Wikipedia talk:Criteria for speedy deletion/Archive 86

CSD for AI generated content
Should there be a CSD for when someone uses AI like Chat GPT, Bing and Brad etc Qwv (talk) 21:48, 18 March 2023 (UTC)
 * This was discussed here in January. It's a good idea in principle, but currently we don't have a completely accurate way of determining whether text is AI-generated (there are various detectors online, but they're a long way from being perfect). Extraordinary Writ (talk) 21:55, 18 March 2023 (UTC)
 * No, because in such cases there ought to be a record of a consensus being reached that the contribution is a product of AI. That is, unless there's another AI product out there that can incontrovertibly detect AI authorship of a text. Largoplazo (talk) 21:57, 18 March 2023 (UTC)
 * I don't see what would be the point anyways. The editor submitting it is taking responsibility for the content. If said content is factual, not copyvioed to some other source, and has a reliable source as a reference that it is accurately summarizing from, then that content would be fine. How the words themselves were created is irrelevant. If the content being added violates existing policies and guidelines, then it should be removed regardless of said word construction source as well. If AI is used to construct better prose while following all of our rules in general, then that's just a benefit. Otherwise, our existing rules are more than good enough. Silver  seren C 22:06, 18 March 2023 (UTC)
 * Does this pass the four criteria at the top of this page? Phil Bridger (talk) 22:27, 18 March 2023 (UTC)
 * I'm thinking one and two, clearly not. Three, disputable. Four, probably OK. Jclemens (talk) 01:26, 19 March 2023 (UTC)
 * If it isn't a copyright violation (hard to say with AI generated content), then WP:PROD/WP:AFD is the better way to go until we have enough experience with AI content to distill it into criteria that pass WP:NEWCSD. —Kusma (talk) 09:53, 19 March 2023 (UTC)
 * I don't think a speedy deletion criterion is going to work - how is a random admin supposed to check that text is AI-generated? I certainly don't know of any tools that can be trusted to judge that. Also, is all AI generated text always so bad as to warrant undiscussed deletion? Jo-Jo Eumerus (talk) 08:32, 19 March 2023 (UTC)
 * No there should not be a speedy delete criterion for this. Some AI generated stuff is fine. (Even when typing here, I use an AI spell corrector). And as mentioned, it is not clear if something is AI or not. It is easier to tell if it has other problems, and they can be considered to determine of the page is delete-worthy. Graeme Bartlett (talk) 11:08, 20 March 2023 (UTC)
 * No. Not only should being written by an AI be speedy deletion reason, it shouldn't be a reason for deletion at all. If it's a copyright violation, then treat it as any other copyright violation, if it's unreferenced treat it as any other unreferenced content, and similarly for any other given problem - treat it exactly as you would human-written content that has the same problem. Thryduulf (talk) 15:31, 20 March 2023 (UTC)
 * The way we would treat human-written content by contributors who deliberately and confidently introduced false sources backing false claims into their content would be to indefinitely block them and then G5-delete any new contributions added after the block by editors appearing to follow the same patterns. So are you saying that's what we should do to AI-generated content as well? Indefinitely block anyone using ChatGPT as a WP:MEATPUPPET of a banned editor? —David Eppstein (talk) 16:58, 25 April 2023 (UTC)
 * No, because ChatGPT is not a banned user, and we block people for their own behaviour not the behaviour of other people. If someone is deliberately introducing false sources backing false claims, then they should very likely be blocked whether they are doing so via ChatGPT or any other method, and existing CSD criteria and other processes will cope with their content just fine. However I would be very surprised if most people using ChatGPT are doing so believing that its output is truthful and correct, those people should generally only be blocked if they are unable or unwilling to respond appropriately when learning that it isn't always (it's worth noting that ChatGPT content can be correct). Thryduulf (talk) 21:58, 25 April 2023 (UTC)
 * We block meatpuppets of vandals even when the meatpuppet is a naive dupe rather than a deliberate vandal. I would argue that someone asking ChatGPT to write an article, and getting back falsely sourced material that the human user is unable to recognize as false, is in a similar category. The purpose of blocking is not to infer the intent of the blocked editor and punish them for bad intent; it is to protect the encyclopedia. —David Eppstein (talk) 23:08, 25 April 2023 (UTC)
 * If someone is posting material that other editors can tell is very clearly false but they insist is not, then they should be be (partially) blocked, but not for meatpuppetry - they are not acting on the instructions of a banned user. The point is that whether the source of the material being posted is ChatGPT or something else is irrelevant - existing processes handle everything that needs to be handled without allowing or even requiring the deletion of material that is not problematic. Thryduulf (talk) 00:47, 26 April 2023 (UTC)

Proposal: New CSD criteria re: Large Language Model/AI Created articles
Proposing the following new criteria for speedy deletion: '''A12. Articles created by large language models or other similar artificial intelligence programs'''

This applies to any recently created article for which it is reasonably determined to have been created by artificial intelligence Large language models (such as, but not limited to ChatGPT) and which the only substantial content of the page was added by its author.
 * Rationale

Recently, several large batches of articles, obviously created by LLMs, and rife with errors, often incomprehensible text, invented or unrelated sources, etc. have been clogging noticeboards and AFD and PROD queues. In vanishingly close to 100% of these creations, it has been determined that the material is not worth keeping, and waiting for PROD or AFD to run its course seems unnecessary here. Recently created and substantial content are presumed to have the same definitions as in other criteria that use those (i.e. G7 and A10).

Discussion (Large Language Model/AI Created articles)
Support as nominator. -- Jayron 32 15:40, 25 April 2023 (UTC)

Preliminary oppose for the reasons in the sections farther up: I don't think a speedy deletion criterion is going to work - how is a random admin supposed to check that text is AI-generated? I certainly don't know of any tools that can be trusted to judge that. Also, is all AI generated text always so bad as to warrant undiscussed deletion? At a minimum, we'd need to specify how to identify such articles. And I'd like to add a clause "Articles written after 1 January 2024 are not eligible" Jo-Jo Eumerus (talk) 15:46, 25 April 2023 (UTC)


 * What special event is supposed to happen on that date? -- Jayron 32 16:02, 25 April 2023 (UTC)
 * An evaluation of whether the AIs by that time have learned to write proper wiki content. I am not optimistic that it will be the case by that date, but some cut-off is needed. Jo-Jo Eumerus (talk) 16:22, 25 April 2023 (UTC)
 * If it ever happens, the CSD can always be repealed. We have repealed several CSDs before.  Setting an expiration date seems unwise, given that we can always just repeal it when and if it becomes inappropriate.  -- Jayron 32 16:27, 25 April 2023 (UTC)


 * Oppose and speedily close. This proposal fails to address any of the reasons any of the multiple other recent proposals have failed to gain consensus, particularly the egregious failures of WP:NEWCSD points 1 (objective) and 4 (nonredundant), and whether it meets point 2 (uncontestable) is debatable. Thryduulf (talk) 22:04, 25 April 2023 (UTC)
 * Oppose: Redundant to existing criteria G1 (patent nonsense) and G3 (pure vandalism and blatant hoaxes). ViperSnake151   Talk  06:06, 26 April 2023 (UTC)
 * Support If the LLM origin is obvious, the article is clearly bad in a way that can't be fixed through editing. I do not think this is redundant with other criteria, since the purpose of stochastic-parrot technology is to generate nonsense that is superficially plausible, i.e., not "patent" or "blatant" in the way CSD's apply those terms. The supposed vagueness of this criterion is no worse than that of the extant CSD's. XOR&#39;easter (talk) 17:43, 26 April 2023 (UTC)
 * If it isn't "patent" or "blatant" then it's not obvious enough to be speedily deleted under any criterion and needs to be discussed at AfD. There is also currently no way (as far as I am aware) to reliably determine whether something is LLM-generated or not. We are slowly working our way through making the existing criteria more objective, we don't need a new criterion that is vaguer than the already highly abused and misused G6. Thryduulf (talk) 19:15, 26 April 2023 (UTC)
 * If the "parrots" thing is a reference to the Gebru/Bender paper, it may be advisable to catch up with more recent literature (the field as it existed in March 2021 is almost unrecognizably different from what exists today). jp×g 10:57, 4 May 2023 (UTC)


 * Oppose By your own description of the content of such articles, other existing deletion criteria already apply. And if we end up with an actually good article somehow that doesn't meet any of the criteria, then there would be no reason to delete said content in the first place, making a CSD criteria such as this actually detrimental as a whole. Silver  seren C 00:29, 3 May 2023 (UTC)
 * Support these AI engines are simply not capable of producing trustworthy content in their current incarnations, as they don't verify that the material they produce is accurate. The content may well qualify for deletion under other criteria or other policies but that typically requires work, e.g. doing research to prove that the claims made are not true. Speedy deletion is intended for such cases. In any case AI generated content frequently is not patent nonsense (if it makes any sense at all it's not patent nonsense) and since it was added in good faith it doesn't fall under vandalism either.  Hut 8.5  17:03, 3 May 2023 (UTC)
 * Speedy deletion is absolutely not intended for pages where you need to do research to determine whether the page should be deleted, indeed it is for the exact opposite of those cases. If the material produced by a LLM is comprehensible and added in good faith then there is no reason to delete it unless it is inaccurate (needs research), unencyclopaedic (A7, A9, A10, A11 or AfD) or a copyright violation (G12) - in other words if it should be speedy deletable it already is, if it isn't it shouldn't be speedily deleted. Thryduulf (talk) 00:40, 4 May 2023 (UTC)
 * I would claim that many G3s (i.e. the hoax variety) require some amount "research" inasmuch as "running a basic Web search for the subject of the article and seeing if there are zero results" or the like is research. jp×g 11:00, 4 May 2023 (UTC)
 * Oppose per Thryduulf. Redundant to the extent that it would be helpful. Elli (talk &#124; contribs) 17:20, 3 May 2023 (UTC)
 * Oppose due to the lack of reasonable use or enforceability. Wikipedia currently does not have a 100% fool-proof way to detect LLMs, which is what we should have. Perhaps @Silver seren puts my thoughts best; just because something is LLM-created doesn't mean it needs to be repealed and replaced.  Invading Invader  (userpage, talk) 21:56, 3 May 2023 (UTC)

G3
Something that was mentioned a while ago on this talk page, which I don't think got a very thorough examination, was the idea of expanding G3 (hoax) speedies to include LLM-generated content with fake sources. Basically, the main way in which these models become problematic for us is when they make false claims, so this would inherently limit the application of CSD to articles that really warranted it and avoid the arduous process of creating a new CSD. The reason I bring it up here is because, previously, I have seen G3 nominations (for pages with fake LLM references that literally cited nonexistent papers) declined as not being obvious hoaxes, so it might be worth pre-empting this by making it explicitly part of the CSD. jp×g 10:52, 4 May 2023 (UTC)


 * I don't see a need for change - if it is sufficiently obvious that it is a hoax then it is already speedy deletable, if it isn't sufficiently obvious then it shouldn't be speedily deleted. If you disagree with someone's interpretation in a specific instance then discuss that specific instance with them, but remember that if people in good faith disagree with whether a page meets a speedy deletion criterion then it does not, because CSD is explicitly only for "the most obvious cases" (this applies to every criterion). Thryduulf (talk) 11:05, 4 May 2023 (UTC)

Does U2 apply to a renamed user's page left for privacy ?
According to public logs, the retailed user Hobomok had requested a deletion of his userpage and renaming to Renamed User iusdhf. K6ka recreated this page before I found it in Uhai's list and tagged it U2 for maintenance. After that, Fastily deleted it. However, Bradv recreated it again and applied fully protection to it.

Only redirects not pages should be excluded from U2. Is User:Hobomok really such an exception? Should U2 be adapted to clarify this situation? NmWTfs85lXusaybq (talk) 10:48, 3 May 2023 (UTC)
 * To answer the question (in general) asked above, I would argue that no, the user page of a renamed user does not fall under U2 since it is explicitly excluded from it. This is an exceptional circumstance and should be dealt with as its own matter; multiple editors were genuinely operating in good faith to deal with some major issues.
 * Hobomok, as mentioned, had their talk page deleted before they vanished, which means that the user page probably should not have been restored (by anyone). If there were issues with trolls/vandals, then the page should/could have been salted.
 * The user talk page, on the other hand was moved with an out-of-process suppression of the creation of a redirect; it should have been moved during the rename and a redirect left in place. Again, if there was harassment or trolling the redirect (and its target) should have been protected.
 * Whether we decide to keep what is currently in place in place is not really a huge concern for me; my comments (transcluded below) were mainly to request that the automatic listing of redirects-from-renames be excluded from whatever dbase report was being used to find potential U2 candidates. The language of U2 is pretty clear, so no clarification necessary in that regard. Primefac (talk) 18:26, 4 May 2023 (UTC)
 * I get your point. Uhai's list is made to simply exclude all redirects from nonexistent user page, no matter the way of creation. You might ask to exclude even pages from renamed user, which needs another query of logging database and miscellaneous commands. User:Hobomok and User talk:Hobomok should have been protected as redirects, or deleted and protected from recreation, which will automatically remove them from Uhai's list.
 * I inspect this case again. JJMC89 have moved User talk:Hobomok to User talk:Renamed User iusdhf without leaving a redirect. From Special:Diff/1151434654, Michael Shellenberger (maybe LGSBB) made the harassment to compel Hobomok to reply to his email. When User:Hobomok and User talk:Hobomok were deleted, he might follow the link of the deletion log directly listed on the non-existent page. It's OK to create dummy pages to prevent this kind of tracking. NmWTfs85lXusaybq (talk) 02:28, 5 May 2023 (UTC)
 * I don't see deletion of the user page being a problem for a few reasons. First only redirects created due to a user being renamed are excluded from U2 (my emphasis). Since that page did not originate as a redirect it can be deleted per plain wording of the policy, in cases where redirects from a move are later converted to pages the most logical option would be to restore the redirect, with protection applied as needed. Secondly the page was U1 deleted prior to the renaming, presumably that user requests deletion still stands and should be respected. Finally, this is an exceptional circumstances case, and I see no issue with deleting the userpage for that reason alone nor was moving without leaving a redirect for the talk page an issue.
 * I'm not sure dummy pages to prevent tracking is helpful; anyone with a modicum of knowledge and even slight determination will be able to track anyway. If we are in a situation where the trail needs to be concealed, oversight (and potentially developers) will need to be involved and log entries will be among the easier things to deal with. 74.73.224.126 (talk) 13:34, 9 May 2023 (UTC)

Empty monthly maintenance categories
I was reviewing some of the links from the speedy deletion categories today and came across Category:Suspected copyright infringements without a source. There are several monthly daughter categories there which are empty, and are flagged by the category template for G6 speedy deletion, but that template doesn't populate the G6 category. Evidently since they're not properly flagged there are some that have been sitting empty for months. These should be eligible for WP:C1 speedy deletion, and I'd like to update the template so that similar empty categories will populate Category:Candidates for speedy deletion as empty categories, but this is one of those complicated templates that calls daughter templates and modules, and I couldn't figure out where to actually put the code.

I also noticed Category:Empty categories awaiting deletion, which seems like it's duplicating the function of the CSD category above, and the articles in it are also not populating the C1 category. They should, shouldn't they? Ivanvector (Talk/Edits) 18:25, 14 March 2023 (UTC) (never mind, answered below) Ivanvector (Talk/Edits) 15:12, 15 March 2023 (UTC)


 * Empty monthly maintenance categories usually appear under G6 rather than C1, which would make more sense (C1 would require the category to be empty for a week before deletion, G6 doesn't). Can they populate the G6 category instead?  Hut 8.5  18:31, 14 March 2023 (UTC)
 * Just as a minor note regarding Category:Empty categories awaiting deletion - this is the "holding cell" of sorts for the C1 nominations that are not yet a week old. Pages in this category should not be deleted as empty because they have not yet been nominated for the required duration. Primefac (talk) 19:01, 14 March 2023 (UTC)
 * Both good points. The Monthly clean-up category templates are already date-sensitive (the current month isn't flagged if it's empty) and since they're monthly there shouldn't be new members added to previous months. If both of those assumptions hold, then any monthly category that is empty and not the current month should be eligible for deletion. I'd be fine with that being G6 deletion instead of C1, although I think making a carve-out in C1 for this purpose would also be sensible. Ivanvector (Talk/Edits) 19:19, 14 March 2023 (UTC)
 * I thought that the bot both created and deleted these monthly categories. If that isn't actually the case, maybe it should be... Primefac (talk) 20:18, 14 March 2023 (UTC)
 * The bot creates them, but it's a handful of admins that delete them. has done a few today. Ivanvector (Talk/Edits) 20:34, 14 March 2023 (UTC)
 * The bot operator did think about writing a bot to delete the categories when no longer needed, but it turned out humans were doing it fast enough there was no need for a bot (see User:AnomieBOT III * Pppery * it has begun... 00:56, 15 March 2023 (UTC)
 * Moving them from the very overloaded G6 to an amended C1 or a specific new C3 would seem sensible for both clarity and as part of the gradual process of fixing the problems with G6. Thryduulf (talk) 00:52, 15 March 2023 (UTC)
 * Agreed. * Pppery * it has begun... 00:56, 15 March 2023 (UTC)
 * I think so too. I wasn't going to bring it up and I don't want to go off on a tangent, but G6 is used too much as a "no other criteria fit" category, when it's supposed to be for unusual cases of uncontroversial deletion, whereas if there's a novel and repeatable situation not covered by existing criteria, we should make new ones. Maybe we just need a C3 for old, emptied monthly maintenance categories, which is the same as C1 but without the 7-day hold. We would have C1 for emptied content categories, and C3 for maintenance categories where the maintenance is finished (other than those with Possibly empty category). Ivanvector (Talk/Edits) 15:22, 15 March 2023 (UTC)
 * See Categories for discussion/Log/2023 March 3 (result: keep). -- Red rose64 &#x1f339; (talk) 22:13, 14 March 2023 (UTC)


 * Sometimes these categories get repopulated when old deleted pages are undeleted. I just ignore the red category link in these circumstances. But should the category be undeleted too? Graeme Bartlett (talk) 11:40, 17 March 2023 (UTC)
 * That would return the article to their correct spot in the cleanup queue, which is probably the right thing to do to avoid articles having a cleanup tag for ten years, being deleted and undeleted, having a new cleanup tag applied and then wait another ten years for the cleanup to actually happen. Or even longer. (We have some super-duper embarrassing things like Category:BLP articles lacking sources from January 2007). —Kusma (talk) 12:08, 17 March 2023 (UTC)

Proposal: C4 Empty maintenance categories
C4: Empty maintenance categories
 * This criterion applies to categories in (a subcategory of) Category:Maintenance categories that are empty and which have no further use, for example dated clean-up categories for dates in the past. This excludes categories that are only temporarily empty, including those tagged with Possibly empty category, categories under discussion at Categories for discussion (or other such discussions) and category redirects.

