This results from a few different kinds of errors — people mentioning categories in talk page discussions without disabling them via a colon or the {{cl}} template, so that the page erroneously gets filed in the category instead of textlinking to the category in the discussion; people erroneously using Wikiproject tracking categories on the article instead of the talk page; people posting full-on copies of articles onto the talk page; etc. — but by and large, people who work on cleaning up categorization errors can only catch these in a scattershot manner by happening to come across them, and could use a comprehensive report similar to the ones that already exist for other types of misfiled content.
This would not need to look for "Draft talk" or "User talk" pages, as those are already caught by the existing draft and user reports — and it should skip listing categories that have been tagged with {{Polluted category}}, as those are typically hidden maintenance categories where mixing types of content is not seen as a problem. But as noted above, it would need to look for Wikiproject categories, because those aren't supposed to be on the reader-facing article.
I've asked for this before at Wikipedia talk:Database reports/Archive 8#Talk pages, but it didn't end up happening. But it seems relatively simple, as it would be coded pretty similarly to the reports that already exist — so would it be possible to get such a report created and put into use? Bearcat (talk) 16:19, 26 September 2025 (UTC)
Also, if possible, it should be formatted more similarly to the user report, which features links to incategory searches for "main" and "user" content in each listed category, rather than the draft report which just lists the categories and requires the cleaner to manually search for the misfiled content. Bearcat (talk) 16:21, 26 September 2025 (UTC)
The change from Wikipedia:Database reports/Polluted categories's query is trivial; the problem is that it's very, very slow, well over the ten-minute limit for SDZeroBot's {{database report}}, and there aren't any good ways to meaningfully improve it. So you'd have to get a "real" bot task to run it. One-time results should show up eventually at quarry:query/97494, unless it times out there, too, and it well might. —Cryptic 17:47, 26 September 2025 (UTC)
Alternately, would something like quarry:query/97507 be of use, or quarry:query/97508 for single categories? They don't make clickable links, and 97507 still takes too long to run with {{database report}}, but I'd be happy to teach you how to run it yourself. —Cryptic 03:30, 27 September 2025 (UTC)
97494 is definitely the most useful in its current state, though it's still hitting a lot of hidden maintenance categories that don't really need to be cleaned up (e.g. "Redirects from/to...", "Articles covered by WikiProject Wikify...", etc.), so there may be a few more rounds of refinement to exclude types of categories that don't actually need to be dealt with.
97507 mixes draft and user and talk content, and thus isn't as useful in that exact form, though it does have potential. I've already got access to an alternative categorized-drafts report that I can update at will (I was given it as a stopgap last year when the regular draft report broke and wasn't updating properly, but the primary maintainer of the bot that updates it was away on vacation and couldn't fix it, and its on-the-fly updatability has actually proven preferable to waiting for the regular report to update just once a week), so I wouldn't actually need it to look for drafts at all — but since the userspace report only runs once a week and often has hundreds of categories to check (quickly approaching or surpassing a thousand if I happen to forget about it just one time), a way to help catch categorized userspace content faster than once a week would definitely make it a more manageable and less time-sucking job, and it definitely helps to find talk pages — but I wouldn't really need it to look for drafts, since there's already another tool for that. And as with 97494, Category:All articles covered by WikiProject Wikify is a category we don't really need to worry about and should probably just be excluded. So maybe, instead of one report that looked for all types of polluting content in the same place, maybe separate jobs that just looked for one specific type of content each?
I already do an incat search of Category:Living people a few times a day, given the total non-viability of finding misfiled content in that massive megacategory any other way — so 97508 wouldn't really be needed in this context, though of course it may be useful for other purposes.
Seems strange that it would take so much longer to find talk page content than draft or user content. Is it just a matter of the raw number of pages being so massive that just finding them all takes too long (which I suspect might be the case as 97494 is landing on a round "precise multiple of 100" number of pages that suggests a capped finding limit rather than there being exactly that many pages on the dot), in which case a concerted effort to get the current situation under control might make a conventional report more viable in the future by virtue of reducing the number of pages there are to find, or is there just something about talk pages that makes it naturally take longer than user or draft pages no matter what?
