk) 22:53, 21 October 2024 (UTC)

• Oppose once material hits the archives. It becomes part of the WP historical record at that point, warts and all, and there is minimal value in spending time on this. Per Mathglot's observation, while we might trust the judgment of Gnomingstuff in altering archived materials, once we say it's fine to tinker with this stuff, we open the door to other parties, too. The potential mess, in my mind, outweighs any benefits from spending time on this, and we shouldn't allow after-the-fact changes to what are closed discussions. Grandpallama (talk) 18:56, 21 October 2024 (UTC)

  Okay, but -- "once it hits the archives" -- you're saying that if a thread is archived in 2025, vandalized in 2027, and noticed in 2029, the men of the bold future can't revert it to the 2025 version? jp×g🗯️ 19:39, 21 October 2024 (UTC)

  I think it's more that I'm saying it's not worth the hassle of making a new set of regulations about when we can/can't play around in the archives. Besides, all the men of the bold future will probably be underwater by 2029. :) Grandpallama (talk) 20:23, 21 October 2024 (UTC)

  No! What we archivists are saying is that (almost) any edits to archives should be reverted. The exceptions have been discussed above (linter, BLP, copyvio). Johnuniq (talk) 21:17, 21 October 2024 (UTC)

• Partial/mostly support - The categories in TPO should tighten when a page is archived. In particular, edits to other people's comments are helpful when there's a specific reason (Copyright, BLP/libel, personal attacks), or you're restoring to the status that it was archived in, restoring comments to how they were before they were vandalized, or when you're improving the functionality of the archive (something got munged while/after being archived). Something like adding a signature to an unsigned comment is probably a net neutral, with the benefits of inline attribution and the drawbacks of archive fidelity roughly cancelling. Tazerdadog (talk) 03:45, 23 October 2024 (UTC)
• Partial support/oppose, archive editing should be limited, but there are clear cases where it should be done. As others have stated above, archives serve a purpose, and that is to be useful records. Restoration. Vandalism to the comments of others is an obvious clear case of edits that should be fixed in archives. If they are left, then the archive does not actually archive the original conversation. Furthermore, as archives lack the page history, it is a bit more difficult to check how a conversation developed. This would also apply to mistaken removals of others' comments, which I have seen a few times. Removing prohibited material. There is going to be an IAR case for removing some prohibited material, it will probably need a stronger case than a live talkpage, but a case can be made. Removing harmful posts/Off-topic posts. Generally oppose editing archives for these purposes, much harm comes from interfering with live discussions and attacking current editors, which are less of an issue in archives. Moving edits to closed discussions. Leaning oppose to this one, little benefit and the timestamps should help verify things in the worst case. Attributing unsigned comments. Support, as this helps archives serve their purpose of being archives. Signature cleanup. Lean oppose, many potential issues with signature might be fixed by adding better attribution afterwards (effectively Attributing unsigned comments) rather than modifying the signature. Non-free images. Probably should be replaced with a link, likely a rare occurrence. Fixing format errors/Fixing layout errors/Sectioning/IDs/Section headings/Removing duplicate sections/Fixing links/Hiding or resizing images/Deactivating templates, categories, and interlanguage links/Hiding old code samples/Review pages/Removing or striking through comments made by blocked sock puppets/Empty edit requests. Oppose, let sleeping formatting lie. I can imagine some exceptions to the opposes might be made for fixing stuff up as it goes into the archive (ie. quite recently archived items), but not enough to specifically call them out from the general IAR principle. I would also add that, again in general terms, similar principles should apply to old talkpage comments that haven't been technically archived. CMD (talk) 06:19, 23 October 2024 (UTC)
• Partial: It's entirely normal to remove vandalism, copyvios, test edits, outing, etc. It's also normal for someone to, say, fix their own typo, and for third parties to do things like fix links that have become broken (linked discussion has itself become archived, two templates have swapped names and what is rendering in the archive page is not what the original poster intended, etc.), and other minor maintenance on archive pages. It is not at all normal, and would be undesirable, for material that is itself subject to discussion to be suppressed after the fact except in unusual circumstances. Strike it if you mean to belatedly retract a personal attack, for example, but do not remove it entirely if it became part of the discussion. In short, do not do violence to our consensus record.  — SMcCandlish ☏ ¢ 😼  16:48, 24 October 2024 (UTC)
• Oppose The value of an archive is its finality. By allowing greater edits to an archive, we invite relitigation. CaptainEek ^{Edits Ho Cap'n!}⚓ 06:02, 26 October 2024 (UTC)

  @CaptainEek: So if someone vandalizes an archive, in violation of what you just said, are you saying that it would be disruptive to revert that? --Tryptofish (talk) 18:41, 26 October 2024 (UTC)

  @Tryptofish What?? That's not at all what I said. I meant that the status quo should remain. And right now, the status quo is that if someone vandalizes an archive, it should be undone. As @Johnuniq said above What we archivists are saying is that (almost) any edits to archives should be reverted.. CaptainEek ^{Edits Ho Cap'n!}⚓ 18:51, 26 October 2024 (UTC)

  Ha! There's confusion about what exactly, this RfC is asking. I suspected that you actually meant what you said in your reply to me, but a lot of editors have been framing that view as "support", rather than "oppose", because the question, as asked, was more about the kinds of gnoming edits that include vandalism reversion, which some editors have actually objected to. --Tryptofish (talk) 18:58, 26 October 2024 (UTC)

  @Tryptofish haha yeah, the framing made me confused if I was supposed to say yes or no lol CaptainEek ^{Edits Ho Cap'n!}⚓ 19:06, 26 October 2024 (UTC)