This is based on the discussion above, and may benefit from wordsmithing (it is numbered C4 as there was previously a C3). If this passes, the first bullet of G6 "Deleting empty dated maintenance categories for dates in the past" would be removed to avoid redundancy. The change should be widely publicised, e.g. in the Administrators' newsletter. Thryduulf (talk) 15:50, 20 March 2023 (UTC)
 * I'm having a hard time seeing what problem this is supposed to solve. All it seems to do is shift these deletions from one category to another, and while I agree that G6 is overused, the proper solution is to talk with admins who are using it incorrectly (and go to DRV if necessary), not to make a new criterion for something that undeniably is "uncontroversial maintenance". Extraordinary Writ (talk) 20:04, 20 March 2023 (UTC)
 * Just to be clear, I oppose per my comment above. If we can think of an alternative to the "uncontroversial maintenance" standard, then we should propose that instead, but as long as that standard remains, C4 would just be a redundant criterion (cf. NEWCSD #4) that doesn't get us any closer to solving the problems with G6. Given the numerical split (Thryduulf, Pppery, Ivanvector, Primefac, and Kusma leaning support, Fastily, Jclemens, Graeme Bartlett, and myself leaning oppose, and everyone else somewhere in between) and the reasonable arguments on both sides, I don't see a consensus at the moment; an RfC would be the next step if anyone really wanted to go forward with this. Extraordinary Writ (talk) 20:26, 14 April 2023 (UTC)
 * Support I would word it even more narrowly as applying only to subcategories of Category:Wikipedia maintenance categories sorted by month and Category:Wikipedia categories sorted by month, but otherwise see no problem with this. * Pppery * it has begun... 22:30, 20 March 2023 (UTC)

Pinging the editors who commented above:. Thryduulf (talk) 00:53, 22 March 2023 (UTC)
 * fixing the ping. Thryduulf (talk) 00:54, 22 March 2023 (UTC)


 * Why don't we just leave them in G6? Creating new SD cats just makes more work for hundreds or thousands of people to learn. Deleting these really is uncontroversial, so they are a good fit for G6. Graeme Bartlett (talk) 03:16, 22 March 2023 (UTC)
 * "Hundreds or thousands of people" seems exaggerated. Surely nobody other than the small handful of admins clearing this deletion queue needs to learn anything new (the tagging is done by template magic). * Pppery * it has begun... 03:28, 22 March 2023 (UTC)
 * Perhaps, but compassing a new SD category is a non-trivial task that does affect people needing to understand it, update (or not update) tools, etc. Thus the question: is the hassle worth the improvement? I'm not seeing a clear "yes" answer either. Jclemens (talk) 05:23, 22 March 2023 (UTC)
 * The categories concerned would normally be bot-tagged, and would rarely (if ever) need to be human-tagged, so the number of people who need to know of the new criterion would be those admins who actually perform deletion (a few hundred, maximum) plus one botop. Only one tool would need to be altered - the bot that scans for qualifying categories. As I see it, part of the rationale for this is to reduce the number of legitimate G6 deletions, so that it is easier to spot G6 misuse. After all, the proposed new speedy deletion criterion does meet Objective, Uncontestable and Frequent, it's just Nonredundant that is the issue - because it's presently bundled with several otherwise-unrelated uses for G6. -- Red rose64 &#x1f339; (talk) 08:37, 22 March 2023 (UTC)
 * because it's presently bundled with several otherwise-unrelated uses for G6 - and yet, we've had a half-dozen discussions in the last year about making G6 easier to understand, trying to get it even more narrow, and eliminating some of the feeling that "well this needs deleting but nothing else fits so G6 it is!" that seems to pervade the criteria. I said it somewhere else, but if the bot isn't deleting the maintenance categories it creates itself, it should. Barring that, the empty categories should be tagged for C1 at minimum. In other words, I'm not sure whether we need a new C4 category, since by definition these empty categories already meet C1. Primefac (talk) 09:34, 22 March 2023 (UTC) whoops. see below. Primefac (talk) 10:01, 22 March 2023 (UTC)
 * One motivation is indeed to reduce the misuse of G6 (G6 at present does not meet the objective and uncontestable points of NEWCSD) by simplifying it and moving points out that clearly do meet NEWCSD.
 * @Primefac C1 explicitly excludes maintenance categories so it would need amending if we want to delete the categories using that rather than G6. We could do that, but I think a new C4 would be both clearer and simpler and I've long been arguing that a long list of short and simple criteria is better than a short list of long, complicated ones. Thryduulf (talk) 09:58, 22 March 2023 (UTC)
 * Sometimes I amaze myself with completely forgetting rules and regs that I've been dealing with for almost a decade... you are correct, so I would support C4 for this purpose. Primefac (talk) 10:00, 22 March 2023 (UTC)
 * We should reduce the number of things covered by G6. Merging into C1 could also work (with or without the one-week wait), but a separate C4 is probably cleanest. —Kusma (talk) 10:10, 22 March 2023 (UTC)
 * Support - thanks for the ping, I made the thread above and then kind of forgot about it, it's a mental Swiss cheese time of year in my world. Monthly maintenance categories are meant to be temporary, even though some may stick around for years we expect that they will eventually be empty and not repopulated. It's also desireable to remove the empty categories, otherwise they clutter the cleanup category views and cause editors to waste time looking for a category with work left to do, rather than only seeing active categories. Thus "empty monthly maintenance categories" easily meet the first three header criteria for new criteria (that's a mouthful). For the fourth, "nonredundant", this is not redundant to C1 because there is no reason for a seven-day hold for this narrow application, and I agree with Thryduulf that many narrowly-defined criteria are preferable to expanding the already-too-broad definitions of G6, and creating a carve-out for this in C1 would be unnecessarily complicated (the modified C1 would apply in different ways to different pages, making it much less objective). Ivanvector (Talk/Edits) 12:44, 22 March 2023 (UTC)
 * As for technical issues, I expected that tagging would be handled by code in Monthly clean-up category. The template already flags the page for G6 deletion if it is empty and not the current month, the proposal really is just to change that to C4. C4 could then be monitored by a bot and members automatically deleted, instead of piling them in with all the other unrelated deletion reasons covered by G6 which really require human review. That would reduce the workload on admins monitoring G6, possibly significantly. Ivanvector (Talk/Edits) 13:05, 22 March 2023 (UTC)
 * No objection to this but I don't see much advantage over having it as part of G6, it doesn't really make the policy simpler and these are uncontroversial maintenance.  Hut 8.5  12:44, 23 March 2023 (UTC)
 * I completely agree that we need to narrow the scope of G6, but I actually think the C1 carve-out is the way to accomplish this. We already have a similar carve-outs in F5 and F7. In particular, F7b/c/d are not given their own criterion because they all stem from the same "issue": an invalid fair-use claim. For C1 and the proposed C4, the "issue" would be the same: an empty category. I would also be fine with not making a carve-out and waiting a week before they are eligible for deletion: they do not need to be deleted urgently. HouseBlastertalk 01:45, 27 March 2023 (UTC)
 * Maintenance categories are currently explicitly excluded from C1, we would need to remove or alter that if we included these in that criterion, which would either make it even messier than with just the one carve out or require a complete rewrite into C1a (content and structural categories, 1 week wait) and C1b (maintenance categories, no delay). I think C1 and C4 are much simpler than C1a and C1b (cf. A7/A9, F4/F6/F7; F1/F8). Thryduulf (talk) 00:22, 28 March 2023 (UTC)
 * Fair enough. I support the creation of C4. I could also live with narrowing the exception to say something like "non-dated maintenance categories are excluded under this criterion". I would oppose a "you can delete these immediately" clause as making a rule more complicated for negligible benefit (see also WP:CREEP and WP:NODEADLINE). HouseBlastertalk 17:00, 15 April 2023 (UTC)
 * Oppose I agree in principle, but the sheer amount of effort required to implement this proposal (e.g. retraining editors, updating bots/templates/scripts/docs) simply outweighs the value we'd get out of it.  Also worth noting that maintenance category deletions are low frequency relative to other types of deletions.  -  F ASTILY   01:20, 28 March 2023 (UTC)
 * Also worth noting that maintenance category deletions are low frequency relative to other types of deletions - The last time I did a G6 taxonomy they were the most common individual type of deletion. * Pppery * it has begun... 01:26, 28 March 2023 (UTC)
 * Is that so? In comparison to other G6's or all CSD? -  F ASTILY   01:30, 28 March 2023 (UTC)
 * It was in comparison to other G6s. See Wikipedia talk:Criteria for speedy deletion/Archive 83. * Pppery * it has begun... 01:43, 28 March 2023 (UTC)
 * Ah I see. My original statement above was made in relation to all CSDs & XfDs.  -  F ASTILY   02:06, 28 March 2023 (UTC)
 * Even compared to all deletions, it's about 1%. Given that there are 40 CSD criteria, that's not as low as you would think, and certainly more frequent than many other standalone criteria. * Pppery * it has begun... 02:22, 28 March 2023 (UTC)
 * Interesting, I had no idea there were so many. If anything, that's a stronger argument to maintain the status quo.  -  F ASTILY   03:03, 29 March 2023 (UTC)
 * Based on other comments in this thread the number of bots that would need to be updated is 1, the number of templates that would need updating is approximately 0 (afaict at a quick look none of the templates in Category:Speedy deletion templates make specific reference to this line of the G6 criteria), the number of templates that would need creating is approximately 1 ( or ). Scripts requiring updating, approximately 1 (Twinkle, but as I don't use it I can't be sure). Editors have coped without any issue since R4 and G14 were spun out of G6. So the amount of effort required to make this change is much less than you seem to have guessed, and the value greater than you seem to have guessed. Thryduulf (talk) 16:19, 28 March 2023 (UTC)
 * Template:Monthly clean-up category/core would need to be updated, I think, and there may be other templates, too. - Eureka Lott 19:28, 28 March 2023 (UTC)
 * It looks like updating that template would take a matter of seconds, changing "G6" to "C4". That's hardly a significant burden. Thryduulf (talk) 19:51, 28 March 2023 (UTC)

I know I'm involved, but I think I'm seeing a slight consensus in favour of this proposal, at least when restricted to subcategories of Category:Wikipedia maintenance categories sorted by month and Category:Wikipedia categories sorted by month. Am I right in this? Are there any other comments? Thryduulf (talk) 11:20, 12 April 2023 (UTC)
 * I believe you are correct in your assessment. There is insufficient opposition to stop this being enacted, even if you (or I) have been involved in the discussion. Primefac (talk) 09:25, 14 April 2023 (UTC)
 * Any further comments? If not I will enact this in a few days (if I can figure out the template work). Thryduulf (talk) 21:52, 7 May 2023 (UTC)
 * You might want to see the comment at 20:26, 14 April 2023 (UTC) by Extraordinary Writ above (second thread, last comment). Apparently there is not as clear a consensus as you and I thought? Primefac (talk) 11:27, 8 May 2023 (UTC)

Creator many not remove tag and redirects created from moves
I think we should clarify that if a redirect was created as a result of a page move that the user who preformed the page move can remove the tag if they disagree unless they were the author of the page or they moved the page to the location where it was moved back from similar to what redirects can be deleted as G7. The reason being that it doesn't make sense to not allow this is firstly it doesn't make sense that a user who moves a page away from an unsuitable title should be disallowed from removing the tag if they think its suitable enough to exist as a redirect, secondly most users who move pages away from unsuitable titles are experienced enough to decide if CSD likely applies anyway.


 * Example 1, User:A moves Foo589 to Foo588 (which was previously a red link), User:B disagrees with the move as they think Foo589 is still more common and reverts the move, User:C thinks Foo588 is an implausible title and tags it with R3. User:B disagreed with the R3 as although they don't think Foo588 should be the title they still think the redirect is plausible though so want to remove the tag.
 * Example 2, User:A moves Foo589 to Foo588 (which was previously a red link), User:A shortly after reverts their move and moves Foo588 back to Foo589, User:B comes along and tags Foo588 with R3 and User:A disagrees and thinks Foo588 should be kept as a redirect and wants to remove the tag.
 * Example 3, User:A moves Foo589 to Foo588 (which was previously a redirect created by User:D), User:B disagrees with the move as they think Foo589 is still more common and reverts the move, User:C thinks Foo588 is an implausible title and tags it with R3. User:B disagreed with the R3 as although they don't think Foo588 should be the title they still think the redirect is plausible though so want to remove the tag.

In the 1st example it would make no sense for User:B to be prohibited from removing the tag as they moved the page away from an apparently unsuitable title though I don't see why User:A couldn't be. In the 2nd example it does make sense for User:A to be prohibited as they had effectively created the Foo588 title as a redirect by their move and revert. In the 3rd example it wouldn't make sense for either User:A or User:B to be prohibited from removing the tag as the redirect previously existed created by a 3rd user.

So without complicating things too much I propose incorporating this.  Crouch, Swale  ( talk ) 18:03, 21 May 2023 (UTC)
 * This seems unlikely to be a common enough case to warrant mentioning here. * Pppery * it has begun... 21:42, 21 May 2023 (UTC)
 * I agree with Pppery. Unless this is happening frequently, and we'd need to evidence of this (and also a much clearer wording), it seems like a good use of IAR. Thryduulf (talk) 12:03, 25 May 2023 (UTC)

Fully protect this page!
Fully protect this page: Criteria for speedy deletion, so only any admin can edit! Also add the gold lock, like this: MihaiEdit (talk) 10:39, 2 June 2023 (UTC)


 * Why? Phil Bridger (talk) 11:04, 2 June 2023 (UTC)
 * Please examine WP:Protection policy, and identify a good reason why the page should be protected. Then file a request at WP:Requests for page protection, stating your reasons. -- Red rose64 &#x1f339; (talk) 11:19, 2 June 2023 (UTC)
 * To prevent vandalism and joke edits! MihaiEdit (talk) 11:35, 2 June 2023 (UTC)
 * As per the protection policy, protection is not intended as a preliminary measure against vandalism/disruption that has not occurred. To consider full protection, there would have to be a history of disruption by established editors or reason to believe that suboptimal edits could cause sitewide damage, both of which I think are highly unlikely in this case. Complex / Rational  13:57, 2 June 2023 (UTC)

How early is the earliest block or ban in G5?
has managed to test the G5 wording to destruction, as has appeared at Deletion Review.

Background: WhichUserAmI created that account. Then, a few months later, they created another account with a stupidly long name. This account was blocked indefinitely for having a stupidly long name. They then created some articles. Subsequently, they were found socking, and all their accounts were blocked as socks, including a re-block of the still-blocked long-name account. then deleted a lot of articles they created, dating back to before the sock-block, but after the long-name-block. WhichUserAmI is now requesting at Deletion Review that these articles should be reinstated as the application of G5 was retrospective of their block for socking. has pointed out that Liz's deletions are after the first block, and therefore still comply with G5.

G5's problematic wording "* When a blocked or banned person uses an alternate account (sockpuppet) to avoid a restriction, any pages created via the sock account after the earliest block or ban of any of that person's accounts qualify for G5 (if not substantially edited by others); this is the most common case for applying G5."

Question 1: Do we need to clarify that G5 applies only to articles that the editor was, by virtue of a ban, unentitled to make? For example, if someone got blocked or banned, subsequently the block or ban was lifted, and then they got re-blocked and did some socking, clearly the editing they did between the end of their first ban, and the point they started socking, was entirely legal and above-board, and shouldn't be retrospectively deleted. Another example, if WhichUserAmI had first created an account with a stupidly long name, and this had been indefinitely blocked as an unhelpful name, and they responded by abandoning that account and creating one with a more sensible name, then again it's not helpful to treat the first block on the stupidly-named account as the point at which they became a "bad editor" whose work should be deleted. There was, in fact, nothing else they could have done, to put right the error they made in choosing a daft user-name.

Question 2: Do we want G5 to go back to the point where socking began, or only to the point where the sock-block was made? This is where WhichUserAmI's situation is very unclear, because they created the long name account after their main account, so it was already a sock when it got blocked for being a silly name. The articles dating between the creation of the long name account and the sock-block were therefore created during the period that the socking had been going on, even though they weren't created in evasion of any particular ban.