Yeah, seems like showing me how to run 97494 and 97507 might be the way forward here. Bearcat (talk) 15:31, 27 September 2025 (UTC)
Specific categories can be omitted from each of these, given a list of names; or if there turn out to be a lot of them, either a pattern of names ("All categories beginning with 'Redirects from *'", for example) or something like "all hidden categories".97507 can be limited to only look for mainspace/talk, if you want, or can add another namespaces like Wikipedia:; my thought was that it would be easier than having to check separate reports for main+talk, main+user+usertalk, and so on. It takes the same (considerable) amount of time to check for them all at once as it does to check for just one.The reason it looks like it took longer to find just talk is that 97494 was deliberately set to find up to 1000 mixed categories, and 97507 just up to 200 - if you look at the sql in those, you'll see "LIMIT 1000" in the first and "LIMIT 200" in the second. There's no way to identify categories with members in mixed namespaces except to look at every category's contents and keep on going until it's ready to stop. Including not just talk but user, user talk, draft, and draft talk too found the first 100 in about four minutes, and the first 200 (as in query 97507's current version) in about fifteen. Looking just for talk, since there were only 700 such categories, it couldn't stop until it had looked at every category, which took half an hour. So if all of these were cleaned up, it would actually take longer to confirm that there weren't any mixed categories left, or just one or two or whatever, whether that means just main+talk or main+(long list of other namespaces).There are step-by-step instructions on how to register at Quarry and to fork (make your own copy of) and run an existing query here. It's been quite some time since I've registered myself, so the third bullet point may no longer be accurate, but it seems unlikely. Only other thing you'll have to change is to use one of these queries instead of 20320; 97508 is probably best to start with of these three, since it runs the fastest. Assuming no hiccups, I can walk you through how to change namespaces and omit specific category names, but user talk: is probably better for that than here. —Cryptic 16:34, 27 September 2025 (UTC)
Okay, cool. I won't be dealing with this right away — a friend's moving to a new apartment this afternoon, so I have to head out shortly to help him unpack stuff — but I'll advise when I get on it, probably most likely tomorrow, and hit you up if I need assistance. Bearcat (talk) 16:43, 27 September 2025 (UTC)
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
WP:CSD U5 kind of sucks. It's vaguely worded, and a yearslong history of its contents being deleted-by-script by a since-desysopped admin left most userspace patrollers under the impression that basically any page in userspace by a new user is eligible. A significant percentage of all nominations are of drafts, even though those are explicitly excluded from U5, they're often nonetheless deleted. Of what remains, a not inconsiderable portion is just userpages saying "I'm John Smith from New York and I work in finance", which are also exempt under the WP:UPYES clause, but likewise get deleted all the time. (Also, let's be honest, more often it's "I'm Ashok Kumar from Delhi and I work in IT"... there's a noticeable region-based bias in what gets tagged.) Links to people's profiles on other platforms, also generally allowed under UPYES and thus exempt, also deleted all the time.
Importantly, these aren't all mere procedural issues. We want people to use their userspace for drafting, and we want people to introduce themselves to us and make themselves feel at home in the editing community. WP:UP lays out these expectations, but U5, both as written and as enforced, doesn't honor them.
Based on discussion with a few others at User talk:Iridescent § U5 in the modern era, it seems to me that there are two reasons we want some kind of userspace cleanup mechanism:
Some people use their userspace to host things that are obviously completely unrelated to Wikipedia. This is things like creative works, persuasive essays, full-length résumés or CVs, a page about how great their cat is, etc.
Pages left idling in userspace indefinitely, while individually mostly harmless, add up to significant problems overall. I've deleted decade-plus-old userspace pages that accused living people of murder or rape. Or such pages can be used to passively promote some entity, although it's not clear how much benefit this actually brings someone who tries it.
What I'd like to brainstorm is a split of the criterion. Handling the second half seems like the easier part: something like
[CSD U6] User subpages of users who have never made a good-faith edit outside of userspace, which have not been edited by a human in at least six months. Promising drafts may be moved to draftspace as an alternative.
This would create a simple procedural mechanism, similar to G13.
This does leave a loophole for users who draft on their top-level userpage. Such drafts already confound userspace patrolling, and make up a considerable percentage of what I decline in CAT:U5. I created {{draftified userpage}} to semi-formalize an informal practice in dealing with them, but it remains, well, informal. So I would propose pairing the creation of U6 with removing the "usually" in "Work in progress or material that you may come back to in future (usually on subpages)" at WP:UPYES, with a footnote explaining that any user may move a draft from a top-level userpage to either a user subpage or to draftspace and replace it with {{draftified userpage}}, and that reverting such a move is an indication that the editor wishes for the page to be treated as userpage content rather than a draft, and thus subject to WP:UPNO.