• Support. If we would remove this material before the talk page is archived, we should remove it afterwards too. The act of achiving is not particularly "holy" and is often done automatically without oversight, so there has been no checking at that stage that the page really is in the state that we want to preserve forever. JMCHutchinson (talk) 11:07, 28 October 2024 (UTC)
• Partial support This may not be an exhaustive list but I hope it is enough to make my thoughts mostly clear. Things that can be edited on archive pages:
  Edits made after the archive was created that are improper edits; clear gross vandalism/abuse/BLP violations/legal issues (with revdel/oversight, if needed).
  Things that should not be edited: striking socks (to preserve the discussion as it occurred); empty/incomplete/nonsensical posts/sections (there's no real advantage to removing them in archives but don't care too much); the majority of fixes for clarity, links, linting, etc (exceptions would be fixing things that are causing larger rendering issues. the vast majority of linting "errors" that are fixed don't matter now and won't matter in 10 years).
  I'd also advise being more careful about referring to edits as vandalism. Somewhat adding, for example, a small amount of stray text on a talk page is rarely vandalism. It's normally a genuine mistake or a literal test to see if they can edit something. Skynxnex (talk) 18:34, 1 November 2024 (UTC)
• Partial support BLP violations and copyvios should still be removed even after archival. Likewise any syntax errors or other problems which do or might in the future make it more difficult to read the comments, as I'm sure we've done a lot in the past with linter fixes and the like especially when technical changes have meant stuff which you to be fine now breaks. I'd also support reverting any vandal modification of someone else's comments whether done before or after archival. Reasonable modification of comments by someone else e.g. RPAs or other reasonable redactions should not be reverted although it might be okay to make it clear who did this if it wasn't made clear. I'd oppose modifying or striking comments simply for being nonsense, dumb or even offensive although it might be okay to sign these comments if they are unsigned and might mislead into thinking they were written by someone else. I'm fine with an exception for anything which qualifies for revdeletion. Nil Einne (talk) 10:43, 3 November 2024 (UTC)
• Partial oppose. Except in extreme cases (like severe libel, doxing, etc.), just leave archives alone. It's wasted effort. They don't need to be brought to publishable standards. If there's cruft or vandalism that got archived, the damage is already done because the work to fix it brings about issues worse than the cure (eg complicating page history). Now if somebody is vandalizing already-achieved pages, then handle it. But I'm talking pre-archive here. Instead of spending time doing this, go improve references or something that improves the quality of the read-facing content. Jason Quinn (talk) 11:56, 3 November 2024 (UTC)
• Yes: I have previously come across several WP:NPA, WP:NOTWEBHOST violations and also deliberate changes to entire paragraphs changing the entire meaning of it. Needless to say, I reverted them. These content really do not belong to Wikipedia and must be removed (revdel-ed if too egregious). —CX Zoom[he/him] ^{(let's talk • {C•X})} 21:22, 3 November 2024 (UTC)

  Also, archives showing up in watchlist is a non-reason. The entire point of watching a page is to keep track of changes. If you don't expect any changes to it, no need to watch it at all. Let the person fighting vandals do it. —CX Zoom[he/him] ^{(let's talk • {C•X})} 21:22, 3 November 2024 (UTC)

  My goal in watching archive pages is to keep track of changes, and I do not expect to see changes to them. So far, we are in agreement. Where we do disagree, crucially, is in your next statement:

  If you don't expect any changes to it, no need to watch it at all.

  Au contraire ! When I do see a change to an archive page, I go check it to make sure there isn't someone running around trying to stealthily vandalize pages in an area where they expect few people to be watching (i.e. archives). Ceasing to watch them would give free rein to vandals, and be completely counterproductive. You just recommended (in GF) not watching them, which however removes the guardrails, and helps them continue. I prefer the opposite tack. Mathglot (talk) 18:39, 4 November 2024 (UTC)

  I think I failed to present my point properly. Yes, if you want to stop vandals, watching archives is fine (I do too). But if someone doesn't deal with vandals but gets angry when they see a reverts to vandals because their watchlist gets long, then they should not keep the archives in watchlist. —CX Zoom[he/him] ^{(let's talk • {C•X})} 11:23, 5 November 2024 (UTC)

• Mostly support per Tryptofish — Martin (MSGJ · talk) 12:49, 4 November 2024 (UTC)
• Oppose per WP:TPO which states A discussion which has been closed with the {{subst:Archive}} or similar template is intended to be preserved as-is and should not be edited. Andrew🐉(talk) 21:57, 6 November 2024 (UTC)

  Andrew, this discussion is about Talk page archives, not about conversations that have been closed (i.e, having a border and colored background) using the confusingly named {{Archive top}} template. Mathglot (talk) 02:28, 7 November 2024 (UTC)

  Talk page archives implicitly close discussions with a similar template such as {{talkarchivenav}} which states emphatically,"Do not edit the contents of this page". Andrew🐉(talk) 09:08, 7 November 2024 (UTC)

• Strong oppose, with the exceptions of removing potentially oversight material, and updating your own username if your account had been renamed since the comment and no one else in the discussion mentions your statement by your own username or obvious reference to it. Animal lover |666| 19:51, 7 November 2024 (UTC)
• Mostly support. The ref tag fix discussed below is an excellent illustration of why archive gnoming should be allowed in some cases. SarekOfVulcan (talk) 21:27, 9 November 2024 (UTC)
• Support its not all that necessary to edit archived discussions but when people think there is a good reason to do so there is no reason why this guideline can't be used for archived discussions. Crouch, Swale (talk) 18:50, 12 November 2024 (UTC)

Discussion re TPO Guidelines

Thank you to Gnomingstuff for starting this Rfc.

The use case that actually provoked this Rfc were some edits to archived Talk pages that were archived many years ago. The prior discussion is here. My concern is, that heretofore, I very, very rarely saw archived Talk pages hit my Watchlist, and now I see them sometimes. I have these issues:

• some of these repair edits occur many years after the page was archived. I do not see how this improves the encyclopedia in any way.
• in the beginning, I didn't know what these edits were, and went to go investigate to make sure they were not some subtle (or not so subtle) form of vandalism. Having examined them, I now trust Gnomingstuff to do the right thing, and no longer need to investigate them, if I see their sig on an archived page. However, if a few more editors follow suit, I will have to start investigating again, until I am persuaded I do not need to; this will lower my productivity on actual encyclopedic pages.
• These edits appear on my watchlist, which is long, and that reduces the number of useful article pages in my Watchlist, which then get bumped off the bottom. Each page taken up by one of these archival repair edits, is a page that runs off the bottom of my list, which I am then not aware of.
• The banner at the top of archived pages say, Please do not edit the page.
• Who benefits? I understand that Gnomingstuff directly benefits; I have mentored users for whom some types of gnoming edits can be a very rewarding and pleasurable experience, and I don't wish to deprive them of that. However, I think the needs of the encyclopedia must be paramount and take precedence.