Obviously some common sense is needed in situations like this, but I wonder if a change in the wording might avoid misunderstanding. I suggest:

"* When a blocked or banned person uses an alternate account (sockpuppet) to avoid a restriction, any pages created in contravention of their block or ban, by any of their accounts, qualify for G5 (if not substantially edited by others); this is the most common case for applying G5." Elemimele (talk) 08:11, 18 May 2023 (UTC)
 * The stupidly-long name account was hard-blocked rather than soft-blocked, so it controls (and should control) in all circumstances, even in the hypothetical in question 1, and the articles created after it was blocked were validly G5ed. That said, it's generally agreed that once someone is unblocked they are entitled to request restoration of any articles G5ed due to them, so the DRV itself is fine. * Pppery * it has begun...  13:15, 18 May 2023 (UTC)
 * I don't want to get side-tracked by that particular example, which is an odd one (a NOTHERE block for vandalism would have made more sense). My main concern is that the scope of G5 should be something that's obvious. When it reaches a situation that we are determining the deletability of an article by a detailed examination of the exact dates and technical natures of blocks buried deep in the block-logs of a set of sock accounts, then for the sake of the encyclopaedia it would be better to avoid G5 and send the articles to AfD, where they can be assessed on their merits rather than the merits of their creator. Elemimele (talk) 16:44, 18 May 2023 (UTC)
 * Sending stuff to AFD would undermine the point of G5. We don't want to decide on the merit because blocking/banning would be pointless if we kept stuff created in violation of a block/ban. In fact, it would just invite more socking. Regards So  Why  19:38, 18 May 2023 (UTC)
 * Speedy deletion criteria explicitly only apply "in the most obvious cases", and this is very clearly not one of the most obvious cases (it's arguably one of the least obvious). A good standard for this is that if there is good faith disagreement about whether a page meets a speedy deletion criterion then it does not. Both of these mean that G5 does not apply here. Thryduulf (talk) 18:24, 20 May 2023 (UTC)


 * So, I'm confused... Are we saying we can't give WhichUserAmI a WP:REFUND because their work was created while socking? If they have been reinstated, I think there's no problem with restoring the work.  I don't think we need to change any wording to G5 to indicate that; on the first hand WUAI's block-and-reinstatement-and-REFUND-request is unusual to the point that I've never seen such a thing before, so I can't see there to be a pressing need to account for such sui generis situations by rewriting policy.  WP:IAR covers weird situations like this; it doesn't benefit the encyclopedia to keep these pages deleted, so let them have the pages back to work on.  That's sufficient for me.  -- Jayron 32</b> 17:12, 18 May 2023 (UTC)
 * I think this is already clarified in G5 that it only applies if the user was blocked (under any account, not counting soft username blocks) at the time the page was created regardless of their behaviour or if they get blocked/banned later, the 1st and 3rd bullet points clarify that. G5 can never be applied if the user wasn't blocked/banned at the time of the creation. See Alamgir Khan (singer) for an example.  Crouch, Swale  ( talk ) 17:17, 18 May 2023 (UTC)
 * Let's deal with this from a "what's best for the encyclopedia" standpoint rather than by wikilawyering. I neither know nor care whether this person had been blocked at the time these articles were created. Phil Bridger (talk) 19:37, 18 May 2023 (UTC)
 * That's what I was getting at: forget the details of the nature and timing of different blocks, just use G5 common-sense to zap stuff that shouldn't exist. But I'm not going to make a thing about this, I think I got an unnecessary bee in my bonnet and I'm quite happy to drop this rather unattractive stick and do something more useful instead. As Jayron32 said, IAR allows an admin (or anyone) to use common sense. Elemimele (talk) 21:15, 18 May 2023 (UTC)
 * Yes, that exactly. G5 allows such things to be deleted, it does not mandate that they must be deleted.  Admins are quite allowed to decline a G5 for as simple a reason as "Look, I get that this was technically a violation, but honestly, it looks like the encyclopedia is better if we just leave it".  -- Jayron <b style="color:#090">32</b> 11:01, 19 May 2023 (UTC)
 * This discussion makes it look complicated but it is not. G5 was appropriately applied, and DRV did not overturn the speedy deletion. This user was blocked and after being blocked they created articles. When a user operates multiple accounts and one of those accounts is hard-blocked, the user is a blocked user, and as such is not allowed to edit per WP:ILLEGIT. The user needed to request unblock, not to create articles. Nothing needs to be changed or clarified in the policy. —Alalch E. 16:36, 3 June 2023 (UTC)

Clarify R3 regarding misspellings of "(disambiguation)"
I think maybe it would be worth clarifying that R3 applies to cases where the term "(disambiguation)" is misspelled/mistyped. This follows a discussion (but not a consensus on this matter) at Redirects for discussion/Log/2023 May 20. ISTM that cases like "Term (diambiguation)", "Term (disabiguation)", and "Term (disambiguation" are very different from, say, "Tirm (disambiguation)" (where "Tirm" is a misspelling of "Term"). The worth of the last case (I'm assuming these are all redirects) depends on the "usefulness" of the misspelling "Tirm", while whatever logic you apply to the first three cases would apply equally to literally any "Term". This would seem to go against the "spirit" of R3, but because it is not explicity covered by that criterion (or G14, or G6 for that matter), people are relying on WP:RDAB, which is not official policy. It's also a little confusing, assuming you have this case in mind, having a caveat in there about R3 not applying to redirects ending with [a properly spelled] "(disambiguation)". Therefore, I propose adding to R3 wording that clarifies that the criterion does apply to cases where the typo is in the term "(disambiguation)" itself. Not sure the best wording to use. Note, BTW, that a search for, say, "Term (diambiguation)" or even "Term (diabiguation)" [two typos] will turn up the page "Term (disambiguation)", if it exists, so such redirects are completely unnecessary. - dcljr (talk) 07:19, 25 May 2023 (UTC)


 * No. That discussion has not closed yet, and so it's completely inappropriate to use it as the basis for a policy change, but even if it had been it and previous discussions makes it clear that deleting misspellings of "disambiguation" (and WP:RDAB as a whole, which is why it's not policy) is controversial and thus fails WP:NEWCSD point 2. Thryduulf (talk) 12:00, 25 May 2023 (UTC)
 * I'm not asking for it to be used as a "basis" for anything. It simply brought the issue to my attention, so I acknowledged that in the interest of "full disclosure". Sorry if I was unclear about that. As for your other point, a user in the other discussion noted the following cases where several such redirects were deleted: "1, 2 and 3". These three discussions from 2020 contain but a single "Keep" !vote that turned out to be based on a misunderstanding and so was struck out and replaced by a "Delete" !vote. I have found this discsussion from 2022 which contains multiple "Keep"s but ended up with a consensus to delete. Can you (or anyone) point to examples where such redirects have been kept? - dcljr (talk) 18:22, 25 May 2023 (UTC)
 * I'd support creating a new R5 criteria for RDAB errors. Yes I know there has previously been proposals on this but from the August 2022 discussion I think the consensus is clearer on this now. Many could arguably be deleted as R3, G14 and G6 but a separate criteria would probably be more helpful to deal with the (not infrequent) times R3 wouldn't apply at least for now.  Crouch, Swale  ( talk ) 20:17, 25 May 2023 (UTC)
 * @Dcljr Speedy deletion is only for things that are uncontroversial - i.e. everyone agrees that everything that could be deleted should be deleted, every time. Where deletion discussions gain good faith recommendations for a course of action other than deletion, as here, that criterion cannot be satisfied regardless of the eventual consensus of the discussions. Unless and until all proposals to delete these are unanimously deleted every time then speedy deletion is completely inappropriate. Thryduulf (talk) 07:19, 26 May 2023 (UTC)
 * I think that's a ridiculously high bar. "Uncontroversial" does not necessarily mean "unanimous". Everything in these CSD guildelines were put there, or at least are being kept there, by consensus, and you can't tell me that every single one of them was unanimously agreed upon. Again, I ask: can anyone reading this point to a nomination of this kind that resulted in a "keep"? - dcljr (talk) 10:56, 27 May 2023 (UTC)
 * The standard isn't "consensus to keep" it's uncontroversial, the most obvious cases and everything that could be deleted should be deleted. Discussions that regularly get multiple good faith recommendations for something other than "delete" are not uncontroversial. Yes, this is a very high standard, but that's what CSD is - it's the exception not the rule. Thryduulf (talk) 19:36, 27 May 2023 (UTC)
 * The criteria on this page are there by consensus. And in particular, the wording of those criteria are in their present state by consensus. I am merely asking that we consider an additional clarifying change in wording, not a whole new criterion. Anyway, I'll wait to see if anyone else weighs in here. - dcljr (talk) 02:19, 29 May 2023 (UTC)
 * The requirements for changing existing criteria are identical for the criteria for new criteria. Thryduulf (talk) 09:00, 29 May 2023 (UTC)
 * R3 is an unusually subjective criterion and it would make sense to address that. I would support adding a footnote after the word "implausible" along the lines of, "An 'implausible' typo or misnomer is one that is unlikely to ever benefit readers. If a redirect has a realistic chance of being kept at Redirects for discussion (RfD), it is not implausible for the purposes of this criterion. Typos and misnomers often deleted at RfD include those containing multiple typos, very unlikely transpositional errors or mishearings of words, and misspelled disambiguators, although all of these may be valid redirects in some contexts. See also ." --  Tamzin  [ cetacean needed ] (she&#124;they&#124;xe) 02:33, 29 May 2023 (UTC)
 * I think most RDAB redirects that were recently created could already be deleted under R3 such as if London (Disambiguation) was created as there is a consensus (from the August 2022 discussion) to delete these but I still think a separate criteria would likely be helpful both for those not recently created as well as avoiding ambiguity.  Crouch, Swale  ( talk ) 16:42, 3 June 2023 (UTC)

Questions re: specific G6 and copy/paste draft scenario
I'd like some feedback regarding a specific scenario involving G6 and drafts:
 * Kaumudi Nemani was previously draftified at Draft:Kaumudi Nemani
 * Blalithk, the article originator, performed a copy/paste move of the draft back to Kaumudi Nemani; but only after a not-insignificant number of contributions from several other editors
 * Mori Calliope fan tagged the copy/paste as G6
 * I took this as a good faith maintenance request and deleted the copy/paste article
 * Mori Calliope fan did not move the draft to mainspace, but did leave a note at User talk:Blalithk explaining the problem with unattributed copy/paste moves within Wikipedia; and in doing so, explained to Blalithk that they can move the draft themselves using the move function.
 * If it matters, it looks like Blalithk has never communicated w/ other editors via any means, be it talk page or edit summary.
 * Pppery subsequently left a complaint on my talk page requesting that I restore the deleted copy/paste because G6 does not apply to copy/pasted drafts. This is apparently something Pppery is doing with all draft-related G6 deletions based on a log entry alone.
 * I realize that G6 states, "An administrator who deletes a page that is blocking a move should ensure that the move is completed after deleting it." Whether this is a should or a must, I concede that follow-up may be a good idea.
 * Questions: What should happen in this scenario?
 * Do I, as the admin who deleted the copy/paste per the G6 tag in good faith, automatically endorse the move from draft to mainspace?
 * In other words, am I obligated to move it rather than assume Mori Calliope fan would?
 * For the record, I don't think the draft is ready for mainspace and could use better sourcing. I believe the original draftification is a good alternative to straight deletion/AFD.
 * The possible outcomes, that I can see:
 * Remind Mori Calliope fan to complete the move, since Blalithk did not perform it themselves
 * Accept an obligation on my own part when performing a good faith maintenance deletion on behalf of Mori Calliope fan's request, in which case:
 * I would need to publish to mainspace whether I agree that it is ready or not, and perhaps PROD or AFD it if I felt strongly enough about it
 * Or, ignore the G6 maintenance request entirely because I don't agree that the article is ready, and leave the copy/paste in place for someone else to deal with.
 * Note that this isn't really a histmerge situation because the deleted copy/paste was just that: one copy/paste edit, and one edit for the G6 tag. Nothing substantive to merge w/ the draft history.