Which brings us to the second half, basically some kind of less vague U5. This is where I'm less sure of what to do, and a big part of why I'm bringing this to VPIL. What I'm currently thinking is something like
[CSD U7] Pages in the userspace of users who have never made a good-faith edit outside of userspace, which are not formatted like an encyclopedia article, where all content on the page falls into one or more of these categories:
Creative or persuasive works unrelated to Wikipedia
Lengthy descriptions of any person's professional accomplishments
Lengthy content about the user's personal life
Links to websites that are primarily commercially-oriented
And where the issue could not be remedied merely by shortening the page or reverting to an earlier version, or where the user has resisted such attempts.
Re U7: Maybe "not intended as an encyclopedia article" rather than formatted? I can see both pros and cons - for example, it would eliminate excuses like "it doesn't have an infobox or references like a real article!" or "there's only one linefeed between paragraphs, so it all displays like one big conglomerated mass!"; and it does technically require mindreading. —Cryptic 17:00, 23 September 2025 (UTC)
"Contains no encyclopedic content" (because what if half the page is a paragraph that could be used in an article, but the rest is test edits)? WhatamIdoing (talk) 17:43, 23 September 2025 (UTC)
This seems far too misinterpretable too. We don't want to be able to say that "This page, which is formatted as and clearly intended to be an article, is about something no self-respecting encyclopedia should cover. Like the author." (See also my back-and-forth with CFA starting here.) I think the idea we should be shooting for is something like "written in an encyclopedic manner, or at least not in the first person". Which of course is no more usable as proposed wording. —Cryptic 01:19, 24 September 2025 (UTC)
I do really like this idea – having more specific speedy deletion criteria is always great, especially when splitting the more vague ones like U5 or G6 that often get misapplied. I also like the "no good-faith edit" criterion, although I do wonder how it would apply to, say, someone writing a promotional draft for their company (in draftspace) and following up with pasting a major CV on their user page.Regarding Cryptic's note, those are good points, but, in both cases, I'm expecting that experienced editors tagging/deleting these pages would have a much stricter view of what counts as an "encyclopedia article" than newcomers writing them. The first case I have in mind is a newcomer wanting to write an autobiography in good faith, but not at all familiar with our guidelines on the topic – where can we draw the line (in style or in intent) between this and lengthy content about the user's personal life? Chaotic Enby (talk · contribs) 17:06, 23 September 2025 (UTC)
Why does would a CV necessarily imply someone was paid? Thryduulf (talk) 17:43, 23 September 2025 (UTC)
If the Draft is for Bob's Big Business, Inc. and the CV says "I'm Bob, president and CEO of Bob's Big Business, Inc.", then how could it not constitute a paid-editing disclosure? (The hypothesized scenario is "someone writing a promotional draft for their company".) WhatamIdoing (talk) 17:44, 23 September 2025 (UTC)
In that case yes, but that's only a subset of CVs. If someone's CV indicates they're a retired mid-level employee at Bob's Big Business, a widget supplier based in Kentucky and they write a draft about a 20th century Russian chess player, that is not at all a paid contributor disclosure. Thryduulf (talk) 18:05, 23 September 2025 (UTC)
Yep, this is closer to what I had in mind – something not focused on simply disclosing their paid contribution, but instead discussing their personal accomplishments at great length and with a clear self-promotion intent. The issue that WhatamIdoing points out is still relevant, as a good-faith user might want to disclose all their potential COIs in advance, and it might be hard to decipher the difference in intent. Chaotic Enby (talk · contribs) 18:20, 23 September 2025 (UTC)
I don't think we should put too much effort into "deciphering intent". No matter what the intention was, if the objectionable CV demonstrates a COI, we should keep that. They can call it "building my personal brand" if they want to, but we should call it "COI disclosure". WhatamIdoing (talk) 21:55, 23 September 2025 (UTC)
True, especially if we're considering the bigger picture. We shouldn't assume that newcomers will be familiar in advance with our specific way of wording things, and we shouldn't fault them for doing something that also happens to be helpful from our point of view.That does raise the question of what to do with WP:NOTCV – currently, it only allows for "limited autobiographical information", but, as you point out, we don't have a meaningful distinction (besides wording) between being transparent about your COIs and posting your résumé. It does in a way loop back to the original U5 criterion – posting a résumé for the sake of posting a résumé, without actually contributing, isn't helpful, but it becomes invaluable if the editor also contributes on topics they might be connected to.However, what about the case of the editor who creates a "user profile", then waits a few days before finding a place to contribute, only to see their userpage deleted, and is now scared away from the project? Ultimately, the cost of hosting a few vanity résumés in userspace is probably less than what we lose by spooking away potential contributors, so maybe that aspect of U5/U7 should be deprecated entirely? Chaotic Enby (talk · contribs) 07:58, 24 September 2025 (UTC)