Although by no means intentional, these edits feel WP:DISRUPTIVE to me in a very tiny way, but I am very afraid that if taken up generally by more editors, it could become genuinely disruptive in a significant way, to a lot of experienced editors, especially to those with long watchlists they attempt to monitor. Please do not encourage edits to archival pages, except in individual cases approved by policy (libel, copyright, maybe some others we could discuss). The rest of them are simply not helpful, and have the potential for causing harm, or at least, lost productivity. Thanks, Mathglot (talk) 18:42, 16 October 2024 (UTC)

I'm trying to imagine a situation in which you want to make such an edit. Are you talking graffiti on a talk page, improper replacement of content on a talk page (i.e. I say something, and then later someone else edits my comment without any good reason), improper deletion of content from a talk page, or what? It would help to have a few examples of edits you've made in this area. Nyttend (talk) 19:27, 16 October 2024 (UTC)

Nyttend, cannot answer the 'why' part, but here are 89 examples (out of 500) you can peruse on this page; highlight them by search-on-page (Ctrl+F) for 'archive'. Mathglot (talk) 19:55, 16 October 2024 (UTC)

The "why" is pretty simple - I think that vandalism is bad, that undetected vandalism is worse, and that reverting it is a better use of my spare time browsing the internet or watching reality shows or whatever. I guess the thing that bothers me the most about this whole argument is people deciding for me what is a productive use of my own time. Gnomingstuff (talk) 20:50, 16 October 2024 (UTC)

Can we be sure that such changes do not break links to archived pages? Alaexis_¿question? 20:50, 16 October 2024 (UTC)

Good question, I don't really have an answer to it. I know that it's possible for vandalism to leave stuff broken -- it frequently messes with subject headers, wikilinks, etc. The only thing I can think of on the other end is restoring ancient markup, which is usually fixed by bots, but more technically inclined people might be able to think of more. Gnomingstuff (talk) 20:54, 16 October 2024 (UTC)

Of the archived talk page changes I have seen, I have not noticed any of them breaking links. Mathglot (talk) 21:15, 16 October 2024 (UTC)

There are people above who are !voting "oppose" or "support", but it's not clear to me exactly what is being opposed or supported. It also appears to me that if rules are made stricter than they presently are, I would be prohibited from reverting vandalism if that vandalism occurred on an archive page: the double-negative fallacy of "two wrongs don't make a right". Similarly, there are people who reply to an old thread after it has been archived, are we to be prevented from reverting those misplaced posts? One thing is certain - archives are not set in stone: for over ten years I have watchlisted each of the WP:VPT archive pages as it was created, and from these I have observed that we have a number of bots that do edit archives on a frequent basis. These include ClueBot III fixing links to archived content, as here; bots that fix "lint" errors, as here; and bots that either subst: or de-transclude templates that are pending TfD deletion, as here. Are we going to prevent bots doing this - or say "bots can do it but humans can't"? --Redrose64 🌹 (talk) 22:13, 16 October 2024 (UTC)

Redrose64, there is a lot of daylight between stricter rules that don't let you fix anything, and cosmetic bot indiscriminateness. As far as confusion about what is being voted on, that was my first impression as well. (I later adjusted it). Mathglot (talk) 03:06, 17 October 2024 (UTC)

OK, this edit by Trappist the monk (talk · contribs) just popped up on my watchlist. It falls under my previous description of "subst: or de-transclude templates that are pending TfD deletion", and that's OK, but look at the left-hand side - here we have Mr Serjeant Buzfuz (talk · contribs) posting a reply to a thread that's already in an archive page; and if we progress just a little bit further back, we get Xaosflux (talk · contribs) doing the same thing. Are these edits revertable, or would the proposal prevent that? In fact, in the history of that archive, there are 21 edits, of which 14 - that's just two-thirds - are legitimate archiving edits, two are valid gnoming, and five should not have been made at all. --Redrose64 🌹 (talk) 21:05, 17 October 2024 (UTC)

Redrose64, are those the faint echoes of WP:CREEP I hear wafting over the hills? If so, I get it. I don't know the answer to your question (in part because I think the Rfc statement was not optimally written). But the linked edits are surprising to me, as they do seem like the continuation of a conversation at an archive, and I think its fair to ask if we want that to happen, or if we prefer to have the discussion unarchived first. (The latter would be my preference.) I don't see those examples as materially different from someone ignoring the statements identifying a closed conversation and replying to the last comment in it, either within the box or outside of it, thus ignoring the shaded background, the header marking it closed, and the footer saying The discussion above is closed. Please do not modify it, and adding a comment either within the box or after it in the section. I doubt one or two such examples would generate a lot of excitement from anybody, maybe a 'Please don't...' right after it or on their UTP, and maybe nothing at all. If we had an editor doing that at bot speed however, I think there would eventually be some kind of reaction, and I doubt we have a policy that specifically covers violation of Please do not modify it exhortations, and it would probably be WP:DISRUPTION or just acting against consensus, if it bothered enough people. I think the current situation is like that.

This brings me back to the Rfc statement, because intentionally or not, it is worded in such a way as to restrict the scope to 'whether TPO covers this', such that a 'no TPO doesn't cover it' leaves the impression that the behavior is now approved for consensus-supported continuation, whereas in reality, this is not about TPO at all, but about disruption, and the analogy with continuing on at a closed discussion, or rather at many dozens of closed discussions, holds.

So, I can't really say whether the Rfc question would cover the linked cases, and I kind of don't care, in a CREEP-ish way; I am inclined to ignore it. What I care about, is if those editors you mentioned started doing that thirty times a day, endlessly. Then I might feel differently about them.

What do you think should happen here? I think reasonableness should rear its lovely head, we don't need new instructions or new interpretations of TPO, what we need is to determine whether some actions here are DISRUPTive, and if not, give them our blessing to carry on, and if so, ask them to stop. Mathglot (talk) 23:16, 17 October 2024 (UTC)

Leaving aside the implication that "30 times a day" is an outrageous amount of edits -- for instance, you yourself made well over 30 edits yesterday -- I would really appreciate if you stopped speaking for me as to what I intended with my statement. The scope is exactly as I said: do the guidelines about appropriate edits to talk page comments also apply to archived talk pages? I don't know how to state that any more clearly. To break it down further, since based on your comment you have the RFC backward:

• "Yes, TPO covers it" = the appropriate edits listed in TPO are also appropriate on archived talk pages
• "Yes, TPO covers it in cases of X, Y, Z" = X, Y, Z are appropriate edits to archived talk pages, and the rest of the edits listed in TPO are not
• "No, TPO doesn't cover it at all" = the appropriate edits listed in TPO are not also appropriate on archived talk pages (i.e., the exact opposite of "the behavior is now approved for consensus-supported continuation")

Gnomingstuff (talk) 01:08, 18 October 2024 (UTC)

I'm sorry I got it backwards, but that was my honest interpretation of it. It seems I am not the only one who did, or at least, who has been confused by the Rfc statement. That will likely make it harder for the closer, if 'oppose' means one thing for editor A, and the opposite for editor B. I guess we'll see how it all turns out.