Any clarity on the matter would be greatly appreciated! --Hadal (talk) 13:33, 6 May 2023 (UTC)
 * As someone who does a lot of these, you have three general options:
 * Perform a histmerge.
 * Move the original page to the new location.
 * Decline the CSD and perform no histmerge
 * Note that these options are valid for Draft → Article but also Article → Article. In this particular circumstance where the only article-space edit was the original creator doing the copy/paste, I would do the second and just move the draft over the article. Much of the time (I would say ~75% for me) I end up doing a histmerge. The rest of the time I am doing #3, largely because there is no content editing by anyone other than the original editor and thus there are no attribution issues. Primefac (talk) 13:40, 6 May 2023 (UTC)
 * @Primefac Wow, that was a quick reply! So you would agree that rather than delete via G6 and assume the nominator will perform the move, that I should (must?) perform the move myself? Does the language of G6 need to be amended to be more explicit, i.e. "must ensure" rather than "should ensure"? --Hadal (talk) 13:48, 6 May 2023 (UTC)
 * The page was tagged (and deleted) three days ago; clearly there is no inclination of the nominator to move the page. I will note that db-move has a (link to perform this move) pipe to Special:Move to make the move that much easier; db-copypaste should probably also have that...
 * Regarding your query specifically, I do occasionally delete without moving but I will then leave a note with the nominator along the lines of "page deleted, move away". Primefac (talk) 13:56, 6 May 2023 (UTC)
 * @Primefac Right, so I can either move the page myself or remind the nominator? I am under no obligation to move it myself? --Hadal (talk) 14:00, 6 May 2023 (UTC)
 * From my perspective, yes. Primefac (talk) 14:07, 6 May 2023 (UTC)
 * @Primefac Thanks for your thoughtful feedback! It may seem like a finer point, but it makes all the difference in this scenario. I'll leave a note on the nominator's page. --Hadal (talk) 14:12, 6 May 2023 (UTC)
 * It appears most of this has been resolved without the need for further comments from me, but I did this because the situation here looked exactly like what was just rejected at . And yes, I plan to keep challenging those deletions whenever they show up at Database reports/Possibly out-of-process deletions, because that's what the community has asked me to do in that section. * Pppery * <sub style="color:#800000">it has begun... 15:02, 6 May 2023 (UTC)
 * I don't think "the community" has asked you to make drive-by complaints based on bot-generated logs that can easily be misinterpreted, and should therefore demand more care than you have shown. Context is king. If you don't have access to the deleted edits to actually know what's going on, ask for clarification first. Please don't go straight to assuming impropriety, and don't link to a withdrawn proposal as some kind of blanket justification for attempting to read other peoples' minds. --Hadal (talk) 15:38, 6 May 2023 (UTC)
 * ... but my interpretation of what happened appears to have been perfectly correct here? * Pppery * <sub style="color:#800000">it has begun... 15:46, 6 May 2023 (UTC)
 * If you believe the copy/paste article should be "restored" you should either ask Mori Calliope fan to move the draft after the maintenance deletion occurred, or move the draft into mainspace yourself. But instead you went straight to requesting the (technically copyvio) copy/paste version be restored, which does not appear to make any sense. --Hadal (talk) 15:54, 6 May 2023 (UTC)
 * It makes perfect sense, because I'm enforcing deletion policy, not making a judgement about the suitability of the article for mainspace. * Pppery * <sub style="color:#800000">it has begun... 15:57, 6 May 2023 (UTC)
 * Oh, I didn't realize you were an enforcer. Please elaborate. If you're requesting an article be restored, and given that it is a copyvio, what possible justification do you have? --Hadal (talk) 16:02, 6 May 2023 (UTC)
 * Because it was speedy deleted in a scenario in which there was explicit consensus several sections above not to speedy delete such pages. Oh, and it's not a copyvio, because nobody other than Blalithk has made significant edits to the draftspace version. We're clearly talking past each other here. * Pppery * <sub style="color:#800000">it has begun... 16:09, 6 May 2023 (UTC)
 * You have repeatedly ignored a perfectly reasonable option: If you want the draft in mainspace, move it yourself. The G6 accomplished two things: It removed a copy/paste move that violated policy; and, as the G6 indicated, made way for the clean draft to be moved to mainspace. The tools are in your hands. Fix it yourself. --Hadal (talk) 16:16, 6 May 2023 (UTC)
 * It is rarely necessary to pre-delete (justified by G6 or otherwise) a redir that is blocking a page move, or even to use the tag. Follow the directions at WP:RM, and the admin who takes it will simply use the normal "move" option, when they will be presented with a warning box reading 'The destination page "Foobar" already exists. Do you want to delete it to make way for the move? (Check the edit history.)' and one extra checkbox: "Yes, delete the page" which they enable before clicking . -- Red rose64 &#x1f339; (talk) 22:57, 6 May 2023 (UTC)
 * Thanks for your feedback but, nobody was talking about redirs here. I don't see the relevance of your statement. Yet, I think you've raised even more interesting questions. Forgive me, I know this has already been a thrilling journey through language and meaning, but:
 * You appear to be saying that non-admins cannot use G6 - delete to make way for a page move/fix move - in CSD nominations, and therefore admins must decline these nominations. They must use RMT.
 * Or, a non-admin can sometimes (?? when do you think this is OK?) use db-move; however, the admin is then obligated to perform the move rather than simply delete it in good faith.
 * Perhaps G6 needs a rewrite if any of the above are becoming common interpretations, as you are suggesting here. --Hadal (talk) 02:22, 7 May 2023 (UTC)
 * It's not necessarily a redirect, although that is the most frequent case. It could be any page that is blocking a move. The opening post of this thread includes I realize that G6 states, "An administrator who deletes a page that is blocking a move should ensure that the move is completed after deleting it." I am pointing out that the admin shouldn't need to delete the blocking page and hope that the move is performed: whether it's a redirect or not, an admin can move another page to take its place merely by passing through an extra step of the page move process, there is no need to delete the blocking page first, and therefore no need to tag it either. -- Red rose64 &#x1f339; (talk) 18:33, 7 May 2023 (UTC)
 * Your logic rests on an explicit assumption that is not borne out in policy: That an admin is obligated to perform a G6-related move, be it via db-move or not.
 * This is not the case, is it? Yet the only logical conclusion to be read from your statements is that you believe an admin must perform the move, which is an entirely separate act from the target deletion. You are also implying by extension that an admin is effectively endorsing the move related to the G6 deletion. Unless you believe that moving and deleting are the same act? In either case, can you support your position with a relevant policy? -Hadal (talk) 19:36, 7 May 2023 (UTC)
 * You are being very argumentative, and reading an awful lot into my comments that I simply did not say. No admin is obliged to do anything. If an admin disagrees with a G6 tagging, they can (a) de-tag the pages concerned; (b) leave a comment on the associated talk page explaining their concerns; (c) ignore it and wait for another admin to pass by. Similarly, if an admin disagrees with a RM#T request, they can (a) decline the request; (b) use the "discuss" link to turn it into a forml RM and say why they disagree; (c) ignore it. -- Red rose64 &#x1f339; (talk) 21:11, 7 May 2023 (UTC)
 * I apologize if you're reading any tone into this that I did not intend, because I was sincerely trying to make sense of your comments. I work with a lot of legal documents, and sometimes it's hard for me to turn off the technical filter in my brain that sounds an alarm whenever something doesn't make sense to me. This is my fault, and I'll make an effort to be more mindful of it.
 * Let's try this again, after stepping back for a minute: You said that it is "rarely necessary to pre-delete... or even use db-move" because these requests are handled at RMT. You said that "there is no need to... tag" an article, because an admin will be handling the move, and the deletion will occur automatically. But as far as I'm aware, there's nothing wrong with non-admins using G6/db-move tags to have target articles deleted, as good faith maintenance requests. Likewise, there's nothing wrong with an admin deciding to delete the tagged target article as a good faith maintenance request -- e.g., there's nothing in the wording of the db-move tag that says the move must be performed by the admin. There's a link for convenience, but this is not an imperative. I'm evidently not the only one who doesn't always perform the associated move after deleting the target per G6 request; perhaps opting to remind the nominator of the clear target instead.
 * If you agree with this, which it now sounds like you might, then everything is in fact working as intended. But if not, we as a community need to have a larger discussion around G6, as there's a vibe around here from a growing number of people that are unhappy with the way G6 is being used or interpreted. Perhaps all such requests should be handled via RMT, which would remove a lot of ambiguity. We'll see. Your feedback is extremely valuable in sussing this out, and I thank you for sharing it. --Hadal (talk) 00:03, 8 May 2023 (UTC)
 * "An administrator who deletes a page that is blocking a move should ensure that the move is completed after deleting it" was my addition. I requested an admin delete a redirect that was blocking a page move. They deleted the redirect but didn't move the page. I didn't get back online for more than a day, so that entire time, that article title was a dead link. That's why the admin who deletes needs to finish the job.
 * If you have any uncertainty about moving the page, don't delete, but bring to a wider discussion, such as at Requested moves Oiyarbepsy (talk) 04:49, 9 May 2023 (UTC)
 * This is another situation caught by my out-of-process deletions report, and it found 6 cases in just the last 3 days of an admin deleting a redirect to make way for a move, and then the move not being finished (all of which I've just done now). * Pppery * <sub style="color:#800000">it has begun... 04:54, 9 May 2023 (UTC)
 * Pardon me Oiyarbepsy, but are you saying that you added that language ("Admins should ensure...") to G6? If so, was it added as a result of any discussion or consensus-building? (I'm asking for my own reference.) I realize policy can be edited by anyone, but if this wasn't the result of consensus I think that the language should be removed. I may do this myself, because I think its inclusion encourages a fundamental misunderstanding of CSD and G6 as a venue for this sort of maintenance request.
 * You say that admins "need to finish the job," i.e. they must perform the move. But this has never been the case since the inception of CSD. You (and others, such as Pppery) appear to be conflating G6/db-move reqs with RMT reqs: You seem to believe they are the exact same mechanism, when they are not.
 * It is reasonable to suggest that you have misunderstood the purpose of G6/db-move deletion requests, as there has never been an explicit guarantee that the associated move will be performed by the admin. It is not always appropriate/ideal for the admin to perform the deletion. (AfC reviews being just one example.) If you wanted to guarantee that the move be made in your absence, you should have made the request at RMT instead.
 * If this is a common mistake, then perhaps a G6 copyedit is in order -- to remind nominators that the move may not always occur immediately following the deletion, and to use RMT if they want to guarantee that the move happens in their stead. But if you want to compel admins to always make the moves associated with G6/db-move, I think that language needs consensus. Such a change in policy would likely have consequences that need to be assessed. -Hadal (talk) 17:56, 9 May 2023 (UTC)
 * I understand the difference between G6 and RMTR, but still shouldn't have to clean up after 5 sloppy admin actions in only 3 days, hence Oiyarbepsy's addition is necessary and I would oppose deletion. The aforementioned clause does not mean the admin has to personally press the move button; it means it is their duty to ensure that the deletion does not show up on my database report the next time it is updated, so it does allow for AfC moves (although personally I don't like the db-afc-move process, precisely because I keep having to clean up after it). * Pppery * <sub style="color:#800000">it has begun...  18:16, 9 May 2023 (UTC)
 * @Pppery While I always appreciate a sense of initiative, nobody is forcing you to "clean up" after anyone. Perfection is not required when editing Wikipedia. If admins are not compelled to make the moves, the language could instead be clarified to say exactly what you have said: That they are encouraged to follow up with the nominator, but are not required to perform the move themselves. Because what you and I think the clause means evidently is not what Oiyarbepsy believes it to mean. And it's likely they're not the only one who reads a sense of obligation into the language. I will make this change in the absence of a good reason (and "I have to clean up after them!" is not a good reason, because nobody has to do anything they don't want to do.) --Hadal (talk) 18:26, 9 May 2023 (UTC)
 * I never thought that it would be controversial that an admin shouldn't leave a task half-done in a way that breaks the encyclopedia Oiyarbepsy (talk) 04:29, 10 May 2023 (UTC)
 * It's "controversial" because you are asserting an obligation that never existed, and you have edited policy based on this erroneous belief. This is more of a "break" to Wikipedia than some "dead link" (the horror!) that nobody noticed for a day. You may not see it, but trying to force this belief onto others would represent some significant consequences that go beyond CSD.
 * Admins are not wet nurses; we're not here to prevent editors from making mistakes, because mistakes are how we learn. The idea that an admin is responsible because you incorrectly tagged an article and then walked away is... absurd, frankly. I'm not seeing any good reason not to rectify this, so I'll start drafting some changes. --Hadal (talk) 10:50, 10 May 2023 (UTC)
 * The "obligation" takes ten seconds to complete, I think it's perfectly reasonable to ask an admin to do it. And if the admin doesn't want to do it, they have the easy solution of not deleting and leaving it to someone else. Oiyarbepsy (talk) 19:49, 10 May 2023 (UTC)
 * If you want to ask the admin to do it, ask them to do it. Don't expect the admin to read your mind. (This is generally a bad thing for anyone to do.) Don't put scare quotes around what you are demanding, as if to minimize it: You want to compel admins to make these moves. For no good reason at all. Again, please use RMT if you want to ensure the page is moved on your behalf, or if you don't want to do the move yourself. Please don't confuse your own expectations with the way the tools actually work. --Hadal (talk) 20:23, 10 May 2023 (UTC)
 * Okay, here goes:
 * "For no good reason at all" - I consider avoiding having a high-profile title be a dead link to be a good reason, which I've said before. Are you saying that you don't?
 * "You want to compel admins..." - Speedy deletion means that admins can delete, not that they must. They are always free to ignore it or to remove the speedy deletion tag.
 * "Again, please use RMT..." - The bureaucracy of RMT is far more time-consuming than an admin (or anyone else) pressing the move button. Oiyarbepsy (talk) 00:50, 12 May 2023 (UTC)
 * I think you're creating a crisis out of nothing. You're grossly overstating something that really isn't a big deal. Perfection is not required when editing Wikipedia.
 * Yes, please use RMT if you expect the page move to be performed for you. You're free to avoid CSD if you don't like how it functions; and you're free to avoid RMT if you don't want to take the minimal amount of effort required. CSD is what it is - and it's not RMT.
 * It's perplexing to me that you're complaining about the "bureaucracy" of RMT while trying to create more bureaucracy here. You essentially want CSD to function the same way as RMT, and if that were to happen we'd end up shifting all move-related reqs to RMT, a concept which has been seriously floated. So you'd end up having to deal with RMT anyway. You're welcome to start an RfC if you believe this is really what you want. I will respect the consensus, as I always have. --Hadal (talk) 02:20, 12 May 2023 (UTC)
 * I have always read that line to mean admins should make sure editors are not using it when they have no intention of actually performing the page move and just want the page deleted. In my view, this is a reasonable request. We are just asking the admin to ensure the CSD used – in this case, db-move – actually applies, like they do for every other CSD. I don't think this requires the admin actually do the move, just that they keep an eye out for potential misuse by disruptive editors. HouseBlastertalk 21:28, 10 May 2023 (UTC)
 * It is clear that several people were reading a sense of obligation into the language. The author of that clause clearly did not see it the same way as you. More generally, one could read your interpretation to mean that admins can't trust the editors making these nominations -- which is counter to assuming good faith. It could also suggest a lack of confidence in non-admins' competency, which is also unfair. (I'm not saying this is what you're saying, but it's a possible interpretation.) In the end, a couple of temporarily "broken links" (deleted redirs?) isn't going to "break" Wikipedia. Put simply, CSD is primarily a forum for requesting deletions, not moves. Misusing tags happens everywhere on Wikipedia, and this can be addressed as needed w/ the 'offending' nominator. --Hadal (talk) 21:46, 10 May 2023 (UTC)
 * I can see how one might interpret my comments in that way, so let me clarify: I believe that non-admins can be trusted – I would hope so, given I am a non-admin ;). I am saying that admins should follow a policy of trust, but verify when deleting db-moves. That is, it should behave like (almost?) every other CSD – e.g., if I tag an article as WP:A7, you are not going to just delete the article under A7 before you read the article to make sure that A7 applies. This does not mean non-admins are untrustworthy, but that they are human.<span id="HouseBlaster:1683756014546:Wikipedia_talkFTTCLNCriteria_for_speedy_deletion" class="FTTCmt"> HouseBlastertalk 22:00, 10 May 2023 (UTC)
 * That's perfectly reasonable on its face, but what more needs to be said that isn't already stated on the CSD project page? Admins are expected to follow deletion policy whenever they delete anything. There's a reminder about this under G6 too. If you want to add a new clause to the deletion policy that says... what? That admins responding to G6 reqs need to be exrra vigilant, not that they haven't been careful otherwise, because these deletion reqs are somehow more sensitive or worthy of extra scrutiny? I guess I just don't see this as some kind of exceptional threat to Wikipedia. If someone wants to misuse these tags, they're not going to do any irreparable damage. Wikipedia will never be perfect, and it will never be free of disruption.
 * Recall that the clause we're discussing was added by someone who was disappointed because the responding admin didn't "finish the job" for them. I.e., they misunderstood the mechanism of G6, and tried to turn it into the same mechanism as RMT, purely for their own desire for their expectations to match reality. Admins aren't cops, we're not mind-readers, we're not anyone's teachers, and we're not here to protect editors from themselves. We're human too. Mistakes are rarely emergencies. --Hadal (talk) 22:30, 10 May 2023 (UTC)

Super crazy idea: remove "pages blocking a page move" from G6
Why is this even a part of G6 in the first place? Is there a reason that ROBIN moves are not sufficient to handle redirects blocking page moves (assuming they have more than one revision, in which case they are automatically deleted)? I realize those one-revision redirects are currently deleted with a log entry mentioning G6, but I believe my point still stands: we can just handle these cases at WP:RMT? It would cut down on G6 bloat. HouseBlastertalk 19:57, 7 May 2023 (UTC)
 * The redirects are being speedily deleted so the speedy deletion policy needs to allow for that, also I believe the policy predates automatic deletion of one-revision redirects becoming a thing the software does - previously you had to delete them manually first. If you want to move this out of G6 to a new criterion where the detail can be expanded upon in the policy then you have my complete support - G6 is overloaded and widely abused (intentionally and otherwise), so splitting it up is a Good Thing. Thryduulf (talk) 21:51, 7 May 2023 (UTC)
 * Genuinely, good luck. Every time a thread on the matter gets started, it is soon sidetracked or simply disappears without anything ever happening. Because heaven forbid we give administrators clear options to work with... Primefac (talk) 08:52, 8 May 2023 (UTC)
 * Hence my chivvying at - things idling out without resolution is a bugbear of mine! Thryduulf (talk) Thryduulf (talk) 11:20, 8 May 2023 (UTC)
 * Indeed. Would this be a new CSD criteria or would we simply not allow it any more? Primefac (talk) 11:27, 8 May 2023 (UTC)
 * My first thought would be something like a new criterion but accompanied by a note saying something like "it is very rare that this tag should be added directly, in almost all circumstances it is better and easier to use WP:RM/T." As we do need to allow the speedy deletion of the page blocking the move and there might be occasions when RM/T isn't the best/a possible option (although I can't think of any off the top of my head, I don't do many page moves).
 * A related issue is temporary deletions as part of a round-robin move. On the one-hand these would logically fit with the page-move criterion, on the other I don't like temporary and permanent deletions appearing the same in the logs (and this is another one of the many issues with the present G6). Thryduulf (talk) 11:42, 8 May 2023 (UTC)
 * temporary deletions as part of a round-robin move -> round robin moves as typically performed don't involve deleting anything. * Pppery * <sub style="color:#800000">it has begun... 14:37, 8 May 2023 (UTC)
 * just as a historical note, non-sysops could always move pages over redirects with no history, although sometimes in those early days following the implementation of Phase III the software could be buggy, and it would still be neccessary to request deletion at VfD. 74.73.224.126 (talk) 18:35, 8 May 2023 (UTC)
 * For AfC acceptances it is better for the redirect blocking the move to be first deleted so the AfC reviewer can do the acceptance steps with the script. ROBIN moves can only be done by page movers also. Note not all redirects with 1 revision are automatically deleted, only ones that point back to the page being moved. Galobtter (talk) 15:52, 8 May 2023 (UTC)
 * I do kind of agree that RMT is the better general forum for doing moves since if there is something blocking the move the move should be double checked to make sure it's appropriate. Galobtter (talk) 15:55, 8 May 2023 (UTC)
 * "They can just be handled at RMT" is a bad reason for removing this from G6. What happens when they are handled at RMT often includes a G6 deletion. So removing this from G6 would make things more difficult/complicated/out of process for the admins handling RMT, with no practical benefit. Maybe what you want, instead, is to deprecate the template that tags pages as needing a G6 deletion, and instead push people to make those requests at RMT rather than through the template? But this would still involve an eventual G6 deletion. —David Eppstein (talk) 19:11, 8 May 2023 (UTC)
 * The problem I'm seeing, especially in the parent section, is that there seems to be a push to make RMT the de jure (rather than de rigueur? but see my last paragraph) pathway for users to request technical page moves. I keep seeing G6 reqs and db-move described as "unnecessary" or somehow depreciated; and RMT described as "better" and "easier." Yet I don't think you can beat the simplicity of a db-move tag. And as you've pointed out, there are consequences to pushing more of these on to RMT by way of rewriting G6 that won't necessarily make things better. At the same time, G6 appears to be some kind of flashpoint (to risk overstating it) in a generally-expressed opinion that admins are frequently acting out of process.
 * Maybe it's because I come from the ancient times when we had a text-only front page, but... there seems to be a kind of bureaucracy creep in many areas of policy interpretation in recent years, and shifting (almost?) all technical moves to RMT may end up adding to that creep. If we require non-admins to go through RMT (almost?) every time, I don't think this is necessarily helpful to them: It could be seen as taking agency away from them, as Galobtter alludes to with just one example (AfC reviewers). Essentially, every move that requires a target deletion would need to be 'endorsed' by an admin, because (presumably) only an admin would ever be making the move itself.
 * At the end of the day, I don't see why G6/db-move reqs and RMT reqs need to be compared to each other, especially as a way of measuring their value. (I.e., I don't think language that amounts to "do this instead because it's better/easier" belongs in a CSD criterion definition.) One pathway is not ulltimately more neccessary than the other, and they both ultimately serve the same goal: they're tools for non-admins to bring attention to maintenance issues that they can't fix themselves. If we're going to consider the removal of some of these tools, we need to be very careful about it. --Hadal (talk) 20:45, 8 May 2023 (UTC)
 * Just responding to the G6/out of process part of your comment: at least personally, I am not concerned about admins acting out of process (deletion or otherwise), as long as their actions improve the encyclopedia. My concern is when admins erroneously cite G6 in the deletion rationale. To use an analogy, expanding articles is more than okay – it is encouraged. But when one is expanding an article, the edit summary should not be "fixed typo". I believe that some editors have other concerns over IAR-style deletions in general, but I wanted to clarify that (at least in my mind) IAR deletion and G6 misuse are two separate matters. HouseBlastertalk 21:28, 10 May 2023 (UTC)
 * To me, this proposal seems to be less about the speedy deletion criteria and more about the venue for requesting speedy deletion. I don't see a reason for fiddling with the venue and I concur that we'd still need a "pages blocking a page move" criterion even if we changed venue. Jo-Jo Eumerus (talk) 08:12, 9 May 2023 (UTC)
 * And any admin should be allowed to fix things if they notice (or are notified) that something is wrong by making page moves and deleting pages that block the move without going to WP:RMTR for everything. There is no need to invent additional bureaucracy. —Kusma (talk) 08:20, 9 May 2023 (UTC)
 * I agree, and now I wonder: Is the root cause for many of these disagreements, including widespread assertions that admins are acting "out of process," the result of a fundamental misunderstanding of G6/db-move deletion requests? It seems to me that people are confusing these reqs with RMT reqs, believing that they are the exact same mechanism... despite clearly not being the same. We should always hesitate when further bloating our bureaucracy. --Hadal (talk) 18:06, 9 May 2023 (UTC)

Changes to G6 per discussions above
I have made changes to G6 based on the discussions above, as it became clear that something had to be done. I waited at least two days for any reasonable objections, but found none. To summarize:
 * The following language has been removed, which was self-attributed to Oiyarbepsy in the parent section:
 * "An administrator who deletes a page that is blocking a move should ensure that the move is completed after deleting it."
 * This language was added without consensus, and it arose from both a mistaken belief that administrators are obligated to perform moves associated with G6 deletions; and, an apparent misunderstanding of how G6 move-related requests function vis-a-vis WP:RM/Technical (RMT) requests.
 * This language perpetuated these falsehoods, and I believe it was contributing to the growing conflation of CSD:G6 and RMT: These are similar, but not identical mechanisms. One is not more necessary than the other; one does not depreciate the other. They are both useful tools that compliment each other.


 * The section was reorganized and language was added with the intent to improve clarity:
 * "Nominators are advised that the responding administrator is not required to perform the page move associated with the deletion request, nor is it always appropriate for the administrator to do so. If the nominator wishes to ensure that the page move is performed on their behalf, the request should instead be made at Requested moves/Technical requests.
 * If the responding administrator elects not to perform the associated page move after the requested deletion is performed, it is recommended that they follow up by informing the nominator of the deletion."
 * I included the second line as more of a nod to courtesy than anything else.