In my experience dealing with the CSD tags, people are far too expansive on their definition of "a CV" when they tag (and admins when they delete), and I'd consider that a bigger problem. Lots of people do make good-faith userpages that twig a typical patroller's idea of what a CV is, but which would be pretty useless as an actual CV/resume. They are deleted, bitily. -- asilvering (talk) 16:18, 24 September 2025 (UTC)
I do think the first sentence of Chaotic Enby's last paragraph is especially crucial—but I do want to point out that I think calling CSD deletions of potential good-faith editors (which is most everyone) simply bitey is insufficient, mainly due to how diluted that term has become. It's not just a bite to find out your very first attempts at editing are getting speedily-deleted, it's a public humiliation in the form of an alligator devouring you while spectators watch. It's also a permanent scarlet letter that, without context, screams "this user didn't perfectly understand how Wikipedia worked" which can significantly influence how people approach initial reactions with that user later on (that is, it'll usually make people reach for the grab-bag of uw-templates instead of, Jimbo forbid, starting a conversation with them). Perryprog (talk) 16:28, 24 September 2025 (UTC)
Indeed. -- asilvering (talk) 16:48, 24 September 2025 (UTC)
Yes, this. It's really easy to forget or understimate the new user experience when you're as established as we all are here.I sometimes edit non-WMF wikis; I typically start with a dozen or so typo fixes or spelling errors, then make a user talk page saying something to the effect of "(Hi, I understand how MediaWiki works, no need for a welcome template or whatever.)" and redirect my userpage to that with a summary like "Not much to say here, but I don't like being a redlink". The popups for "Congrats, you've just made your first/tenth edit, keep it up!" are always surprisingly encouraging, even though you'd think I'd be used to them by now. But about half of the time, my first interaction with another human user is an admin thanking me for my spelling fixes and then deleting my user and user talk pages with a log entry like "not necessary". Those wikis I never edit again. —Cryptic 17:09, 24 September 2025 (UTC)
That last point you mention is, in my opinion, one of the most important reasons for us to have strictly defined CSDs, with some amount of supervision through pages like WP:Database reports/Possibly out-of-process deletions, instead of the ability to delete pages for any reason. That way, we're shifting which parameters control these new user experiences. It isn't admin temperament, which is impossible to control at a large scale, even if we have selection processes like RfA. Instead, the parameters are community-defined criteria, for which we can discuss the specific interactions they might have with editor experience. Making these criteria more specific, and reducing the grey area in their wording, can only help avoid these kinds of bad experiences. Chaotic Enby (talk · contribs) 17:25, 24 September 2025 (UTC)
And maybe adding a reminder about MFD directly in overused criteria (e.g., "For ordinary amounts of self-promotion, use MFD instead of U5")? WhatamIdoing (talk) 22:08, 24 September 2025 (UTC)
I mean, I would argue to just not waste your time at all if it isn't that flagrantly blatant. Perryprog (talk) 22:17, 24 September 2025 (UTC)
Yes, but you probably don't go looking through the User: namespace in the hope of doing Important™ Work as an apprentice Defender of the Wiki whose goals are truth, justice, and tidiness. But if someone believed that this dreadful page's existence sullies the wiki, then they might stop abusing CSD if we remind them that MFD exists, and that deleted via MFD is every bit as deleted as any other process. WhatamIdoing (talk) 22:55, 24 September 2025 (UTC)
Do we need an explanation somewhere about what a CV looks like? That we're looking for a bare list of bullet points with schools, job titles, and dates, and that we're not looking for a paragraph of prose? WhatamIdoing (talk) 17:18, 24 September 2025 (UTC)
My first thought is that there needs to be some time frame in which someone can demonstrate that they are here in good faith before U7 would apply. We don't want to bite future good editors who start by writing a detailed userpage setting out their credentials under the impression that this is required before going on to edit constructively elsewhere. How long should this be? I don't know, but an absolute minimum of 7 days, maybe 4 weeks would be better? Thryduulf (talk) 17:42, 23 September 2025 (UTC)
That's the range that I'm thinking about. WhatamIdoing (talk) 17:46, 23 September 2025 (UTC)
I wonder whether the problem is "CSD taggers do the wrong thing" or "this CSD criteria doesn't do what we want".
If CSD taggers are regularly doing the wrong thing, we could set a bot to watch for obvious, objective violations (e.g., any five unreverted edits outside userspace), revert it, and leave a note for the tagger saying "BTW, that's not what this CSD is for. If you hate it, use MFD instead".