Beyond that, when you wrote, "The scope is exactly as I said," I believe you. Unfortunately for the purposes of a neutral Rfc, I think the scope was poorly chosen. See § Non-neutral Rfc statement below. Mathglot (talk) 05:04, 18 October 2024 (UTC)

I'm guessing continuing a conversation onto an archive page is probably the result of subscribing to a section, which by design persists when a conversation is moved to another page, so a contributor could miss that they are editing an archive page. I agree with Mathglot that the conversation should be unarchived in that case, since most users do not watch archive pages, so the ongoing conversation can be seen by the talk page watchers. isaacl (talk) 23:47, 17 October 2024 (UTC)

Subscribing to a section or not, I've seen such behaviour several times on archive pages going back some years before subscribing was a thing (August 2022). One thing that I have noticed is that various paid staff do this fairly often, as here. --Redrose64 🌹 (talk) 21:59, 22 October 2024 (UTC)

• I have a very broad watchlist myself, and I saw one of these edits the other day, and I thought to myself "that's odd but I recognize this users' name and I'm sure they are doing what the edit sumarry says they are doing" and I moved on with my day. Is it super helpful? Maybe not, but I don't see how it is harmful. I work with archiving a lot and I often remove garbage from talk pages rather than archive it, but I've also noted that others are less careful and will archive talk content that rightly coud have just been removed from the page at any time. Saying it's a waste of time is not a valid argument in my opinion. How a user chooses to spend their time, so long as it is not harming the project, is their own business. Just Step Sideways ^{from this world ..... today} 22:17, 16 October 2024 (UTC)
• Read the top bit twice and still not sure what the scope of the question is. Is it just "does TPO apply to archives?" If it's narrower, is it specifically "do these prohibitions still apply to archives?" or is it "do these allowances still apply to archives?" Sounds like the latter? IMO there are only three good reasons to edit an archive, erring on the side of not editing for the simple fact that none of the original participants will see the change: to fix syntax errors, to remove egregious attacks/vandalism/BLP issues, and to update a link to a separately archived thread for posterity. — Rhododendrites ^talk \\ 00:52, 17 October 2024 (UTC)

  It's just "do the guidelines about editing comments in TPO apply to archived talk pages in addition to active talk pages." The RFC is because some people (myself included) feel strongly that they already do and some people feel strongly that they don't. Gnomingstuff (talk) 06:51, 17 October 2024 (UTC)

• How about this situation? Editor makes this edit, which includes a <ref> tag but no matching </ref> tag. It's not noticed at the time, and gets archived in the same state. Some months later, another archiving edit takes a valid <ref>...</ref> pair into the archive, and the MediaWiki software matches that new </ref> tag with the <ref> tag from months earlier and miles further up the page. Result: everything between those tags vanishes. But no way can we call this the result of vandalism, either in the archive or in the original - it was a simple mistake that anyone might make. Should it (i) be left alone because we don't alter archives even when they're clearly broken; or (ii) be fixed because otherwise we don't see any threads between April and November 2023? --Redrose64 🌹 (talk) 21:28, 26 October 2024 (UTC)

  Apply the general principle: preserve the appearance of the discussions while they were active. The closing </ref> tag should be added so all the appearance of all the other threads can be preserved. As far as I can see, the reply to the edit in question was visible on the talk page at the time, so its appearance is properly preserved as well. isaacl (talk) 21:40, 26 October 2024 (UTC)

  I concur with, Isaacl. In this case, you are fixing the discussion so it can be read as it was intended. So, if you want go ahead and do it if you feel like it helps people read the archive. I like Isaacl's "preserve the appearance" maxim. Jason Quinn (talk) 18:53, 4 November 2024 (UTC)

• Somebody reformatted the bold font in the comments of various editors (including mine) in this RfC. It's no big deal, so I'm not going to revert or really complain, but I think it's ironic that anyone would do that, on this particular talk page. --Tryptofish (talk) 22:29, 9 November 2024 (UTC)

Discussion on a taxonomy of nonconstructive edits

It might be helpful to break down the types of edits I have been classifying (or not classifying) as vandalism, since a few people have said that some things are OK to remove but not all. The taxonomy on WP:TPO is a good starting point but this discussion is getting more granular than it does.

I think these should be removed from all talk pages, including archives (these types of edits are a subset of WP:TPO, and more narrow than it):

• Threats, illegal, and defamatory material
• Blatant crude vandalism to other people's comments
• Non-trivial changes to other people's comments, e.g., someone changing someone's words to mean the opposite
• Blatant crude vandalism as standalone comments
• Self-insert vandalism, e.g., "jayden is awesome"
• Drive-by blanking of constructive comments
• Obvious spam
• ChatGPT nonsense -- not people simply using ChatGPT as a tool to write legitimate comments, but the weird repeated multi-header stuff that started in 2022 when ChatGPT came out, it's hard to describe but you know it when you see it
• Gibberish/nonconstructive test edits

I think these should not be removed from archives (includes some things in WP:TPO and some things that aren't):

• Comments by sockpuppets/banned users, because they're a legitimate part of the record (per Tryptofish)
• Similarly, vandalism that people have responded to or struck, e.g., declined semi-protected edit requests, unless it is defamatory/suppressable
• Debatably off-topic comments/soapboxing/statements of opinion, e.g., someone commenting on the talk page for a book that they liked it
• Heated arguments or personal attacks that are not vandalism, e.g., two people in a political dispute calling each other evil fascist assholes who should die
• False statements that aren't defamatory
• Typos/spelling/grammar errors, and/or people fixing other people's typos/spelling/grammar/syntax errors
• Any comments on user talk pages unless they're spam, defamatory, or suppressable

I could go either way/don't really care:

• Private information/"dox" like phone numbers/state ID numbers, a lot of this seems to fall into a gray area of "the person thought the talk page was email," often with a language barrier, and it's hard to tell intent
• Self-promotional comments that are probably spam but it isn't obvious
• Fixing formatting, layout and/or confusing stray markup like "Insert bold text here" inside otherwise constructive comments - I was doing this for a bit then stopped because it was too tedious even for me
• Stuff I am 99% sure is vandalism but cannot prove because I truly do not find the diff and the original text because it's from some long-lost merged page or manually copied over or just... not there. — Preceding unsigned comment added by Gnomingstuff (talk • contribs) 07:23, 17 October 2024 (UTC)