 * I hope that these changes make the functional differences between move-related G6 reqs and RMT reqs better to understand for everyone, but especially for nominators.
 * Note that I have not created any new rules and G6 should function the same as it has, but with (I hope) less ambiguity. I ask that any objections be discussed here first, rather than simply jumping to reverts. I have tried my best to be clear and fair in my wording.
 * If Oiyarbepsy or anyone else would still like to compel administrators to perform these moves, I think such a change would require a wide-reaching request for comment, and a firm consensus. Such an obligation would likely have consequences that go beyond CSD.

If you've read this far, I thank you for considering my thoughts. --Hadal (talk) 20:59, 10 May 2023 (UTC)
 * If you are going to have this much detail about considerations required for processing deletions to make way for moves then it no longer really fits in G6 and feels like an entirely new criterion. * Pppery * <sub style="color:#800000">it has begun... 21:43, 21 May 2023 (UTC)
 * Agreed. The more complicated a criterion is the more likely it is to be unintentionally used incorrectly and/or misused. G6 is already an amalgam of various semi-related things, and making it more complicated is the opposite of what we should be doing. Thryduulf (talk) 11:20, 22 May 2023 (UTC)
 * I don't think it's very helpful to have a long discussion about whether an administrator is or is not obliged to move a page as a result of a G6 request. I would have thought it was obvious that administrators are not obliged to move a page just because they deleted a page. The bit which says that in some cases "the request should instead be made at Requested moves/Technical requests" is also misleading because any moves made as a result of requests there are also done under G6. I suggest we just get rid of the two paragraphs under "Move-related" to simplify things.  Hut 8.5  18:31, 22 May 2023 (UTC)
 * Since it seems clear nobody actually supports the changes Hadal made, I've reverted them (after having to clean up after yet another admin deleting something to make way for a move and the page still being red days later). This does not preclude any further changes to G6 if they obtain proper consensus. * Pppery * <sub style="color:#800000">it has begun... 02:53, 11 June 2023 (UTC)

TimedText pages
Are orphaned (TimedText:Bon_Jovi_-_It's_My_Life.ogg.en.srt and TimedText:Lady_Gaga_-_Telephone.ogg.en.srt) TimedText pages and these for files on Commons (TimedText:This-prounciation-audio-nonlabial.ogg.en.srt and TimedText:US_Army_Special_Forces_Dive_Team-1956.ogv.en.srt) eligible for speedy deletion under G8 and G6 (respectively)? The first seems to be straightforward, the second one not. Jo-Jo Eumerus (talk) 09:59, 16 June 2023 (UTC)


 * Yes to the first and no to the second; if the file exists on Commons, but the TimedText does not, then shouldn't the text be copied to Commons? And only then deleted under a slight extension of F8? —Kusma (talk) 10:07, 16 June 2023 (UTC)
 * TimedText makes no sense without the original media file, so WP:G8 is imho the correct criterion to use for orphans. After all, it just says "pages dependent on a non-existent or deleted page", it does not limit the kind of page (the list is just of examples). If the file exists, eligibility to copying to Commons should be checked and if it exists on Commons, F8 applies. On a side note, I think we should change "Images" to "Files" in the header for F8 since the "F" already implies that it applies to all files. Regards So  Why  12:39, 16 June 2023 (UTC)
 * I was thinking of F8 as well but TimedText isn't the File namespace, whereas G6 does encompass all namespaces. Jo-Jo Eumerus (talk) 15:37, 16 June 2023 (UTC)
 * Honestly, given the absolute rarity of this situation, IAR applies. Now just as a thought exercise, assuming a strict CSD interpretation, then imo neither G6 or F8 legitimately apply; we'd either need to extend G6 to include TimedText or create a new CSD criteria for orphaned/trans-wiki'd TimedText content.  -  F ASTILY   05:26, 20 June 2023 (UTC)
 * IAR is most definitely not appropriate. If you think something should be deleted but it isn't covered by a speedy deletion criterion and doesn't need to be oversighted then the only acceptable course of action is to take it to PROD or XfD. In this case it isn't eligible for PROD and MfD is the correct venue. You can propose a new or expanded speedy deletion criterion, but you are wasting your (and others') time unless your proposal meets all four of the WP:NEWCSD requirements. I can't imagine anything involving TimedText and not covered by one or more existing criteria being remotely frequent enough. Thryduulf (talk) 12:55, 20 June 2023 (UTC)
 * The TimedText namespace is obviously an extension of the File namespace. Going through MfD for a TimedText page that should be on Commons like the associated File page would be a WP:TROUT worthy violation of WP:NOTBURO. —Kusma (talk) 12:59, 20 June 2023 (UTC)
 * In your opinion. In my opinion, speedy deleting a TimedText page under an F criterion is the as deleting it using an A or C criterion - i.e. serious questions about your competency worthy. Remember that no admin may delete any page without an explicit consensus to do so, the speedy deletion criteria being the a list of the explicitly very narrowly worded exceptions where there is consensus in advance. As far as I can tell no such consensus exists for the TimedText namespace. Thryduulf (talk) 15:50, 20 June 2023 (UTC)
 * Regarding "no admin may delete any page without an explicit consensus", yeah sure, you're entitled to your own opinion, but that's definitely a minority opinion. -  F ASTILY   07:06, 21 June 2023 (UTC)
 * It's not just an opinion, it's the entire foundation on which the deletion policy is built. Deletions must either meet one of the criteria listed on this page or have consensus. Thryduulf (talk) 08:27, 21 June 2023 (UTC)
 * As Policies and guidelines says, "Policies and guidelines should always be applied using reason and common sense." The deletion policy is not an exception where the rules must be understood ultra-literally. —Kusma (talk) 09:40, 21 June 2023 (UTC)
 * Please explain how following the exact wording of the policy, in several places, including
 * is understanding the policy "ultra-literally".
 * Note also the very-long-standing consensuses that criteria must be interpreted narrowly and that if editors in good faith disagree about whether a page should be speedily deleted it should not. [all quotes taken from Criteria for speedy deletion except the last from Deletion policy. Thryduulf (talk) 10:53, 21 June 2023 (UTC)
 * I don't understand your query. I understand "following the exact wording of the policy" to be the same as "understanding the policy literally". Doing that in all cases, even when it defies common sense, is wrong, as policies and guidelines are not there to be taken literally, but to be applied using reason and common sense. —Kusma (talk) 11:09, 21 June 2023 (UTC)
 * "Applying policy using reason and common sense" is not the same thing as "applying policy in way directly contrary to how it is written". Applying a policy which applies to only "files, images, and other media." to something that is none of those is neither "common sense" nor "reasonable" - especially when consensus going back over a decade says that the policy should be applied very conservatively. Thryduulf (talk) 11:29, 21 June 2023 (UTC)
 * You will not convince me with your argument that it makes a fundamental difference that TimedText pages are stored in a separate namespace instead of being sub-pages of the File page. TimedText pages belong to their corresponding Files, and just because nobody bothered to add them to F8 doesn't mean they should not be covered. But sure, if it makes you happy, you can add the TimedText namespace to F8 explicitly. —Kusma (talk) 11:43, 21 June 2023 (UTC)
 * If namespaces weren't fundamental then CSD would not be split by namespace. Thryduulf (talk) 12:29, 21 June 2023 (UTC)
 * Now reported at Database reports/Possibly out-of-process deletions * Pppery * <sub style="color:#800000">it has begun... 01:37, 25 June 2023 (UTC)
 * I see no reason why G8, "pages dependent on nonexistent pages", doesn't apply. It doesn't say "pages dependent on non-existent pages in the same namespace or in the associated Talk namespace".Largoplazo (talk) 12:05, 21 June 2023 (UTC)
 * I think we should change "Images" to "Files" in the header for F8 Not just in the header. Clearly applies to all files, not just images; amended. —Kusma (talk) 10:21, 21 June 2023 (UTC)
 * "Applying policy using reason and common sense" is not the same thing as "applying policy in way directly contrary to how it is written". Applying a policy which applies to only "files, images, and other media." to something that is none of those is neither "common sense" nor "reasonable" - especially when consensus going back over a decade says that the policy should be applied very conservatively. Thryduulf (talk) 11:29, 21 June 2023 (UTC)
 * You will not convince me with your argument that it makes a fundamental difference that TimedText pages are stored in a separate namespace instead of being sub-pages of the File page. TimedText pages belong to their corresponding Files, and just because nobody bothered to add them to F8 doesn't mean they should not be covered. But sure, if it makes you happy, you can add the TimedText namespace to F8 explicitly. —Kusma (talk) 11:43, 21 June 2023 (UTC)
 * If namespaces weren't fundamental then CSD would not be split by namespace. Thryduulf (talk) 12:29, 21 June 2023 (UTC)
 * Now reported at Database reports/Possibly out-of-process deletions * Pppery * <sub style="color:#800000">it has begun... 01:37, 25 June 2023 (UTC)
 * I see no reason why G8, "pages dependent on nonexistent pages", doesn't apply. It doesn't say "pages dependent on non-existent pages in the same namespace or in the associated Talk namespace".Largoplazo (talk) 12:05, 21 June 2023 (UTC)
 * I think we should change "Images" to "Files" in the header for F8 Not just in the header. Clearly applies to all files, not just images; amended. —Kusma (talk) 10:21, 21 June 2023 (UTC)

==Discussion at Wikipedia:Village pump (proposals) § RfC on draftifying a subset of mass-created Cricketer microstubs== You are invited to join the discussion at Wikipedia:Village pump (proposals) § RfC on draftifying a subset of mass-created Cricketer microstubs. – Novem Linguae (talk) 16:32, 12 July 2023 (UTC)
 * Posting this here because this proposal involves changing G13 from 6 months to 5 years for a subset of 1000 drafts initially, then probably more later as more of Lugnuts' stubs are draftified. – Novem Linguae (talk) 16:32, 12 July 2023 (UTC)

RfC: Removing template exclusion from R2
Should CSD#R2 have the template exclusion removed? SWinxy (talk) 06:22, 13 July 2023 (UTC)

The current wording, with the proposed removal struck out: "R2. Cross-namespace redirects: This applies to redirects (apart from shortcuts) from the main namespace to any other namespace except the Category:, Template:, Wikipedia:, Help: and Portal: namespaces."

Redirects to templatespace are rare and mostly unused: Category:Redirects to template namespace counts approximately 31 non-shortcut redirects from mainspace, e.g. Cite web (though it is certainly an undercount). Half of the templates affected (see collapse) are highly plausible to be or redirect to legitimate articles (e.g. Malaysian music). The remainder are implausible. Redirecting to the template namespace takes readers to a place they did not expect or even want. If supported, the links or transclusions to these pages will be fixed and the redirects may be deleted via the updated R2, or the target be changed to point to somewhere in mainspace.

Survey (Nixing templatespace from R2)

 * Oppose per the initiator's comment that which means the change would be a gratuitous failure of WP:NEWCSD point 2. Thryduulf (talk) 10:09, 13 July 2023 (UTC)
 * Say those are converted to redirect to somewhere in mainspace. Would that still be a failure of criteria 2? SWinxy (talk) 13:47, 13 July 2023 (UTC)
 * Yes. Because speedy deletion criteria are only for situations where everything that could be deleted should be deleted, redirecting is not deleting. Thryduulf (talk) 15:40, 13 July 2023 (UTC)
 * Oppose, if it isn't even clear we want to delete these redirects, we should certainly not debate extending a CSD to them. —Kusma (talk) 10:16, 13 July 2023 (UTC)
 * Oppose, in addition to point 2, I think the evidence is lacking for point 3 as well. RFD is not being overwhelmed with main to template XNRs. 74.73.224.126 (talk) 12:59, 13 July 2023 (UTC)

Discussion (Nixing templatespace from R2)
It's a severe undercount. Full list. Why on earth did you rush directly to an RFC? —Cryptic 08:12, 13 July 2023 (UTC)
 * Just as a note, I've converted the Quebec districts page into an actual list article. Primefac (talk) 11:36, 13 July 2023 (UTC)

F11's 30-day rule
Hello! F11 appears to currently contradict itself (to my eyes, at least). It says that: which implies that anyone can nominate a file under this provision, without further involvement of VRT agents. However, the very next sentence states that: (emphasis mine). So, which is it? (Or am I misinterpreting the criterion?) HouseBlastertalk 23:55, 20 July 2023 (UTC)
 * It's the "without waiting an additional seven days" that makes it different, I think: usually F11s have to be tagged for seven days prior to deletion (just like WP:C1), but a check from a VRT agent allows for immediate deletion as soon as the thirty days are up. Extraordinary Writ (talk) 00:44, 21 July 2023 (UTC)

Disagreement on interpretation of CSD
Posted at Administrators' noticeboard/Incidents. Not by me, because I think this is the correct page instead, but I am reluctant to move it. 2601:5CC:8300:A7F0:A5BE:494E:58AF:9E7C (talk) 05:01, 9 September 2023 (UTC)


 * Replied there, but to reiterate, there's no rule that says only admins can remove CSD tags, and I'm sure community consensus would come down very strongly against creating such a rule. Elli (talk &#124; contribs) 05:06, 9 September 2023 (UTC)

CSD G5 on sock created articles that subsequently suffer public relations editing
IOActive was created as a stub by a sock and subsequently directed edited upon by a disclosed paid public relations clerk of the subject company which was highly promotional. This was mostly cleaned up only to have promotionally edited by a single purpose account. Because of having been edited by someone other than the article creator sock, it was denied CSD G5 removal. Shouldn't this loop hole be plugged up? In my experience, it's common for an article to be created as a stub, then contents added in great amount by COI public relations activity. There's been no real activity on this article other than by disclosed COI/U editing and a WP:SPA that appears to be COI. Graywalls (talk) 03:08, 14 July 2023 (UTC)


 * List at AfD. SmokeyJoe (talk) 03:56, 14 July 2023 (UTC)
 * That's always an option but promotional editing always look for a way to make deletion as difficult as possible this loophole plays along with what they're desiring. CSD G5 could be modified so contribution by COI/U, disclosed PR agents, or presumptive UPE (such as SPAs) don't count Graywalls (talk) 04:00, 14 July 2023 (UTC)
 * Unless I’ve missed something recently, the community has consistently failed proposals for a consensus to declare undisclosed paid editing as a reason for deletion. Maybe individual AfDs have seen substantial UPEditing as a compelling reason to delete, but I have not seen that.  In the meantime, attempting to use G5 where it doesn’t apply to get around the difficulty, that AfD won’t agree, is an unacceptable use of speedy deletion.
 * My suggestion is to engage COI editors to declare the topics they have a COI, and to remind them of the rules at WP:COI. And even more so for undeclared paid editors.  But be careful, WP:OUTING of their personal information is not allowed, even if you know they are lying. SmokeyJoe (talk) 04:40, 14 July 2023 (UTC)
 * I agree with SmokeyJoe, UPE by itself is not a reason for deletion, and if someone other than a sock has edited a page it is no longer a G5 candidate. Primefac (talk) 12:27, 14 July 2023 (UTC)
 * I agree completely with Smokeyjoe and Primefac. UPE has been explicitly rejected as a reason to speedy delete an article multiple times. Thryduulf (talk) 13:20, 14 July 2023 (UTC)
 * Without having looked at the article - If the non-creator's edits were so promotional that they needed to be rewritten, then those revisions were by definition deletable as G11. I suppose if they've already been rewritten (not just reverted) by the time the socking is discovered, you'd need the agreement of the rewriter too to delete their revisions as G7; but the mere fact of there being contributions other than from the creator is immaterial if all those contributions are also speedyable. —Cryptic 04:17, 14 July 2023 (UTC)

Something I just learned is that G5 seems to be treated like a "soft delete". If sock puppetry is identified and G5 is invoked during AfD, it will fail subsequent G4 deletion when another sock comes back and re-create the G5'd article because it's not treated as letting the AfD going all the way through regardless of AfD was on track towards consensus of delete. Articles_for_deletion/Big_Dean%27s_Ocean_Front_Cafe Graywalls (talk) 09:50, 9 September 2023 (UTC)
 * G4 can only be used if the XfD closed as delete. It can't be used if there was a procedural close of the XfD because the content was deleted through a different mechanism, such as speedy deletion. For that reason, it could sometimes be desirable to wait out the XfD instead of deleting speedily.
 * I see that the account used for re-creating the article just has been blocked for sockpuppetry, so I don't see why it can't be deleted as G5 a second time. In fact, User:Bilby did this just as I was writing this comment. --Stefan2 (talk) 12:49, 9 September 2023 (UTC)
 * I see what happened here as a clear bureaucratic fail. If the same person goes back and re-re-create it a week later from a different location using a re-written composition to look different enough from the original, it might slip through C/U test and I'm concerned it'll be denied G4 again over the same reason taken to the letter. So, I think patiently letting the original Afd until the end would realistically have created a sturdier delete. Graywalls (talk) 16:37, 9 September 2023 (UTC)
 * If the same sock is recreating it, though, it could be easily G5 deleted. Elli (talk &#124; contribs) 21:09, 9 September 2023 (UTC)
 * If it is clearly the result of a banned user (whether the same one or not) then G5 may be used to delete it (assuming no significant contributions by others, etc). If it is unclear then it cannot be - and this is a Good Thing. Except in the cases of attack pages and copyright violations, there is never any harm in waiting for a full XfD rather than speedy deleting. Thryduulf (talk) 06:57, 10 September 2023 (UTC)
 * There seem to be a few unfounded assumptions by the OP here. The "standard" way of deleting an article is by consensus at AfD. Speedy deletion is for clear-cut cases. Blatantly promotional articles can be speedily deleted by WP:G11, and rather less promotional articles can still be deleted at AfD for that reason. Articles that have been speedily deleted, rather than gone through an AfD discussion, can be speedily deleted again for the same or a different reason, just not via WP:G4 which is designed for pages that have been discussed. There is no "loophole" or "bureaucratic fail" here. Phil Bridger (talk) 19:47, 10 September 2023 (UTC)
 * I would say the "standard" way of deleting an article is via speedy; there are more articles speedied in an hour than AFDd in a day. Stifle (talk) 13:30, 21 September 2023 (UTC)
 * You'd be wrong to say that. XfD is the standard deletion process in all cases, PROD and speedy deletion are explicitly limited exceptions to the normal process. This is why objections to PROD or speedy go to XfD, and any cases where it is not clear whether speedy or PROD is applicable go to XfD. Thryduulf (talk) 15:16, 21 September 2023 (UTC)
 * Depends how you define "standard". I define it as "most usual, most frequently invoked", and by that definition speedy deletion is absolutely standard. Stifle (talk) 13:53, 22 September 2023 (UTC)
 * By the number of deleted pages it is the most common, but that does not mean it is standard. Speedy deletion is one of the two exceptions to the standard deletion policy that requires explicit affirmative consensus to delete a page (the other exception is PROD, where presumed consensus based on lack of opposition is accepted). Thryduulf (talk) 23:58, 22 September 2023 (UTC)

What's the point of G5?
Why bother deleting an article just because it was made by a blocked/banned user?