We could also re-write the U5 criteria to be clearer. I think the current wording (where the owner has made few or no edits outside of user pages, except for plausible drafts and pages adhering to Wikipedia:User pages § What may I have in my user pages?) might leave editors divided on whether this can apply to someone who has made (say) six edits to a plausible but declined draft, plus two more that are promotional stuff on the User: page, or if the six edits to a draft exempt them from U5 permanently. WhatamIdoing (talk) 17:58, 23 September 2025 (UTC)
The bot is an interesting idea and I might run it, but we'd have to consider carefully where to draw the bot's line to avoid too many false-positive reverts while still being useful. And probably give humans a parameter for {{db-u5}} and similar templates to tell the bot "a human has carefully checked this, don't remove" for when FPs slip through anyway. Anomie⚔ 18:21, 23 September 2025 (UTC)
To be fair, CSD tags are reviewed by an administrator making the decision to delete or not, so we should already expect tags to be double-checked. A bot checking for a numerical requirement might be too strict, but maybe we could give more explicit guidance to deleting admins in the CSD tag's wording? Chaotic Enby (talk · contribs) 18:22, 23 September 2025 (UTC)
Yes, but part of the complaint is that admins are being overly generous with their acceptance of U5 tags.
The bot could flag the CSD tag rather than reverting it. A message like "A bot has determined this is likely an incorrect use of the tag. U5 is only acceptable for people who have made 'few or no' constructive edits outside the User: space, and the creator of this page has at least 23 non-reverted edits." WhatamIdoing (talk) 21:57, 23 September 2025 (UTC)
Flagging the tag could be a much better way to go at it, and reminds me of how we have similar bots at WP:PERM for example. Chaotic Enby (talk · contribs) 22:12, 23 September 2025 (UTC)
Flagging could well be done too, I like that idea. The bot could add a parameter to the template (maybe |non-reverted-non-userspace-edits=42, suggestions for a better name welcome) and then the template adjusts its messaging accordingly. But what should the the lower limit? Should the bot add even |non-reverted-non-userspace-edits=0 so admins can differentiate "bot checked this and it's fine" versus "bot hasn't checked this one yet"? Should the bot update the parameter if the count changes, and if so when? 0↔1 sure, 1000↔1001 probably not, but where's the line?Or would it be better to go in a different direction, have the bot update a subtemplate with a {{#switch}} or JSON giving the edit counts for users with active U5 (or U6 or U7) tags, which the template can use to fetch the count? With that we could have the bot update the counts frequently without having to worry about flooding the history of the flagged pages. Anomie⚔ 13:05, 24 September 2025 (UTC)
I dislike flagging the =0 or =1 cases because it's too easy for those counts to change. I'd probably flag only for accounts with ≥5 unreverted edits. Less than that, the default message could be a reminder to admins that this only applies to accounts with few or no unreverted edits, and that they'll have to check manually. WhatamIdoing (talk) 17:28, 24 September 2025 (UTC)
I worry about creating an incentive for someone to see a user with a number of edits close to the threshold and hunt for edits of theirs to revert. That would clearly be gaming the system, but it would not be easy to detect and I'm not sure it's something we can (should?) expect admins patrolling the speedy deletion category to check for? Thryduulf (talk) 19:03, 24 September 2025 (UTC)
While this is starting to be a bit too elaborate, checking in which context each edit was reverted seems natural, especially as reversions might themselves be mistaken. Beyond that, while I believe that it is good to have a numerical benchmark to avoid out-of-process deletions, this incentive you mention could be avoided if it was a flexible range? For example:
... where the owner has made few or no edits outside of user pages. While there is no set number, a user with more than 5 unreverted edits will usually be exempt from this criterion. Below that, edits should be reviewed individually to make an informed decision.
That way, we have both a safeguard against CSD abuse, and a level of flexibility is retained (cf. Wikipedia:What U5 is not). The original discussion that led to U5 being enacted focused explicitly on SPAs as the source of the problem, and many users with only a handful of contributions are visibly not SPAs.An alternative possibility, which has the advantage of providing a better-defined boundary, would be:
... where the owner has made no edits unrelated to the tagged material outside of user pages.