I for one agree with all the items you say that should always be removed except Self-insert vandalism and gibberish/nonconstructive test edits, which I'd deal with on a case-by-case basis (or just leave them there, to be honest ... I think they'd do a minimal amount of harm). I think it'd be helpful if you could provide an example of the "weird repeated multi-header stuff "; I don't know what you're talking about. I wouldn't disagree with anything in the iother two lists. Graham87 (talk) 12:37, 17 October 2024 (UTC)

Thanks for opening this section! Having the space to think through and discuss this taxonomy is definitely useful, since I imagine people will have a range of opinions about which types of edits should or shouldn't be removed. From my own POV, and speaking about archived talk pages specifically, I think the following information should be removed or reverted from archived talk pages: threats/illegal/defamatory material, non-trivial edits to other users' comments (including blanking of constructive comments), obvious spam or promotional edits, and oversightable private information. I think the commonality that these types of edit share is that either their being visible has the potential to cause real-world harm (illegal material, private info) or their being present at all subverts the goals of the page (by hijacking the page as a promotional platform or by distorting the record of what was said). For other types of non-constructive edit, such as self-insert vandalism or gibberish, my opinion is that (in Talk: space) their harm to the encyclopedia comes mainly in their ability to disrupt productive discussion of the page topic. Thus, talk pages are impacted relatively significantly by new vandalism, as it can clog watchlists or derail ongoing discussions - but the harm of that same vandalism is likely to decrease substantially once it's grown stale and passed into an archive. ModernDayTrilobite (talk • contribs) 15:20, 17 October 2024 (UTC)

This diff [4] is a good example. Might not be ChatGPT specifically so much as some kind of mobile phone AI thing but I almost never see this pattern of edit before 2022. Gnomingstuff (talk) 15:33, 17 October 2024 (UTC)

Yeah I know what you mean now. They're not always multi-header though. I always remove those when I encounter them on live talk pages, but I wouldn't remove them from talk page archives, because they cause minimal harm there. Graham87 (talk) 04:22, 18 October 2024 (UTC)

As I said in my RfC comment, I find it useful to ask whether the revert will restore the record of the discussion to what it was at the time of the discussion. For me, that's the bottom line.

If something was included in the discussion originally, and left unaltered up to the time of archiving, then the reasons for reverting it from the archive later need to be pretty compelling. As noted, a sock who was undetected at the time is one example of something best not reverted later, but I don't think it's the only one. I wouldn't correct a spelling error or other typo that was left at the time. Threats, defamation, and the like should generally have been dealt with at the time, so such material should be examined carefully if discovered later in an archive. If it should have been dealt with before archiving but wasn't, then it should be corrected according to the WP:Harassment, WP:BLP, and WP:NLT policies, indicating something like "redacted" if it was responded to at the time. (If it was added by a vandal after archiving, it should be reverted, and rev-deled or oversighted if appropriate – because there's no reason to preserve it as part of the original discussion.)

Other kinds of post-archiving edits should be encouraged, because they actually help to preserve the original discussion. One not listed above, but that has been discussed higher up in the discussion, is when a bot fixes something like a linter error. That's a good edit, because it fixes a formatting error in such a way as to restore the appearance of the discussion to how it looked at the time. --Tryptofish (talk) 22:49, 17 October 2024 (UTC)

IMO we can split the types of edits into four categories:

1. Content so problematic that it requires revision deletion or oversight: revdel/oversight as appropriate, regardless of when it was added (although do be aware this has the potential to create a Streisand effect)
2. Content added/changed before archiving that doesn't require revdel or oversight: Do not change it.
3. Edits made after archiving that add, change or remove content: revert
4. Technical changes (linter errors, substing templates, etc) to maintain the archive integrity that don't affect content: Enact.

This is my understanding of the guidelines as they stand at present and is my preference for what the guidelines should be. There will be occasional exceptions, but they will be occasional and must come with a strong justification. Thryduulf (talk) 23:00, 17 October 2024 (UTC)

"Threats, defamation, and the like should generally have been dealt with at the time" -- Yes, they should, as should any vandalism. The problem is that that they are not, and this is not happening to the tune of thousands of instances.

I really do not understand the rationale for grandfathering in vandalism that would have been perfectly acceptable and encouraged to revert if it was found just 1 day prior to when an archive happened to be made. There should be no reason to ever keep it around. Revdel is an extreme bar and vandalism does not have to meet it to be removed. Gnomingstuff (talk) 01:00, 18 October 2024 (UTC)

You need to look more carefully at what I actually said. I'm not arguing for grandfathering in vandalism. For vandalism that falls below the high bar for rev-del (or the even higher bar for oversight), I'm saying that the reverts should restore the original discussion to what it was when the discussion took place. If the vandalism was somehow replied to, then use something like "redacted". If it altered the original good-faith comments of another editor, then restore what the good-faith editor originally said (I guess I didn't make that latter point clearly enough). But if it had no effect at the time, then correcting it later is like correcting a spelling error later, not very urgent or necessary. --Tryptofish (talk) 21:02, 18 October 2024 (UTC)

Perfect! This is in reply to Thryduulf, see my comment below. In summary, do not adjust archives unless actually needed. Johnuniq (talk) — Preceding undated comment added 00:53, 18 October 2024 (UTC)

On another matter where there seems to be confusion, Johnuniq, your edit summary in [5] asked me whether I had accidentally omitted the word "not" when I wrote "should be encouraged". I meant "should be encouraged", as written. --Tryptofish (talk) 20:54, 18 October 2024 (UTC)

@Johnuniq: it's not clear whether you are replying to me or to Gnomingstuff. Thryduulf (talk) — Preceding undated comment added 01:23, 18 October 2024 (UTC)

I was endorsing your comments which correctly outline what should happen with an archive. Unfortunately, Gnomingstuff inserted their comment (diff) above mine which changed its meaning. Please do not do that! Johnuniq (talk) 02:04, 18 October 2024 (UTC)

@Gnomingstuff: Re failure to "understand the rationale", you might respond to the substance of comments such as mine at 06:10, 17 October 2024. Johnuniq (talk) 02:10, 18 October 2024 (UTC)

@Gnomingstuff: Re failure to "understand the rationale", see assume good faith. There are editors here who have, in good faith, views different from yours. That does not imply that they favor "grandfathering in vandalism". This discussion is not about deciding between the views of one group who are opposed to vandalism, as policy requires, and some other folks, who, unaccountably, are in favor of vandalism, and are trying to "grandfather it in". If that is your view of opposition argumentation, it is little wonder you don't understand it; I wouldn't either, under that interpretation. Mathglot (talk) 03:58, 18 October 2024 (UTC)

The reason I call it "grandfathering in vandalism" is that when (not if, when) vandalism to a talk page goes undetected before the page was archived, and if there is a restriction that vandalism cannot be removed from archived talk pages with the reasoning that it was "part of the state of the discussion at the time," then that vandalism will exist forever. It will be "grandfathered in" on the rationale of being around for a long time. This is the inevitable consequence of what you are arguing for. Gnomingstuff (talk) 22:16, 18 October 2024 (UTC)

Just wanted to update the list here as far as kinds of edits people have brought up that I didn't think of.