I understand why you'd delete it if it was in a violation of a topic ban but why do it if it was made by someone who's been banned/blocked from editing any page?

If the article follows all of the guidelines such as Wp:GNG then why bother? Isn't that just a waste of time? 172.58.241.204 (talk) 15:07, 8 September 2023 (UTC)
 * The waste of time is the banned user evading their ban, not admins deleting it. * Pppery * <sub style="color:#800000">it has begun... 15:09, 8 September 2023 (UTC)


 * If a user is blocked or banned, this is a way of saying go sit in a corner. If content created by the blocked or banned user is allowed, this is a way of saying that the user does not need to sit in the corner, he just needs to contribute under a different name. It is an invitation for sockpuppetry. --Stefan2 (talk) 15:48, 8 September 2023 (UTC)
 * Such an edit is a violation of a topic ban. The user has been banned from editing all topics. The justification is that it wastes time to check whether the edit is legitimate when the editor has been shown to make illegitimate edits. Phil Bridger (talk) 16:13, 8 September 2023 (UTC)
 * Then why ban/block anybody for anything? If someone commits 100 acts of vandalism for every good edit, just keep reverting the vandalism while leaving the good edits in place and let the person keep wasting everyone's time, is that it? A block is meant to be a block, not just an inconvenience forcing the blocked user to create a new account before carrying on as before. It's also meant to relieve the rest of us of having to scrutinize the user's edits and additions individually.
 * Note, however, that G5 doesn't require a page created by a banned user to be deleted. If you find a page listed for G5 deletion and believe it's a good article, you can remove the tag. Largoplazo (talk) 16:26, 8 September 2023 (UTC)


 * What Phil Bridger said about wasting time. Users are banned because their editing is so problematic that it's net negative to allow them to edit at all, either in a particular area for a topic ban, or altogether for a site ban. G5 (and related policies like WP:BANREVERT) are saying, "your editing has been so bad that you were banned, so we can presume any new edits you make are also bad and not waste our time checking." In the event that some other editor in good standing does review the edit and adopts it then G5 no longer applies, but then that user takes full responsibility for the edit's content. Ivanvector (Talk/Edits) 16:33, 8 September 2023 (UTC)
 * Otherwise it creates an incentive to keep trying to sockpuppet, which wastes even more time. –&#8239;Joe (talk) 17:52, 8 September 2023 (UTC)
 * DF67 has a good allegory about this. Primefac (talk) 18:06, 8 September 2023 (UTC)
 * The fact that the example uses paintings is actually interesting, given that sockpuppetry is not valid as a sole reason to delete images on Commons. -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 20:34, 8 September 2023 (UTC)
 * The problem I have is people like Slowking4. This person has been blocked for over 10 years. Since the block, they've created innumerable articles, many of which have been kept. The person is still a very active member of this project. 70 times now sockpuppets of this person have been brought to scrutiny. Yet, we still don't delete articles created by this person. In so doing, we are allowing them to be an editor here despite their ban from the project. --Hammersoft (talk) 11:24, 10 September 2023 (UTC)
 * We certainly have G5-deleted many contributions by Slowking. I G5-deleted several uploaded images (along with rollbacking all added article content) from one Slowking sock just yesterday. —David Eppstein (talk) 17:11, 10 September 2023 (UTC)
 * Really? (nothing against you) That explains why confirmed socks of Slowking4 have created more than 70 mainspace pages in the last month alone that remain undeleted. What is the purpose of a ban from the project if the editor continues editing? We are actively rewarding their behavior. This middle ground creates an undue burden on the project. Either we welcome Slowking4 back with open arms, or we start enforcing G5 against all their creations. As is, we're doing the former already. --Hammersoft (talk) 17:33, 10 September 2023 (UTC)


 * G5's purpose is revenge and punishment. It has nothing to do with actually improving anything on Wikipedia and, in most cases, is actively detrimental by deleting perfectly normally written notable articles. But getting revenge on banned users is more important to many than building an encyclopedia. Thus, the purpose of G5. Silver  seren C 17:14, 10 September 2023 (UTC)
 * This is exemplified and shows claims above of checking edits taking too much time as being an incorrect reasoning by the fact that getting a G5 deleted article undeleted so that it can be worked on and kept is a massive hassle. Because admins don't want to do it. Because it's not about an editor in good standing taking responsibility for said article content, it's about punishing the banned user. And having the banned person's articles be undeleted for any reason isn't punishing them. Therefore, while G5 should be absolutely a reason why WP:RFU can be used (and, indeed, G5 isn't on the list of ones not allowed at RFU), it's close to impossible to get a G5 undeletion request on there approved. I've only managed it a handful of times after massive amounts of time and effort arguing, time and effort that showcases the claims of those above to be wrong. Silver  seren C 17:19, 10 September 2023 (UTC)
 * G5's purpose is revenge and punishment is absolutely not correct. I wrote an essay last year here which attempts to explain the intent and spirit behind G5: G5 is not a firm rule. Mz7 (talk) 01:13, 13 September 2023 (UTC)
 * In practice, I've found G5 to be used pretty aggressively. Mass deletions of non-problematic articles by editors such as and  come to mind. – Novem Linguae  (talk) 18:57, 13 September 2023 (UTC)
 * "You are banned from Wikipedia. This means you are not allowed to create new articles."
 * "Ah, but I am allowed to create new articles! So I'm not really banned!"
 * We have few rules, but we do have some. DS (talk) 23:10, 24 September 2023 (UTC)


 * The point is that banned means banned. Stifle (talk) 13:54, 22 September 2023 (UTC)

G2 in user namespace
The rule that G2 does not apply to the user namespace is being rampantly violated by admins actually doing speedy deletions. I think it's time that we reflect the de-faco consensus here and delete pages in the user namespace from G2. * Pppery * <sub style="color:#800000">it has begun... 04:30, 16 September 2023 (UTC)


 * No, we should not. G2 in user space is currently only applied against newbies (like U5). If it is officially extended to all userspace, some people with too much time on their hands will go around and tag all pages called "test" or with a creation summary of "test" in your and my userspace for G2, which would not be enjoyable for any non-troll. —Kusma (talk) 07:46, 16 September 2023 (UTC)
 * The correct solution is to advise those admins that their actions are contrary to the policy, and start a discussion to see what the current consensus is. Personally I am opposed to removing the userspace exemption, as someone who quite often does testing in my userspace. Thryduulf (talk) 10:52, 17 September 2023 (UTC)
 * I wouldn't be too excited to see G2 expanded either. The broader problem here is that the only way to remind admins about the criteria is to leave a message for them and/or go to DRV, which, if we're being honest, most people (including me...) aren't usually willing to do because it typically comes across as making a big deal over something trivial. Pppery does yeoman's work in this regard, but it's not realistic to expect him to do it all himself. What other options do we have? It looks like the edit filter works on deletions, so would it be possible to use it to warn admins who are about to make a questionable deletion? This would require care and wouldn't work on everything, but for some things (e.g. the "wrong namespace" section of the database report) it might be worth considering. Any other ideas? Extraordinary Writ (talk) 23:31, 18 September 2023 (UTC)
 * Uh oh. Two of those deletions in the current report are mine.  You're welcome to bring them to DRV as a test case.  I promise I won't be offended? —Cryptic 23:59, 18 September 2023 (UTC)
 * Well, those could just as easily be called G6s (of the "unambiguously created in error" variety), and either way I still wouldn't be comfortable spending a week of DRV's time to undelete pages that everyone agrees should be deleted some way or another. A low-effort, low-drama way of reminding admins about this sort of thing on the spot would be worthwhile, though, and I'd be interested to hear thoughts about whether the edit filter idea or anything similar would be workable. Extraordinary Writ (talk) 00:17, 19 September 2023 (UTC)
 * If it's workable, and I'm not qualified to say whether it is or not, the edit filter seems like a good idea for the wrong namespace issues. Maybe a bot could iterate through the report and deliver a message saying "this deletion you did was possibly of process because X. If you think this is an error, please discuss it [here?]". Obviously there will be some false positives (e.g. misclicks on the deletion dropdown). Not instant feedback but reliable feedback, and it'll possibly highlight any obvious patterns. Thryduulf (talk) 00:24, 19 September 2023 (UTC)
 * I've posted specific abusefilter code for "wrong namespace" over at Edit filter/Requested/Archive_4. * Pppery * <sub style="color:#800000">it has begun... 03:45, 29 September 2023 (UTC)
 * An edit filter for wrong namespace should be workable, although I think the exact criteria would need to be rethought - when writing these database reports I aim hard toward favoring false positives over false negatives, whereas an edit filter should be designed more toward minimal false positives. "No criterion" is also technically edit-filter-able, and "not recently-created" probably is but I'm not 100& sure.  None of the other sections are edit-filter-able, usually because they rely on some data that the edit filter doesn't have access to. If someone wants to write a bot to monitor my report and notify people, they are welcome to, but it's not a project I feel like taking on. * Pppery * <sub style="color:#800000">it has begun...  01:34, 19 September 2023 (UTC)
 * We could remove G1 and G2 from MediaWiki:Deletereason-dropdown for userspace pages, since they shouldn't be used. I've just checked on Twinkle and both of them aren't available in userspace. I wouldn't be surprised if people starting filling it in manually, but hopefully that would make a bit of a difference. Sdrqaz (talk) 00:40, 19 September 2023 (UTC)
 * If that's technically possible (I have no idea), we should do that for all criteria not valid in a given namespace. I don't think it will completely fix the problem but it should help. Thryduulf (talk) 01:37, 19 September 2023 (UTC)
 * It's certainly possible - if you go to delete User talk:Cryptic, none of the article criteria, nor "G8: Talk page of a nonexistent or deleted page" will show up. It's something of a blunt instrument, though, since if you go to delete User talk:Cryptic/g2, that specific G8 variant still won't show up, even though it should.  It's unclear how much parser function trickery MediaWiki:Deletereason-dropdown supports - I suspect not much, since otherwise surely someone would have made the "OVERSIGHTER ONLY CRITERIA, DON'T USE THESE YOU PLEBE" only appear to overseers - and I'm too lazy to go to a testwiki to find out.  But plain namespace restrictions?  Those are easy. —Cryptic 02:14, 19 September 2023 (UTC)
 * I think the dropdown supports all parser functions, but parser functions can't do user-specific logic so the oversighter text has to be shown to plebs. * Pppery * <sub style="color:#800000">it has begun... 04:55, 19 September 2023 (UTC)
 * It seems plebs can rejoice at not bearing witness to the dropdown reasons for oversighter's given MediaWiki:Deletereason-dropdown-suppress now exists. &#8211;<span style="font-family:CG Times, times"> MJL &thinsp;‐Talk‐☖ 16:36, 27 September 2023 (UTC)
 * Yes, that's a very good idea, and one of us should probably just do it BOLDly unless there's an objection. Something else that's helpful is having the CSD templates display an error message when they're used in the wrong namespace. I see, , and a few others already do this, but there might be some benefit in expanding it to , , etc.—maybe someone skilled in template wizardry could take a look? Extraordinary Writ (talk) 06:15, 19 September 2023 (UTC)
 * I wrote the aforementioned warning message for G1 and G2 many years ago. I didn't do it for other criteria partly because there were a lot of them and partly because I assumed people would know better if the namespace letter was right there blaring at them. * Pppery * <sub style="color:#800000">it has begun... 03:45, 29 September 2023 (UTC)
 * Those two pages could properly have been speedily deleted as G8 given they consisted only of a single link to a page that was deleted/doesn't exist, although the one in . When speedy deleting it is as important to use the correct criterion as it is that the page meets a criterion, this is because it offers an easily understanable explanation to both other admins and to the editor whose page was deleted. Thryduulf (talk) 00:38, 19 September 2023 (UTC)
 * Linking to a non-existent page is even less a speedy deletion criterion than creating editing tests in someone else's userspace. The same user's sandbox would be just as deleteable if these were proper G8s. —Cryptic 01:10, 19 September 2023 (UTC)
 * G8 for pages that are dependent on deleted or non-existent pages. A page consisting of literally nothing but redlinks meets that criterion. The user's sandbox could technically be deleted per G8, but there is no benefit to doing so Thryduulf (talk) 01:33, 19 September 2023 (UTC)
 * Lists of red links can be highly valuable. If pages consisting only of red links can be deleted under G8 (I don't think they can), G8 will need to be fixed. —Kusma (talk) 06:18, 19 September 2023 (UTC)
 * Years ago I was accused of being prejudiced for thinking that, first by using G8 and then PROD; I never did get an explanation as to why I was a racist to think this was eligible for deletion. The Blade of the Northern Lights ( 話して下さい ) 16:18, 21 September 2023 (UTC)
 * There are non-redlinks there now so it definitely wouldn't be subject to G8 today, and even if they weren't all red I wouldn't speedy it as a bit of context could easily make it an article (whether it would survive AfD is irrelevant). Thryduulf (talk) 17:43, 21 September 2023 (UTC)
 * What other options do we have? A simple solution might be to ping everyone in that report to this discussion, so they know not to do that anymore. Normally "G" means "all namespaces", so it's a reasonable mistake to make. – Novem Linguae (talk) 16:30, 19 September 2023 (UTC)
 * It's not a crime to create a page in somebody else's user space - I offer [//en.wikipedia.org/w/index.php?title=User:Whispyhistory/Userboxes_DYK&action=history User:Whispyhistory/Userboxes DYK] for consideration. The scenarios that may happen include:
 * User:A creates the User:A page or User:A/subpage
 * User:A creates the User:B page or User:B/subpage, and User:B is registered
 * User:A creates the User:B page or User:B/subpage, and User:B is not registered
 * Pages in case 1 are normally speedyable under U1, if requested by User:A
 * Pages in case 2 are also speedyable under U1, if requested by User:B
 * Pages in case 3 are normally speedyable under U2 - anybody can request this
 * In all three cases, G7 may be used by the page creator, and some other G criteria may apply - such as G10. -- Red rose64 &#x1f339; (talk) 21:10, 19 September 2023 (UTC)

U5 equivalent for drafts
I've done some CSD tagging in my time, quite often, pages that meet U5 also meet G11, though not always. Occasionally I find a (usually autobiographical) draft page that would meet U5 if it were in userspace and A7 if it were in mainspace, but I am not sure if it meets G11 per se. Should there be a new criterion or expansion of an existing criterion to cover this circumstance? TornadoLGS (talk) 02:49, 11 October 2023 (UTC)


 * No because draft space is where we tell people who want to write about themselves or who have a conflict of interest to create a draft and submit through AFC so they can get feeedback and correct any POV issues. -- Whpq (talk) 03:00, 11 October 2023 (UTC)
 * The point of draftspace is where articles can be worked on and improved until they meet the standards for mainspace without fear of the work being deleted because it's not immediately up to scratch. Generally the only CSD tags you should be placing on drafts are G5, G7, G10 and G12. Everything else will either be improved or deleted via G13 when the time comes. Thryduulf (talk) 10:28, 11 October 2023 (UTC)
 * G3 as well. Fram (talk) 10:50, 11 October 2023 (UTC)
 * Interesting that you don't mention G11, since drafts are regularly deleted under that criterion. TornadoLGS (talk) 00:44, 23 October 2023 (UTC)
 * That's pointless busywork at best (drafts are not indexed so they're doing no harm) and harmful at worst (biting contributors who want to contribute encyclopaedia articles, although this isn't the case with every draft it is with some). They're only an issue if they are disruptively resubmitted without improvement, and that's well established as a reason to delete them at MfD. Thryduulf (talk) 02:12, 23 October 2023 (UTC)
 * I can see your point there. Though the issue I've had on that matter is promotion-only accounts getting autoconfirmed and then moving or copying their drafts into mainspace. Aside from G11-ing drafts, how should it be made clear that promotional editing is not permitted? TornadoLGS (talk) 03:37, 23 October 2023 (UTC)
 * By engaging with the editors concerned, e.g. by using talk pages and template messages on the draft. Thryduulf (talk) 12:06, 23 October 2023 (UTC)
 * I'd be very wary of expanding U5 w/o first spelling out what sort of content actually falls under "webhost", I see too many instances of folks calling drafts or sandbox-like pseudo-articles "webhost" Jo-Jo Eumerus (talk) 09:33, 14 October 2023 (UTC)
 * See /Archive 84, which spun off into a lot of general discussion of what the criterion means. Personally I've decided to just never review U5s as an admin in part because the criterion is so ambiguous. * Pppery * <sub style="color:#800000">it has begun... 14:45, 14 October 2023 (UTC)
 * I tend to put off the category until last, and then delete the overwhelming majority as G11s, and the overwhelming majority of the remainder under some other criterion. Pages that meet every arm of U5 as written are pretty rare; pages that fail every arm of U5 other than being in userspace, but get tagged as such anyway, are quite common, and the userspace police get really upset if you decline them. —Cryptic 22:17, 14 October 2023 (UTC)
 * Make that three of us. I also give other categories precedence and usually only delete under U5 if I would delete the same page in draftspace (i.e., it meets another CSD), or if a page is (borderline) eligible under another criterion and is clearly not intended to be an article draft, userpage, or essay. Complex / Rational  23:03, 14 October 2023 (UTC)
 * I'd like to throw WP:A10 into this mix. I just refused a G1 request at Draft:Vehicles - arguably A10 would apply if it was in article space, but it isn't. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  12:54, 19 October 2023 (UTC)
 * My bad, that's already covered under WP:NOTCSD. <b style="color:#98F">W</b><b style="color:#97E">a</b><b style="color:#86D">g</b><b style="color:#75C">ge</b><b style="color:#83C">r</b><b  style="color:#728">s</b><small  style="color:#080">TALK  12:57, 19 October 2023 (UTC)