This excludes occasional users with only a handful of (all constructive) edits, while including, say, a spammer promoting their product in both userspace and draftspace. Chaotic Enby (talk · contribs) 19:46, 24 September 2025 (UTC)
The hole that that leaves is someone who creates two distinct instances of tagged material: say, an autobiography on their userpage and a startup business on a subpage, without an explicit link between the two; and edits outside userspace exclusively related to one or the other, but never both. (WT:Speedy deletion/Archive 84#Quantifying "few or no other edits" in U5 is still fairly recent, and on-point to the first part of your comment.) —Cryptic 20:10, 24 September 2025 (UTC)
The reverse is what I'm concerned about: "Hi, I'm a neurologist who specializes in autism" with someone saying that good edits to articles about Autism are "related to the tagged material" and therefore don't count. WhatamIdoing (talk) 20:15, 24 September 2025 (UTC)
This wouldn't supersede the main U5 criterion requiring the material to be a WP:NOTWEBHOST issue, which a brief introduction isn't. Beyond that, I think the "relatedness" criterion should be defined more broadly. If you write a user page about your research in autism, editing about autism in general wouldn't be an issue (in fact, we welcome subject-matter experts!), but only editing about your own research and self-citing it would put that same user page under a different light. Chaotic Enby (talk · contribs) 20:49, 24 September 2025 (UTC)
I'm pretty convinced that both taggers are doing the wrong thing and that the wording is insufficient. I mentioned in the contextual ANI Clovermoss linked below that labeling this criterion as "Non-contributor's misuse of Wikipedia as a web host" was a terrible idea, and I've only become more convinced of that since. Most of that title has nothing to do with the criterion itself, and most taggers don't look at anything but the title. And, also in that contextual ANI, I made the point that a fair number of the pages that get deleted as U5 that don't qualify for it (or any other criterion) should still be speedies. —Cryptic 01:19, 24 September 2025 (UTC)
@Cryptic, got any ideas for a better wording for the label? -- asilvering (talk) 02:31, 24 September 2025 (UTC)
I think this idea from Cryptic has some long-term potential. Maybe ==Fan fics, job applications, and similar content with no attempt to contribute to Wikipedia==? My idea is to name some of the most extreme examples, in the hope that people will realize that this is about the "oh, my goodness, you are hopelessly on the wrong website" situation and not the "eh, I think that's a bit too self-promotional for my own subjective, culture-b(i)ased tastes" one. WhatamIdoing (talk) 22:15, 24 September 2025 (UTC)
Separately, the idea that "except for plausible drafts and pages adhering to WP:UPYES" could be read as an exception to the "few edits", instead of as an exception to what can be deleted, occurred to me too recently, during this deletion review (which is relevant to this discussion though not this particular point, and is worth a read-through). If I were truly confident that the original meaning of "any page plausibly intended as a draft can't be deleted" still had broad consensus, instead of "any draft that could plausibly be promoted to an article" or this nightmarish interpretation of "graduating from userspace and editing drafts too still doesn't get you out of newbie U5 vulnerability", I'd have already moved the clauses around. —Cryptic 01:19, 24 September 2025 (UTC)
I note users who have never made a good-faith edit outside of userspace leaves a gigantic loophole: a single typofix in a non-userspace page or a comment on a talk page outside of userspace would enough to invalidate the deletion. Even good-faith creation of a redirect that is deleted would invalidate it.I also note that the example a page about how great their cat is wouldn't fit under any of the things in the proposed U7, unless you stretch "the user's personal life" a bit. userspace pages that accused living people of murder or rape wouldn't fit under anything there either, unless you consider murder and rape "descriptions of any person's professional accomplishments" (and there are enough alleged murders and rapes to be "lengthly"), or maybe if the user is alleging that they were the one murdered or raped (in which case "the user's personal life" clause might fit, again if it's "lengthly" 🙁). used to passively promote some entity also wouldn't fit anything there if the entity isn't a person and they avoid it being just links to commercial sites. There are a lot of other things at WP:NOTWEBHOST and WP:UPNO that wouldn't fit the proposed U7 either, such as non-commercial linkfarms, non-Wikipedia discussion that's not "creative or persuasive works" (e.g. using a userpage as a forum or a dating site), playing games, 'publishing' original research, uploading files (e.g. base64-encoded text so they don't have to be in the file namespace), keeping enemy lists, descriptions of the personal life of people other than the user themself, etc. But maybe all that is ok to have to go to MFD? 🤷 Anomie⚔ 18:07, 23 September 2025 (UTC)
That's a good point - how much of this stuff (both what's proposed to be included and what Anomie highlights as potential additions) goes to MfD at the moment? Is there enough of this to meet the frequency requirement? Thryduulf (talk) 18:15, 23 September 2025 (UTC)