Edits I personally believe are good to make to archives:

- Attributing unsigned comments that the bot didn't catch for some reason

- Non-free images: I've never seen this and it feels like it'd always fall into WP:IMAGEPOL or Commons jurisdiction, but sure

Edits I don't think should be made to archives:

- Updates to signatures (unless the signature itself was previously vandalized): these are decorative and not really part of a discussion

- Updates to usernames: way, way, way, way too much potential for confusion

- Most things in WP:NOTWEBHOST: a lot of this isn't necessarily off-topic and it feels a bit callous to remove things like memorials from talk pages at all, let alone archived ones

Edits I could go either way on/don't care:

- Broken links: clear value in this, but then you get into issues of which version of the page the original writer intended and whether that version is the one that got archived and it feels really easy to change the meaning

- Linter stuff: this is just so trivial, I'm baffled by the arguments that linter stuff is worse than vandalism and at the same urgency level of revision deletion(!!!). Gnomingstuff (talk) 02:55, 9 November 2024 (UTC)

Gnomingstuff: This is a pretty good list (except I know nothing about the ChatGPT stuff) for me, too, but my thoughts are more closely aligned with Graham87 (talk · contribs)'s: case-by-case on the small stuff, esp. as it appears to bother watchers who don't like the traffic. The four-way dichotomy from Thryduulf is too harsh ("do not change it") on the second point, IMO, as I think a clean-up in Aisle 9 is useful, even if nobody much goes down that aisle anymore. But again: case-by-case. That kind of sidesteps the RFC question, I realize, but I think it's more likely that TPO should explicitly mention changes to archived pages, possibly with a narrowing of scope for them. — JohnFromPinckney (talk / edits) 11:35, 18 October 2024 (UTC)

Non-neutral Rfc statement

In my opinion, the Rfc statement fails the WP:RFCNEUTRAL paragraph, which is entitled, § Statement should be neutral and brief.

The Rfc question asks whether the TPO guideline applies to archived Talk pages, and then is immediately followed by a longish statement defending archive page edits as within the bounds defined by TPO. But from my point of view, that approach is both a straw man argument, and also excludes the major reason that some people might oppose such edits, which is not TPO but the WP:DISRUPTION guideline.

The user Talk page discussion that sparked this Rfc never mentioned TPO even once. What was mentioned there, was disruption and achieving a consensus to make such edits. A neutral Rfc statement would have been one that included just the behavior itself formulated as a yes-no question, something like: "Are edits to archived talk pages to be encouraged if they involve repair of vandalism?" without asking about TPO, which was not part of the original objection.

By couching the question as one about whether WP:TPO applies, other questions (like, does WP:DISRUPT apply?) seem to be off-topic, or at least, are not uppermost in the minds of responders, who, understandably, respond to the question asked. In my view, it is the wrong question—it is the straw man question. It is probably too late to do anything about this now, but it is a shame, as the chief question is for the most part not being addressed, and was relegated to the margins in favor of a narrower scope chosen at the outset based on a single guideline that may favor that view a little bit more than a neutral statement might have, as manifested in the heading at the top of this section. It is instructive to note, for example, that the subsections originally entitled "§ Survey" and "§ Discussion", were changed to "§ Survey re TPO Guidelines" and "§ Discussion re TPO Guidelines" by a third party, so the Rfc title appears to have had an influence on how this Rfc is viewed, in ways that are counterproductive, imho. Mathglot (talk) 04:53, 18 October 2024 (UTC)

You are correct that the RfC is broken. In a couple of cases, it is not clear what the Support and Oppose votes mean. The issue is whether archives should be gnomed which is nothing to do with TPO. Even if this RfC resulted in a supported close, there would be no actionable result because it would still be disruptive to make gnoming edits on archives. Johnuniq (talk) 04:58, 18 October 2024 (UTC)

I'm not sure I understand your point here. Are you saying that you personally believe that gnoming archives impedes building the encylopedia, regardless of whether consensus is that archives should be gnomed? jlwoodwa (talk) 19:35, 18 October 2024 (UTC)

I have added a note clarifying what "support," "partial support/oppose," and "oppose" mean. WP:TPO is explicitly about appropriate edits to other people's comments on talk pages. That is what TPO stands for: Talk Page (Others' comments). My concern is vandalism, others' concerns are linting errors etc. All of these are explicitly listed in WP:TPO, very clearly.

I don't understand what's unclear about this, and I truly do not know how on earth I can make it any clearer for people. Gnomingstuff (talk) 22:06, 18 October 2024 (UTC)

I don't see problems that would rise to the level of disqualifying this RfC. For what it's worth, I interpreted the RfC question as supporting or opposing the kinds of edits that you have made, and I answered it on that basis. But your clarification is fine with me, and it really doesn't change how I would have answered. --Tryptofish (talk) 22:17, 18 October 2024 (UTC)

There's another dimension that your choices don't allow for: if the edits to be reverted or modified were made before or after archiving. Edits after archiving alter how the discussion appeared when active, so reverting them restores the original appearance. If the edits were made before archiving, reverting or modifying them would be confusing since it would alter the history of discussion, so there needs to be a strong justification for it. The considerations for this may not fall neatly into specific categories and may need to be judged on a case-by-case basis. isaacl (talk) 22:43, 18 October 2024 (UTC)

I'm not opposed at all to reverting vandalism that was made after archiving. I assumed that was implied by my previous statements, but I didn't spell it out.

As far as "altering the history of discussion," though, the kind of timeline I'm talking about usually looks something like this:

• 2007: Some discussion takes place.
• 2009: Somebody vandalizes the discussion in one of the above ways. From what I have seen, this almost happens months or years later, after the discussion has been abandoned. Alternatively, someone creates an unconstructive edit in a separate header that receives no acknowledgement.
• 2010: The page is archived, including the vandalism from 2009 that went undetected.