CSD G6
Just a notice that Category:Candidates for uncontroversial speedy deletion has been changed to Category:Candidates for technical speedy deletion because of a CFD closure, Categories for discussion/Log/2023 September 30. I thought this change should get a little more publcity, especially for admins who patrol CSD categories. I thought "Housekeeping" was descriptive enough but we now have "technical speedy deletion". Liz <sup style="font-family: Times New Roman; color: #006400;">Read! Talk! 22:09, 28 October 2023 (UTC)

Wording of the G7 criterium
So, sometimes it happens that someone (me, in this case) does a bunch of speedy deletions while forgetting about a part of the pertinent CSD criterium. In this case, the other than a userspace page, a category page, or any type of talk page disqualifier in the WP:CSD criterium. That said, before re-un-deleting a bunch of these pages I'd like to query whether that formulation is meant to exclude items like Template talk:K League 2 Young Player of the Year from G7 deletion or not. From Wikipedia talk:Criteria for speedy deletion/Archive 53 it seems like the purpose of that clause was to exclude scenarios where pages with multiple non-trivial editors get deleted? Say a talk page with multiple editors attached to a main page with only one. Jo-Jo Eumerus (talk) 12:49, 21 October 2023 (UTC)


 * I don't see any reason not to delete talk pages with no substanial content and no other editors like that one. The other 2 exceptions to G7 already have a CSD for them, WP:C1 for categories and WP:U1 for user pages. Either we could amend G7, or we could use WP:G6 for pages that were clearly created in error. Iffy★Chat -- 17:56, 21 October 2023 (UTC)
 * Would it work better if it said If the sole editor blanks a page other than a userspace page or a category page, this can be taken as a deletion request., and put something under G8 to say that if a main article is legitimately tagged under G7 but there are edits by other editors on the talk page, the talk page may not be eligible under G8? That might deal with the problem mentioned in the archived discussion. Jo-Jo Eumerus (talk) 06:44, 23 October 2023 (UTC)
 * This would revert back to the pre-2009 version of WP:G7. Looking back at that, I don't see the justification for that change so I'm happy to go back to what it was back then. For G8, I don't think we need to change the wording. We have, which anyone can add if they think the page shouldn't be deleted under G8. Iffy★Chat -- 13:47, 24 October 2023 (UTC)
 * Actually, in that discussion some editors (e.g ) did see a problem with the previous wording. Granted, it seems like one that could be resolved via a change to the G8 criterium rather than G7, but let's see what others think. Jo-Jo Eumerus (talk) 09:08, 25 October 2023 (UTC)
 * What problem is trying to be solved here? While the template talk page you link above is not the sort of thing the criterion is intended to block the deletion of, I have to ask why you want to delete it in the first place? If you nominated the page for deletion at MfD I doubt you'd get a consensus to delete it, which indicates that it should not be speedily deleted. Thryduulf (talk) 10:13, 25 October 2023 (UTC)
 * Mostly that it doesn't seem correct to blanket-exclude talk pages from G7. It's mostly neatfreaking - true, that particular blank page can simply be restored to the wikiproject template instead of deleted, but that's a different consideration that wouldn't apply to e.g Talk:Southern Miss Golden Eagles basketball statistical leaders. Jo-Jo Eumerus (talk) 10:37, 25 October 2023 (UTC)
 * In that case I'd either have left a blank page or added an appropriate WikiProject template, and I'm struggling to think of any situation where deletion would be better than that and something like G10 or G12 doesn't also apply. Thryduulf (talk) 11:20, 25 October 2023 (UTC)
 * I am not sure why a blank page would be better than a deleted page; certainly the existence of the blanking clause in G7 does not argue in favour of this notion. Also, that presumes that the WikiProject tagging will always be correct, which I am not sure of. My question tho' is less about one specific page or another but about whether the blanket-exclusion of talk pages from G7 is necessary (rather than a note on G8 about multi-user talk pages of a G7-deleted main page not necessarily being fit for G8 ) Jo-Jo Eumerus (talk) 08:39, 29 October 2023 (UTC)
 * If the WikiProject tagging is incorrect then it can be fixed. Blanking a page is something anyone can do whereas deletion requires an administrator. I'm not understanding why you think admins should be deleting pages when not required (deletion should always be restricted to occasions when it is better than some alternative, speedy deletion even more so. Despite thinking about this off and on since the topic was first raised here I've yet to come up with an example of an occasion where deleting a talk page under G7 where some other speedy deletion criterion doesn't apply that offers any benefits at all. Thryduulf (talk) 12:36, 29 October 2023 (UTC)
 * None of the deletion policy criteria (except maybe G10 and G12) require deletion, so that doesn't sound like an argument. Is there an advantage/reason why a talk page blanked by its creator and sole editor is better left in a blanked state than deleted? 'cause it can distract when people are paging through page and edit lists to find content that needs cleanup. Jo-Jo Eumerus (talk) 12:51, 29 October 2023 (UTC)
 * What you need to establish is what benefit there is to deleting such a page? I don't understand how a blanked page adds interferes with cleanup? Thryduulf (talk) 13:06, 29 October 2023 (UTC)

These cases have different reasons for not treating blanking as a request to delete. —David Eppstein (talk) 16:35, 29 October 2023 (UTC)
 * For userspace pages: users should be able to manage their pages and their history, including blanking and re-starting, without having the history wiped out from under them.
 * For category pages, if the category is still in use, then it should stay blue even if blanked.
 * For talk pages, users often blank talk pages as a way to archive or close a discussion. This is ok for user talk pages but may also mistakenly be done for other kinds of talk pages. It should not be used as a reason for wiping the history of the talk page.

Unintended malfunctioning redirect
I unknowingly created a malfunctioning redirect Epidemiology of Familial Mediterranean fever. Thank you. Invasive Spices (talk) 22:24, 3 November 2023 (UTC)


 * Double redirect fixed. Not really a topic for a discussion page about our criteria for speedy deletion. Whether the other newly created redirect Epidemiology of Familial Mediterranean fever in the United States of America is worth keeping is unclear to me, but also not really a topic for this discussion page. —David Eppstein (talk) 22:40, 3 November 2023 (UTC)

Request for new redirect criteria
Over the past couple days, there have been a lot of nominations at WP:RFD for redirects with unusual wikipedia spaces (such as Philip Dunne(writer)). I suggest to add a criteria for redirects basedd off WP:RDAB Yoblyblob (talk) 16:06, 17 November 2023 (UTC)


 * @Yoblyblob The example you gave looks like G6. It was created obviously in the wrong title, so it probably can be speedied under there. I think G6 is well equipped to handle this case. Awesome Aasim 19:53, 19 November 2023 (UTC)
 * We should still have a clear criterion for this (assuming the consensus is that they should be speedy deleted) rather than stretching G6 and relying on the CSD-ing admins to mind-read back to 2005. * Pppery * <sub style="color:#800000">it has begun... 20:06, 19 November 2023 (UTC)


 * I don't think we need a new criterion for this. Despite the recent spate of nominations, I don't think the missing space issue arises frequently enough to justify a new criterion. If you're talking about a new criterion for WP:RDAB in general, I don't think that would be wise, because "errors in disambiguation" is open to interpretation. Discussions where it's invoked, like the conversation underway at Redirects for discussion/Log/2023 November 13, don't necessarily arrive at a decision to delete. - Eureka Lott 20:21, 19 November 2023 (UTC)

Was there an unintended change in scope of G14?
This edit by back in Jan 2021 was described as a "restructure for clarity", but it seems to have also resulted in a significant change to the scope of G14. Whereas previously, the bit about "pages that perform a disambiguation-like function (such as set index articles or lists)" was limited to -- after the restructure, it appears that all aspects of G14 are now also applicable to set indexes as well as to disambiguation pages. As a result, a well-meaning editor has been changing set index pages to redirects because they only contain one existing article. However, one of the reasons for the existence of set indexes distinct from disambiguation pages is precisely to allow for entries that do not have an existing article. Was there ever any discussion about this, or was this change in scope accidental? older ≠ wiser 17:00, 1 December 2023 (UTC)


 * I'm not familiar with the previous discussion about this, but I did experienced this issue when I nominated G14 candidates of SIAs and lists before, including MV Buccaneer, List of Development Regions of British Columbia, List of chairmen of the State Council of the Republic of Adygea and List of chairmen of the Chamber of the Nations (Czechoslovakia). The related discussion includes User_talk:EurekaLott, Special:Diff/1158664588, User_talk:Liz/Archive_49 and Articles for deletion/List of chairmen of the State Council of the Republic of Adygea. NmWTfs85lXusaybq (talk) 18:34, 1 December 2023 (UTC)
 * I am not aware that there was any previous discussion on this point. I think the edit was only intended as clarification -- but inadvertently resulted in expanding the scope of G14. Previously, the applicability to set indexes was limited to . older ≠ wiser 18:40, 1 December 2023 (UTC)
 * I don't remember, and can't find, any discussion about changing the scope. I would support changing the scope back to the original. Thryduulf (talk) 19:00, 1 December 2023 (UTC)
 * I’m not sure if this is related but Deletion%20review/Log/2023%20November%2027 shows what appears to be an emerging consensus to overturn multiple G14 closes as invalid.--67.70.103.36 (talk) 23:43, 1 December 2023 (UTC)
 * That's not related, other than both being a response to NmWTfs85lXusaybq's actions. It's about a different misinterpretation of G14. * Pppery * <sub style="color:#800000">it has begun... 23:48, 1 December 2023 (UTC)

Proposed update
This applies to disambiguation pages and redirects to disambiguation pages:


 * 1) Disambiguation pages that have titles ending in "(disambiguation)" but disambiguate only one extant Wikipedia page.
 * 2) Regardless of title, disambiguation pages that disambiguate zero extant Wikipedia pages.
 * 3) A redirect that ends in "(disambiguation)" but does not redirect to a disambiguation page or a page that performs a disambiguation-like function (such as set index articles or lists).

If a disambiguation page links to only one article and does not end in (disambiguation), it should be changed to a redirect, unless it is more appropriate to move the linked page to the title currently used for the disambiguation page.

Bullets are changed to numbers to facilitate referencing. There is minor change to the lead-in. #3 has been updated to clarify the scope. older ≠ wiser 21:05, 2 December 2023 (UTC)
 * I support this update. However, what about the case that all topics in a disambiguation page are actually in the same article? Should these pages be changed to a redirect? For example, see Macmillan ministry and Spider-Man trilogy. NmWTfs85lXusaybq (talk) 00:55, 3 December 2023 (UTC)
 * I don't see that as a question particularly relevant to CSD, and it likely varies depending on the article in question. Elli (talk &#124; contribs) 01:00, 3 December 2023 (UTC)
 * There will always be edge cases. If it isn't clear whether the criteria apply, it is perhaps best to discuss rather than try to force changing to a redirect based on speedy deletion criteria. older ≠ wiser 01:02, 3 December 2023 (UTC)
 * I am wary of dismissing such cases. Clearly there are two distinct Spider-Man trilogies, which are ambiguous to each other, and may be referenced ambiguously to each other. It would theoretically be possible to have separate pages on each, and the redirects go to distinct sections of the page, not generally to the article as a whole. Perhaps those aspects should be the test. I would keep this one as a separate disambiguation page. BD2412  T 01:21, 3 December 2023 (UTC)
 * I wasn't dismissing them -- I just don't think CSD is directly applicable. The note here is just a hint about how to handle pages that do not fit the CSD criteria. If more detailed guidance is needed, WP:DAB or WP:MOSDAB is a more appropriate place than CSD criteria. older ≠ wiser 02:13, 3 December 2023 (UTC)
 * Which two series are the trilogies? It depends on whether "No Way Home" counts as one installment for each of the three Spider-Man universes. It's doubly ambiguous! Largoplazo (talk)
 * Support. Regarding the question above, it is a general principle (arising out of the requirement for CSD to be used only in the most obvious cases) that if it is unclear or arguable whether a criterion applies to a given page then it doesn't. Thryduulf (talk) 14:58, 3 December 2023 (UTC)

RfC: Does CSD:G4 apply to a page deleted in AfD and immediately recreated as a redirect page?
An ongoing DRV discussion brought up the question of whether the "sufficiently identical" language in CSD:G4 applies to pages deleted in AfD, and immediately recreated as a redirect by someone who wasn't pleased with the result of the AfD. Several editors believe that since the redirect page is different from the one deleted, any editor is welcome to recreate the deleted page, even if the "Redirect" option was brought up in the AfD and failed to reach consensus.

The options, as I see it, are:
 * 1) G4 applies to all cases where consensus was not followed, including recreation of a page as a redirect.
 * 2) G4 does not apply to the recreation of a deleted page as a redirect, even if that option was discussed and rejected in the AfD.
 * 3) G4 applies to the recreation of a deleted page as a redirect, but only if that option was brought up and rejected in the AfD.
 * 4) Other?

To be clear, we are talking about a page recreated shortly (<30 days) after an AfD closed as Delete.

Either way, there seems to be enough confusion about this to justify adding a sentence about recreation as a redirect to the language in CSD:G4.

Would appreciate more views on the subject. Thank you! Owen&times; &#9742;  21:02, 28 November 2023 (UTC)


 * OwenX, would you mind changing the bulleted list to a numbered list to make it easier to reference each option? Firefangledfeathers (talk / contribs) 21:04, 28 November 2023 (UTC)
 * Done. Thanks! Owen&times; &#9742;  21:11, 28 November 2023 (UTC)
 * This is not a valid RFC, not having a neutral statement of the question to be decided. I also note that the only confusion seems to be by the OP. Everyone else in the discussion can see that a redirect is in no way identical to an article. Phil Bridger (talk) 21:18, 28 November 2023 (UTC)