Would userspace pages that accused living people of murder or rape already fall under WP:G10? Chaotic Enby (talk · contribs) 18:18, 23 September 2025 (UTC)
If unsourced, yes, but if they included sources for each accusation (and avoided it being "legal threats" or "intended purely to harass or intimidate") then I think it wouldn't qualify. I don't know what exactly Tamzin was referring to when mentioning I've deleted decade-plus-old userspace pages that accused living people of murder or rape above. Anomie⚔ 18:27, 23 September 2025 (UTC)
To be clear, my point about such pages is that they show why it's bad to let user subpages fester indefinitely. Sure, they may be deletable under one CSD for another, but only if someone looks for them, like I have by searching phrases like "is a rapist" or "is a murderer". So this isn't really about my U7 proposal but my U6. My idea is for most of all this to fall under U6, even a lot of stuff that could technically be U7, much like G13 sweeps up a decent number of pages that might be G1/G2/G3 or even G11. -- Tamzin[cetacean needed] (they|xe|🤷) 19:01, 23 September 2025 (UTC)
The question isn't whether we should let these pages "fester". The question is only whether the route to deletion needs to be an undiscussed CSD nomination, rather than one of the other deletion options. WhatamIdoing (talk) 22:00, 23 September 2025 (UTC)
If the page is "my cat is the greatest cat ever", that's fine. If the page is many paragraphs on how great the cat is, that's lengthy content about their personal life; actually as I was drafting the above I flirted with something like "or entities they are personally familiar with such as their friends and pets", and I'd be fine including that if it's clearer. If the page is many paragraphs and reads like an actual biography of the cat, it's a draft and can be handled under the proposed U6, or MfD in the less likely event that it's associated with someone who's actually contributed to the encyclopedia. -- Tamzin[cetacean needed] (they|xe|🤷) 19:07, 23 September 2025 (UTC)
As I said, it depends how far you want to stretch "the user's personal life". Having seen just how pedantic some people at WT:CSD can be about the definitions, I'd be inclined to write for the narrowest interpretation possible. Anomie⚔ 19:51, 23 September 2025 (UTC)
We already have WP:Database reports/Possibly out-of-process deletions#U5, which I've been reporting on for a while but never had the will to do anything about largely for the same reason Tamzin started this discussion (U5 being a mess of conflicting interpretations). Likewise, I've almost always been ignoring U5 requests when I do speedy deletion patrolling, for the same reason. * Pppery *it has begun... 18:43, 23 September 2025 (UTC)
There's a lot to reply to. I'll start with changes to userpage policy (WP:UPYES)
1. Change Work in progress or material that you may come back to in future (usually on subpages) from usually on subpages to always on subpages.
2. Append (always on subpages) to User space archives.
3. Change Experimentation (usually on subpages) by changing usually to always.
And reword the rest of the policy as needed to be consistent with these changes.
Having work in progress on the main userpage is confusing to other editors who are looking for info about the editor in question and confusing to casual wiki users (I've seen those pages misrepresented as encyclopedic information). There's also no compelling reason to have work in progress on the main userpage. — rsjaffe🗣️ 18:46, 23 September 2025 (UTC)
A second change to policy is to always properly template pages in userspace that look like Wikipedia pages. There's several templates that should be ok, including {{Userspace draft}}, {{User page}}, {{User sandbox}}, etc. The purpose of this is to avoid misunderstanding a userspace page as a Wikipedia page. This change alone would markedly increase what I would tolerate in userspace. — rsjaffe🗣️ 19:56, 23 September 2025 (UTC)
And other people could add templates to the page if the author fails to do so (e.g., a newbie who doesn't know about this). — rsjaffe🗣️ 19:57, 23 September 2025 (UTC)
I'd have your proposed U7 replace the current U5. It needs some tweaking--e.g., I'd change the no real edits provision to be something like "fewer than 10 non-userspace edits" (or something like this). That is, a real contributor. Also, I'd change the not formatted like a Wikipedia article to something like "not formatted like a Wikipedia article or not suitable to be a Wikipedia article". — rsjaffe🗣️ 20:02, 23 September 2025 (UTC)
"Not suitable to be a Wikipedia article" could be clarified a bit, as otherwise it might be too strictly interpreted as "not suitable for mainspace in its current state", which is not ideal. Chaotic Enby (talk · contribs) 20:10, 23 September 2025 (UTC)
I agree. I'm just struggling with wording there. — rsjaffe🗣️ 20:16, 23 September 2025 (UTC)
With the change to what's allowable on the main userpage (which is a bit more restrictive than what you proposed), I strongly agree with adding something allowing others to move a nascent article on the main userpage to a better location and to use the {{Draftified userpage}} template. Leaving a hard redirect on the main userpage is user-hostile, as it confuses visitors to the page (who may not notice they've been redirected) and prevents a naive user from properly utilizing the userpage as they probably won't figure out how to remove the redirect.