In cases like this -- which, again, are the overwhelming majority of undetected vandalism cases on talk pages -- I really don't see the benefit of preserving the vandalized state of the discussion in 2010, since its existence distorts the actual discussion from 2007. Even gibberish/test edits make discussions harder to follow, while adding nothing of value that, IMO, is worth preserving. If it would have been uncontroversially reverted in 2009, or uncontroversially reverted in 2024 if it appeared anywhere else but an archived talk page (or sandbox), I truly don't see why it should not be reverted in 2024 just because it got archived in the interim. Gnomingstuff (talk) 22:55, 18 October 2024 (UTC)

Sure, I didn't mention anything about your viewpoint. The point is that the three options you provided don't consider when the edit being considered for reversion/modification was made. I think the answer is different based on whether or not the edit was made during the active discussion, or sometime afterwards with no responses. Thus personally, as several people have alluded to in their comments, I would prefer to provide guidance based on general principles: preserve the state of the discussion at the time active participation ceased. If you're not sure about whether or not an edit should be reverted or modified, you can ask others for guidance. Also, I suggest giving priority to removing vandalism from talk pages, so it doesn't get into the archives. isaacl (talk) 23:36, 18 October 2024 (UTC)

Yes, it certainly seems like one consequence might be a heightened sense of urgency to catch vandalism before pages get archived, especially since archiving is often done automatically by bots who don't/can't check for vandalism beforehand. "There is no deadline," except when one is created I guess. Gnomingstuff (talk) 04:30, 19 October 2024 (UTC)

(ADDENDUM) I am also unsure how WP:DISRUPT is relevant to the discussion about the broader guidelines to WP:TPO. I am very familiar with that page, and the only things on it that seem even remotely related to editing archived talk pages are:

• "Wrongly accusing others of vandalism" (indirectly via WP:TENDENTIOUS, mentioned on WP:DISRUPT). The guideline there seems to be mostly about how one communicates with the other editors, and when they're not around anymore no communication is really taking place. But sure, I'm willing to be more granular in edit summaries. (It's noteworthy, however, that WP:TENDENTIOUS contains a section (WP:AOTE) about how accusing others of tendentious editing can be inflammatory without clear evidence.)
• "Fails to engage in consensus building/rejects community input": The reason I made this RFC was to receive community input. I probably should have made it sooner, but now that I have, I don't see how these apply.
• WP:MEATBOT: Not mentioned on WP:DISRUPT but adjacent to it, even though I'm the only person who brought it up. As I said I have tried to take this into account, including spacing edits out more. However, that guideline also states that merely making a lot of edits is not necessarily disruptive.
• WP:COSMETICBOT: This is about cosmetic vs. substantive edits, which has come up in this RFC. That guideline states clearly that "changes that are typically considered substantive affect something visible to readers and consumers of Wikipedia." The definition of "cosmetic" is not in that guideline, but it is in Wikipedia:Bots/Dictionary: "A cosmetic edit is one that doesn't change the output HTML or readable text of a page"; the example they give is "whitespace optimization." I don't believe there is any instance where reverting vandalism or test edits would be considered cosmetic by this definition. Like... it indisputably changes both the output HTML and the readable text.

Notably not anywhere in WP:DISRUPT or associated pages are:

• Editing archived pages. Likewise, WP:ARCHIVE makes no mention of anything disruptive.
• Adding edits that appear on editors' watchlists; WP:WATCHLIST makes no mention of that being disruptive either. The only mention is above, in regard to cosmetic edits.
• Reverting vandalism or test edits; the only mention of vandalism is in reference to whether vandalism itself is disruptive.

I am not aware of anything else in policy that is relevant. Gnomingstuff (talk) Gnomingstuff (talk) 23:25, 18 October 2024 (UTC)

The behavior visible in these Talk page contributions, as well as on your Talk page appear to fit WP:DISRUPTSIGNS numbers 1, 4, and 5. Mitigating the last one was bringing it here for discussion, which was a positive step, although I disagree with the statement of it for reasons previously stated. Mathglot (talk) 00:02, 19 October 2024 (UTC)

I have addressed all of #1, #4, and #5 -- specifically, those exact three -- in the above bullet points. The first bullet point addresses #1. The second bullet points address #4 and #5, as they go together. Meanwhile, so far you have not provided any grounding for your accusation besides "because I said so." You have also accused me of making "cosmetic bot indiscriminateness" based on some unstated definition of "cosmetic" of your own that contradicts the definition set forth in the project guidelines, and accused me of "acting against consensus" when no such consensus existed. As WP:TENDENTIOUS clearly states:

"Making accusations of tendentious editing can be inflammatory and hence these accusations may not be helpful in a dispute. It can be seen as a personal attack if tendentious editing is alleged without clear evidence that the other's action meets the criteria set forth on this page." Gnomingstuff (talk) 04:18, 19 October 2024 (UTC)

You have not engaged with any points made by people who do not want archives "fixed". You are responsible for a lot of wasted time here. That is the definition of disruption. Do not fiddle with archives. Johnuniq (talk) 04:38, 19 October 2024 (UTC)

...I feel like all I have done for the past few days is respond to and engage with points made by "people who do not want archives 'fixed.'" I am doing that right now. This comment is me doing that. I really don't know what points you think I am not engaging with -- unless by "engaging with" you mean "agreeing with," which is not what "engaging with" means.

I resent the implication that I am "responsible for" anyone's time, and it feels like a personal attack. Your time is your own. My time is my own. You are free to spend it doing something else, as am I. Gnomingstuff (talk) 04:50, 19 October 2024 (UTC)

(Update) You specifically mentioned your comment at 06:10, 17 October 2024. I did in fact reply to that comment, which you saw. I addressed the "setting a bad precedent" part elsewhere -- essentially, it's a slippery-slope argument that addresses the remote possibility of people suddenly starting to do something that they have not done in over 10 years, at a time when we had many more active editors. As far as "having to examine history," I have been providing the diffs for any edits to archived pages that do not contain timestamps, so all someone has to do is check that diff, not the entire history -- or, for that matter, assume good faith that an editor in good standing is making legitimate edits Gnomingstuff (talk) 05:01, 19 October 2024 (UTC)

You are changing archives. We have to spend time here explaining why that is a bad idea. There are exceptions (linter, serious BLP problems, copyvios) where removal would be ok, but removing nonsense means that the original discussion is changed from how it was when people commented. If archives were never read, we could just delete them. The point of an archive is to allow easy searching for old discussions to see what has happened before. When doing that, we should not have to waste time checking the history of the archive, then checking diffs of passers-by "cleaning" the archive to see whether any meanings have been changed. Just do not do it. Johnuniq (talk) 06:36, 19 October 2024 (UTC)

Shall anyone here volunteer to have me go write "love too diarrhea shit my pants" in a randomly selected archive of their user talk page, and then we can see how many days of not being detected it takes for it to become an immutable permanent part of the page that nobody is allowed to revert? jp×g🗯️ 12:54, 19 October 2024 (UTC)

Your use of "we" here is telling. You are framing this situation as my disregarding the orders of some unanimous authority/authorities. In reality -- as you can see clearly throughout this discussion and the survey section in particular -- there is a range of opinions on this topic, some of which align with yours and some of which don't. Before the RFC, I was told by two people that reverting vandalism to archives was acceptable (plus some other people using the "thank you for this edit" feature on similar edits to archives, I do not have a list but it certainly was more than two); I created the RFC when two people had raised objections.