 * Option 2, sort of. We should keep G4 to just the cases of duplicate article creation. RfD is enough of a working process that there is an avenue available to those who believe the redirect should be deleted or retargeted. I think the difference between redirect as an AfD decision, which leads to the article history being preserved, and a post-deletion redirect creation is significant enough that someone doing the latter is not overturning the result. I don't think additional guidance is needed in the text of G4, and I'd rather just talk to any editors who tag, or admins that action, G4s that don't quite apply. Firefangledfeathers (talk / contribs) 21:22, 28 November 2023 (UTC)
 * So are you saying that after the AfD, we'll need an RfD for the same page to delete the redir? Wouldn't the same arguments for changing the original article to a redir still apply? I understand your point about the history. Do we need an AfD !vote along the lines of, "Do not recreate as a redir if deleted"?  I would think most who !vote "Delete" also do not wish to see the article instantly resurrected as a redir, even if they don't spell it out. Owen&times;  &#9742;  21:31, 28 November 2023 (UTC)
 * Yes, that is what I'm saying. My experience with AfD and RfD is that the arguments tend to be quite different. In the rare (as far as I'm aware) case where AfD participants really dig into the criteria for redirect deletion, the case for an eventual RfD would be pretty open-and-shut. Firefangledfeathers (talk / contribs) 21:38, 28 November 2023 (UTC)
 * I understand. Why do you object to adding clarifying words to G4, then? This doesn't seem intuitive at all. Owen&times; &#9742;  21:49, 28 November 2023 (UTC)
 * I could be persuaded, but I'm having trouble seeing this as a common issue. Have you looking into how frequently this comes up? Since my sense is that it's rare—rare for redirects to be created post-deletion, rare for such creations to be contentious, rarer still for them to be speedily deleted—I'd rather not complicate the language of the criterion. Firefangledfeathers (talk / contribs) 21:52, 28 November 2023 (UTC)
 * Very rare indeed, as you say. Most editors accept that AfD consensus is binding, and if consensus was against turning an article into a redir, we shouldn't turn it into a redir, including post-history-deletion. But if we accept your interpretation of G4, it means that there's no point in !voting "Redirect", as we can now just wait out the AfD, and show up the next day to recreate the page as a redir, and get a second shot through the RfD. I don't think keeping this option a secret is the best way to prevent it from becoming prevalent. If it's allowed, it should be spelled out as such, and if not, it should be explicitly prohibited. There seem to be enough people on the DRV I linked above who seem willing to use this option to bypass an AfD deletion, and I'm sure the popularity of the technique will spread quickly once the word is out. Owen&times; &#9742;  22:06, 28 November 2023 (UTC)
 * I have trouble seeing the situation so nefariously, but I've said about as much as I feel is wise, and I look forward to hearing from others. Firefangledfeathers (talk / contribs) 22:11, 28 November 2023 (UTC)
 * Oppose changes: This is entirely unnecessary. It excludes pages that are not substantially identical to the deleted version, and pages to which the reason for the deletion no longer applies. — A redirect replacing an article clearly fits this description as a redirect is substantially different from a page that was deleted at AfD. Hey man im josh (talk) 22:17, 28 November 2023 (UTC)
 * @OwenX: I've read through the DRV and the only person arguing that G4 applies to redirects for articles deleted at AfD is you. There was consensus there that G4 did not apply. Hey man im josh (talk) 22:42, 28 November 2023 (UTC)
 * So why keep it a secret? Why not spell it out in the wording of CSD:G4, and not waste people's time on AfDs debating redirect as an option, when they can simply show up a week later and recreate the page as a redirect anyway?  Beyond simply repeating what you've already said in that DRV, I don't see you addressing any of the policy issues in this RfC. Owen&times;  &#9742;  22:43, 28 November 2023 (UTC)
 * @OwenX: There is no secret, the wording is already clear to the point that you've been the only one advocating that G4 applies in this situation. Hey man im josh (talk) 22:50, 28 November 2023 (UTC)
 * "No secret"? 'Redirect' is one of the most commonly !voted choices on AfDs. Do you really think editors would be wasting their time arguing about it, if they know they could just show up the next day and recreate the deleted article as a redir? If this is what our policy is, let's make it abundantly clear, and stop wasting people's time on AfD discussions. Owen&times;  &#9742;  22:59, 28 November 2023 (UTC)
 * About there being no point in !voting 'redirect': When the outcome is 'delete', the page is deleted. A 'redirect' outcome as an alternative to deletion means that the page is not deleted . Deleted vs. not deleted. The point of a 'redirect' !vote and a 'redirect' outcome is that there is no need to execute the deletion in a technical sense because it's fine to preserve the history underneath the redirect, based on a premise that the content is not something that must not remain publicly accessible even in history (bad BLP content, etc.). Conversely, when the page is deleted and a redirect is created at that name, history is not preserved. —Alalch E. 23:01, 28 November 2023 (UTC)
 * That's like saying "why keep it a secret that G4 doesn't apply to Draft:X when X was deleted in an AfD". It's the same with a redirect, except it's even more obvious that it doesn't apply to a redirect which can't, categorically, be "suffuciently identical", and a draft can. These are different types of pages each with different deletion discussions. The type of deletion discussion must match the type of the page. I think that this is obvious to most people. —Alalch E. 22:55, 28 November 2023 (UTC)
 * But that's exactly what CSD:G4 does spell out! Read it and see: "It excludes pages in userspace and draftspace". If Draft:X was important enough to explicitly exclude, why not redirs? Owen&times;  &#9742;  23:01, 28 November 2023 (UTC)
 * Because redirects are not substantially identical to articles that have been deleted. Hey man im josh (talk) 23:08, 28 November 2023 (UTC)
 * It mentions drafts because people do sometimes get confused and think that draftspace content copied from a deleted page is G4 liable, because the content is identical which is the primary element of G4, and, more importantly, it extends G4 to respond to circumventing deletion to host content on Wikipedia. It doesn't mention redirects because it's harder to get confused when the content is not identical, or even similar at all (a redirect is not similar to an article), and you can't circumvent content being deleted by creating a redirect at the same name, as redirects have no content. —Alalch E. 23:11, 28 November 2023 (UTC)
 * Option 2. I already made a comment about this in a deletion review that led to this RfC—special:diff/1187380309.—Alalch E. 22:35, 28 November 2023 (UTC)
 * Option 2 (strongly). G4 specifically does not apply to articles that are not substantially identical to the deleted version.  It makes no mention of being substantially identical to a different version of an article discussed in the AFD (such as a redirect).  Any redirect can betaken to RFD, where the arguments tend to be quite different from those made at AFD.  Frank   Anchor  23:05, 28 November 2023 (UTC)
 * As others have said, G4 does not, and should not, apply. Would suggest withdrawing this RfC as it appears there's no chance it'll gain consensus (nor was much workshopping done here, which should've happened before opening this RfC). Elli (talk &#124; contribs) 23:44, 28 November 2023 (UTC)
 * Option 2 per everyone except the OP. A redirect is not an article and therefore cannot be substantially identical to one. Thryduulf (talk) 03:01, 29 November 2023 (UTC)
 * #2: G4 is about substanially identical recreations. For example, a non-notable article getting recreated a tenth or twentieth time, or an identical redirect already debated at RFD. It does not apply if the recreation addresses the reason it was deleted. Awesome Aasim 14:51, 30 November 2023 (UTC)
 *  Comment - I vote Option 2. --Jax 0677 (talk) 22:12, 1 December 2023 (UTC)
 * Option 2 is, and has been throughout recent memory, CSD policy. It's now 9-1 by my count, with no one else yet supporting your preferred option. Thank you for phrasing this as you did and putting it here, which is the precisely right place to get the best feedback. I suggest reading WP:1AM, even though this is a policy interpretation matter rather than an article content dispute. Jclemens (talk) 08:00, 3 December 2023 (UTC)
 * Option 3, reworded: “G4 applies to the recreation of a deleted page as a redirect, but only if that option was brought up, and was rejected in the close of the AfD”. And if in doubt, ask the deleting admin . I support G4 being useful in broadly supporting consensus at AfD for six months after the AfD. SmokeyJoe (talk) 08:32, 3 December 2023 (UTC)
 * To clarify, do you believe option 3 as reworded is supported by the current wording of G4 on this policy page, or are you suggesting that this policy page be changed to encompass this usage? Jclemens (talk) 21:50, 3 December 2023 (UTC)
 *  Reply - I could buy into something like this, but would advise saying "EXPLICITLY rejected". --Jax 0677 (talk) 22:39, 3 December 2023 (UTC)
 * Creating a redirect explicitly rejected at a recent xfd (imagining a G10 style reason) could get the creator blocked for disruption. SmokeyJoe (talk) 02:56, 4 December 2023 (UTC)
 * It could... but if the redirect is G10-able, then G4 needn't apply because G10 does. This interpretation of "the spirit of G4" (immediately below as I compose this) is not only at odds with how CSD criteria are applied (strictly) as well as NEWCSD #2, uncontestable Almost all pages that could be deleted using the criterion, should be deleted, according to consensus. In fact, the vast majority of redirects created after an AfD deletion are kept, not deleted. Jclemens (talk) 07:03, 4 December 2023 (UTC)
 * I believe that it is the spirit of G4, would not be surprised to see it happen, but I think it happens extremely rarely if ever. Probably not worth the fuss. SmokeyJoe (talk) 02:54, 4 December 2023 (UTC)
 * It’s not worth the extra words. Extra words would fail WP:NEWCSD#3 Frequent. Use RfD. SmokeyJoe (talk) 03:49, 4 December 2023 (UTC)
 * Option 2. If a redirect is useful, it should be created, period. BD2412  T 03:17, 4 December 2023 (UTC)
 * Indeed, but who gets to decide whether it is useful or not? The same can be said of an article, yet we still enforce consensus when it is deleted in an AfD. Why do you feel redirects should be handled differently when consensus was against creating one?  I don't think your terse response add any clarity. Owen&times;  &#9742;  13:18, 4 December 2023 (UTC)
 * Re: "who gets to decide whether it is useful or not?" The community decides. We have an entire process addressing that. BD2412  T 13:23, 4 December 2023 (UTC)
 * So are you saying that a consensus reached in an AfD against a redirect has no bearing on the subsequent recreation as a redirect, and we have to go through another round of XfD to remove that redirect? Aren't we putting process above practicality here? Owen&times;  &#9742;  13:42, 4 December 2023 (UTC)
 * I think that's a whole different discussion than what G4 is meant to cover. That moves outside the territory of non-controversial deletion. Hey man im josh (talk) 13:48, 4 December 2023 (UTC)
 * In very few cases do AfD discussions consider the merits or otherwise of a redirect, and even when they do very few participants express any opinions about the suggestion so they cannot be considered to have been the subject of a discussion as required by G4. There are occasional exceptions (e.g. Articles for deletion/Oakbrook Mall) but in the vast majority of those cases either the presence or absence of a redirect matches the consensus in the discussion. That leaves cases where a redirect is recommended but not created, which is not a matter for any sort of deletion, and redirects created contrary to consensus - which are so infrequent that CSD is completely inappropriate - especially as there may be alternative targets not considered by the AfD. This is not process for the sake of process, but process for the sake of getting the best outcome for the encyclopaedia. Thryduulf (talk) 16:20, 4 December 2023 (UTC)
 * Option 3 specifically talks about AfDs that discuss a redirect as an outcome. If you wish to ignore Option 3 because you believe it to be rare, you are free to do so. But rare or not, I don't see how ignoring consensus in an AfD that specifically discusses--and rejects--a redir solution helps get the best outcome for the encyclopedia. The two-phase solution you and others here seem to imply, of an AfD immediately followed by a RfD, is anything but cruftware. We specifically wrote G4 to avoid these unnecessary steps. Owen&times; &#9742;  16:38, 4 December 2023 (UTC)
 * We specifically wrote G4 to avoid these unnecessary steps? Without searching archives, I strongly doubt that.  G4 is primarily about deleting reposts.  Some keeps an offline copy, waits a week, then tries to quietly out it back. SmokeyJoe (talk) 03:04, 5 December 2023 (UTC)
 * I have done some wikiarchaeology, the very first version of what evolved into today's G4 dates from June 2003, it read (votes for deletion was the ancestor of all today's XfD processes). So, no, G4 was very much not written specifically to avoid AfD followed by RfD or anything similar. Thryduulf (talk) 05:26, 5 December 2023 (UTC)


 * The earliest mention I can find in the deletion policy is this revision from June 2003 where MartinHarper added
 * In this September 2003 edit to Deletion guidelines for administrators Cimon Avaro added "Candidates for speedy deletion" including as point 5: . That is identical to the then current version at Deletion policy
 * In the earliest version of Criteria for deletion (January 2004) the relevant criterion (at that time numbered 5) read making it clear the reasoning is for material that has been reposed after being deleted, not other pages created at the same title.
 * In April 2005 at Wikipedia talk:Criteria for speedy deletion/Archive 3 (the earliest discussion I have found) myself, SPUI, Dante Alighieri, Jondel and Furrykef expressed opinions, the loosest of them was that if the reasons for deletion are (potentially) addressed then it cannot be speedily deleted.
 * In July 2005 there was extensive discussion at Criteria for speedy deletion/Proposal/G4 about the wording and intent of the criterion. The explicit intent of the change was to clarify that and makes it clear that being "substantially identical" is the key aspect of the criterion.
 * In September 2005 Encephalon stated at Wikipedia talk:Criteria for speedy deletion/Archive 4 {{tpq|Why do we have it? In its fundamental, most essential form, the G4 rule simply means this: if an article was found unsuitable for WP in the past, and you then find an exact re-creation of it, you do not have to go through the whole rigmarole of formally discussing it again before deleting it once more.{{pb}G4 is recognition that the community has dealt with a problem before, come to a decision, and doesn't need to expend resources to reach that same decision every time the problem is recreated. If the article is substantially improved, however, or very different, then common sense dictates it can be considered anew in the usual fashion.}} (Italics in original).
 * At Wikipedia talk:Criteria for speedy deletion/Archive 31 UnitedStatesian asked a similar question regarding redirects created after AfDs were closed as delete (the earliest discussion specifically addressing this I've found). All discussion participants (Ilmari Karonen, Protonk, SoWhy, Dcoetzee and CharlotteWebb) agreed that these were not speedily deletable and did not feel there was a need to change this.
 * In December 2016 Nyttend stated at Wikipedia talk:Criteria for speedy deletion/Archive 60 that (emphasis in the original)
 * In September 2018 Wikipedia talk:Criteria for speedy deletion/Archive 72 (started by me and with only four other participants, (Tazerdadog, SoWhy Adam9007, and Nyttend came to the conclusion that


 * Thank you for the thorough wikiarchaeology! The September 2005 quote from sums up my thoughts nicely: G4 is recognition that the community has dealt with a problem before, come to a decision, and doesn't need to expend resources to reach that same decision every time the problem is recreated. Yes, the redirect is indeed very different from the article deleted, but it is not different from the option discussed--and rejected--in the AfD. If the community participating in an AfD considered the option of turning the article into a redirect, and consensus was against it, then we shouldn't expend resources having to deal with the same question all over again in RfD. This was the intent of G4 in 2005. Alas, as is often the case with old laws, the words gain Scripture status and are revered, and common sense gives way to blind observance, while the original intent is all but forgotten. Owen&times;  &#9742;  09:32, 5 December 2023 (UTC)
 * Except, when you read the actual context, that wasn't the intent in 2003, or 2004, 2005 or even 2022. The intent was and is to deal with content that had been discussed and reposted (almost) verbatim. As for your second comment the entire purpose of the RfD criteria is that they are interpreted per the literal meaning of the words, because they are the very limited exceptions to the deletion policy which states that everything must be discussed before being deleted. Thryduulf (talk) 12:51, 5 December 2023 (UTC)
 * Fantastic research Thryduulf, thank you for digging all of this up. I think it may be time for an uninvolved admin to put this RfC to rest because I think consensus was clear then and it's also clear now. Hey man im josh (talk) 13:03, 5 December 2023 (UTC)
 * I feel like you're putting words in Encephalon's mouth. Encephalon's intention was to avoid repeating the same debate over and over. If the option of a redir was discussed and rejected in an AfD, it's pretty clear from their wording that they would not wish to see the debate take place again in an RfD. Didn't we handle both articles and redirects in VfD back in 2005? Owen&times;  &#9742;  17:36, 5 December 2023 (UTC)
 * No. RfD was established in November 2003. Encephalon was not talking about redirects at all. Thryduulf (talk) Thryduulf (talk) 17:46, 5 December 2023 (UTC)
 * I'm interested in your reply to my last comment. You seem to imply that there's zero weight given to views about redirect on an AfD, since opposition to a redirect can only be handled by an RfD. Did I get that right? Owen&times; &#9742;  16:09, 4 December 2023 (UTC)
 * If you're talking about a situation where in the course of a well-attended AfD someone suggests, "redirect this title to Foo" and multiple other editors say "do not redirect to Foo, that would be a bad redirect" (as opposed to just voting "delete" without commenting on the redirect), and the closing admin finds that there was a policy-based consensus specifically against having a redirect, and an editor who participated in the discussion says something along the lines of "forget you guys, I'm going to make the redirect anyway", I suppose that might be a speedy case. However, literally anything outside of that should go to RfD, which is well-equipped to evaluate the propriety of redirects. BD2412  T 17:21, 4 December 2023 (UTC)
 * You've just described the AfD that triggered the DRV that prompted this RfC. The DRV has since been closed, without prejudice as to the validity of G4 in that case. However, the views expressed in that DRV suggest that the majority of editors disagree with you and me about the applicability of G4 in such a case. Does this happen often enough to justify an extra sentence in the wording of G4? I don't know, but I'd rather err on the side of over-verbosity than leave things in the air. Owen&times;  &#9742;  17:59, 4 December 2023 (UTC)
 * It doesn’t happen often enough. SmokeyJoe (talk) 03:05, 5 December 2023 (UTC)
 * The subject AFD most certainly did NOT have consensus against redirect. There was disagreement about a redirect target but not consensus in opposition to a merge/redirect in general.  However, the sole topic of this discussion is whether G4 covers a redirect when the deleted version is an article.  And many users including myself see it as a strong “no.”  Frank   Anchor  04:59, 5 December 2023 (UTC)
 * Option 2. Allowing G4 of redirects violates WP:NEWCSD: 1) It is not objective, because reasonable people can disagree whether an AfD actually supports a consensus against redirecting (if not explicitly stated by the closing admin). Meanwhile, requiring the closing admin to specify explicitly every single time seems like extra work for very little gain (given how rarely this situation arises). 2) It is not uncontestable, because AfD and RfD draw different crowds and it is not clear that a consensus formed at AfD would be sustained at RfD. Given the expertise of the RfD crowd, it is better to hold redirect-related discussions at RfD. 3) It is not frequent enough to be be worth including in the criteria in order to avoid an RfD. -- <b style="color:red">King of ♥</b><b style="color:red"> ♦</b><b style="color:black"> ♣</b><b style="color:black"> ♠</b> 18:15, 4 December 2023 (UTC)
 * Comment I've made a request at WP:ANRFC for an uninvolved editor to close this. Although it's only been open 6 days consensus is already very clear and a formal closure will be of benefit going forwards. Thryduulf (talk) 13:22, 5 December 2023 (UTC)
 * Comment: Very well, I'm not going to fight consensus here. If the scope of G4 does not extend to articles recreated as a redir, I won't use it as such. My issue at this point is one of informing editors participating in AfD. Many are debating, in good faith, the option of changing an article to a redir, not realizing they could simply do so themselves if the article gets deleted. I'm not a fan of secret rules. Shouldn't we publicize this so that less time is wasted on "Redir" !votes in AfDs? Right now, the guide for AfD makes it sound as if "Delete" and "Redirect" are two distinct, mutually exclusive options, when in reality any deleted article may be instantly recreated as a redirect. An extra sentence in the AfD guide could--and should--clarify the issue. Owen&times; &#9742;  18:27, 5 December 2023 (UTC)
 * It's not a secret rule, it's an obvious interpretation of the wording of G4. Redirect and delete are two distinct options, one of which would preserve page history while the other (assuming that the page hasn't been salted) would allow for the page to be re-created as either an article or redirect. Hey man im josh (talk) 18:49, 5 December 2023 (UTC)
 * Exactly this. Deletion does not prevent recreation unless the title is salted, and that's been the way things have worked on Wikipedia for at least 20 years and in that time, as far as I've found, only one other person has found the interaction between that, redirects created after deletion and G4 anything other than clear - and that other person didn't have any problem with it. There is no problem that needs solving, and speedy deletion would be the wrong wrong tool to solve it if it was a problem. Thryduulf (talk) 19:44, 5 December 2023 (UTC)
 * We've already moved on from the question of speedy deletion. The question before us now is whether we should add a sentence in the AfD guide to inform participants of the fact that if they don't care about the page history, they are free to recreate as a redir any article deleted in an AfD. It may be obvious to you, but realistically speaking, I doubt more than 10% of those !voting "Redirect" on AfDs know this totally-not-secret fact. Owen&times; &#9742;  20:47, 5 December 2023 (UTC)
 * No, it's unnecessary. There's nothing that stops a user from re-creating a page at the destination that was just deleted unless the page has been salted. Users are also free to immediately create a new article at the same location as well if they so wish (provided that article addresses the reasons that it was deleted). Hey man im josh (talk) 20:53, 5 December 2023 (UTC)
 * Right, and having them waste their time debating redirects on AfD is ever so much fun, because clearly they have nothing better to do, since they all know they can just recreate the page. Got it. Thank you.
 * My goal here is to make AfD discussions more efficient, not to sit back and go, "Well, it's not my problem they chose to waste their time." Owen&times; &#9742;  21:02, 5 December 2023 (UTC)
 * Not sure the pessimism is necessary. As mentioned, there's a very clear and distinct difference between deleting and reflecting a page. It doesn't change the course of an AfD discussion to further elaborate on this unless we start salting more articles. Hey man im josh (talk) 21:31, 5 December 2023 (UTC)