What is happening here is that you are choosing to spend time here explaining why you personally think editing archives is a bad idea. But as of right now, that is as far as it goes. Gnomingstuff (talk) 17:13, 20 October 2024 (UTC)

I was actually curious about the number of people who approved of similar edits to archives, so I went and checked. It's 5. I've redacted the names because if they don't want to be dragged into this mess I'm not going to do that, but they are five separate people, all of whom are editors in good standing and only one of whom is involved in this discussion.

• 23 August 2024: [user #1] thanked you for your edit on ‪[page].
• 20 August 2024: [user #2] thanked you for your edit on [page].
• 18 August 2024: [user #3] thanked you for your edit on [page].
• 27 June 2024: [user #4] thanked you for your edit on [page].
• 18 May 2023: [user #5] thanked you for your edit on [page].

Gnomingstuff (talk) 22:43, 21 October 2024 (UTC)

• Sorry for not reading the whole previous conversation, but in my defense the last part seems really boring and petty. At any rate, it is very kooky to say that people can't revert someone vandalizing ana rchived talk page comment with "jklsadjklfahjklfw3lk" and "pee pee poo poo" because it spams up watchlists. Come on. At this point maybe we should just admit that watchlists are a disruptive worthless timesink and remove them from MediaWiki entirely, because not only are they used as a justification for preventing people from fixing the normal stuff, but now it is literally being argued that we should let people vandalize pages because reverting it would spam up watchlists? Deeply unserious. I disagree with removing actual comments decades after the fact, if they were actual comments (e.g. I am resolutely against the deranged practice some people have of deleting talk page threads from 2012 because they think the person's asking a dumb question). But if something was not even made as a comment, just a random cigarette butt thrown onto the talk page, then who cares. jp×g🗯️ 12:52, 19 October 2024 (UTC)

  To be clear: why the heck would this be an improvement? That wasn't even the original comment! jp×g🗯️ 12:59, 19 October 2024 (UTC)

  Yes, it was only a partial revert of this edit, and this vandalism wasn't reverted at all. --Redrose64 🌹 (talk) 15:25, 19 October 2024 (UTC)

  Thanks for pointing this oversight out. I use keyword search to find instances of vandalism so it's very possible to miss stuff. That being said, I am now afraid to change it lest someone think I am doing something wrong. Gnomingstuff (talk) 04:41, 20 October 2024 (UTC)

  I'm not sure whose comments you're referring to with "boring and petty" but I apologize if they are mine. I can be boring but I try to not be petty. Gnomingstuff (talk) 05:20, 20 October 2024 (UTC)

Use of FAQ template not documented

{{FAQ}} appears to be used on some talk pages but I couldn't see a mention about its use on this guideline. Commander Keane (talk) 00:39, 1 November 2024 (UTC)

I went ahead and created a section about using the FAQ template. —The Mountain of Eden (talk) 15:29, 1 November 2024 (UTC)

Suggest in-article references to FAQs?

Does anyone read talk FAQs before boldly editing articles or making proposals on talk? Granted, FAQs are useful to cite when responding to an edit or proposal. But can we make them more preventative? Perhaps we should mention the option of adding in-article footnotes or hidden text that refer to the appropriate FAQ. - Butwhatdoiknow (talk) 17:07, 1 November 2024 (UTC)

That's a good idea. I have incorporated it into the project page. The Mountain of Eden (talk) 17:14, 1 November 2024 (UTC)

We could recommend that individual articles can be given a brief editnotice, something like "Before editing this article, please read the frequently-asked questions", with the last three words linked to the approptiate talk page section or subpage. --Redrose64 🌹 (talk) 08:58, 2 November 2024 (UTC)

That would have been a good idea if ordinary editors could create Editnotices. According to WP:EDNO, "only administrators, page movers, and template editors can create or edit editnotices in any .. namespace [that's not User or User talk]".

So perhaps we could suggest that editors should request an administrator to add an editnotice to read the FAQ (with a link to the FAQ) for edits that have been repeatedly rejected. The only question would be what would be the appropriate noticeboard for such requests? The Mountain of Eden (talk) 20:07, 2 November 2024 (UTC)

"Wikipedia:STAYONTOPIC" listed at Redirects for discussion

The redirect Wikipedia:STAYONTOPIC has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2024 November 8 § Wikipedia:STAYONTOPIC until a consensus is reached. Veverve (talk) 09:54, 8 November 2024 (UTC)

re-factoring/editing of others comments

While removing unambiguous personal attack is often reasonable, I question this editorializing of my comment by a third party and I feel it's an overzealous editing. Is this within the leeway given to any editor for what they consider "potentially PA" ? Graywalls (talk) 18:04, 18 November 2024 (UTC)

Graywalls, In my opinion, that is a TPO violation, and you can revert their edit, citing WP:TPO in your edit summary. To demonstrate an abundance of good faith with respect to that editor, instead of simply reverting their edit (which was also made in good faith), you could remove it from inside your comment and move it to *after* your comment,properly indented, while taking care to mention your change in the edit summary as "refactoring" their interpolated comment (which is also a TPO violation on your part of their comment, but a justified one imho). This both acknowledges their disagreement with your edit, while still preserving the original form of your comment undisturbed, along with a refactored version of their comment. Be sure to retain their wording, even if critical of you, to demonstrate your good faith. Since they didn't sign their interpolated comment, you should append the following to the end of their comment after you move it down: {{subst:unsig|Dustfreeworld|16:36, 18 November 2024 (UTC)}}. A little tricky, but that's how I would handle it. Mathglot (talk) 10:50, 20 November 2024 (UTC)