By the way, the preponderance of user reactions to moving the "article" from the main userpage has been gratitude, with I think one negative reaction (moving the page back). Many of the people starting a page there are just doing it because they don't know the right place to write it.
I even move somewhat inappropriate autobiographies from the main userpage to draftspace. This gives them a longer amount of time and more feedback to provide the case for inappropriateness rather than a lightning strike kill of the article. I think people tend to be more accepting when they receive feedback from multiple people over a range of time rather than a single no. It does waste some time of the reviewers, though. — rsjaffe🗣️ 20:20, 23 September 2025 (UTC)
Really bad stuff gets moved to a user subpage or deleted, though. — rsjaffe🗣️ 20:22, 23 September 2025 (UTC)
That's really good to know! I was wondering about that and I do think it's worth considering, but I still contend that it doesn't really matter if it's something that's living as a userpage or as a subpage. It really just feels like a pointless distinction in my mind. Like, yes, it's not the "intent" of userpages but... so? If it's someone who sticks around for any length of time they will likely figure out on their own that they can have it as a subpage or in draftspace instead, and I feel like that is arguably preferable to having a weird process (and what will appear to newer users as them having violated some unspoken rule) being taken on their work for no clear reason. Perryprog (talk) 20:25, 23 September 2025 (UTC)
Why do you move it from User:Example to Draft:Example (to be deleted in six months) instead of moving it to User:Example/sandbox (no time limit)? Are you intentionally trying to get more of them deleted during the next year? Because your actions are actually resulting in more of them being deleted during the next year than if you left them alone or moved them to a user sandbox page. WhatamIdoing (talk) 22:05, 23 September 2025 (UTC)
Ah, I did somewhat misread this—I had assumed it was moving to the user's sandbox or similar. In my opinion the only time that we should be moving to draftspace is once a page is submitted for AfC review, and it doesn't pass any quick-fail criteria. (I'd rather quick-fail it in their userspace rather than the place that starts a ticking clock on their work.) Perryprog (talk) 22:09, 23 September 2025 (UTC)
Such an overhaul is long overdue - U5 is widely misunderstood and widely misapplied, and hasn't proven sufficient to handle the problems that could arise with userspace pages. I like U7 as written (it should be U5 though, if we're overhauling that?) and I would caution against introducing language like "non-encyclopedic content" or "not intended as an encyclopedia article", because not infrequently userspace personal info is people hosting material they want in a mainspace biography but that shouldn't be in a mainspace biography. I'm less certain about U6 - it strikes me as too narrow. Why does making a single typo-fix in mainspace allow you to incubate drafts indefinitely? We ought to tie it to editor activity and page activity. I'm less concerned about attack-page material - we have means of deleting those - and more about the leavings of young editors writing semi-anonymous bios that don't serve an encyclopedic purpose and could plausibly cause harm later. Vanamonde93 (talk) 18:54, 23 September 2025 (UTC)
Numbering isn't a huge concern for me, but I provisionally went with calling it U7 more for the practical reason that it'd require patrollers to go read the new rule after they see that {{db-u5}} throws an error, rather than be looped in piecemeal to the narrowing of the criterion. But again, not a huge deal either way. -- Tamzin[cetacean needed] (they|xe|🤷) 19:04, 23 September 2025 (UTC)
Addendum: "Lengthy descriptions of any person's professional accomplishments" is a necessary addition, IMO, because really what we're trying to address is the use of userspace as an end-run around WP:NOTCV. That said, we're going to have to have a discussion about what "lengthy" means, because we have tolerated personal accomplishment tallies from experienced editors to a degree that we never would in a newbie. See, for instance, User:Marine_69-71, and the varied feelings expressed at ARBCOM over their use of userspace. Vanamonde93 (talk) 19:07, 23 September 2025 (UTC)
So you want to ban "lengthy descriptions", because you don't want people to have an opportunity to express their varied feelings about specific instances? WhatamIdoing (talk) 22:11, 23 September 2025 (UTC)
I'm saying we need to work out as a community what quantum of personal biography is permissible in userspace - how are you getting what you said from what I wrote? Vanamonde93 (talk) 16:52, 25 September 2025 (UTC)
I'm not sure I understood your comment. You think it's necessary to ban "lengthy descriptions". Okay, why do you think that? Specifically, why would a rationale person think that length descriptions need to be deletable under CSD? The obvious answer is: if we don't delete them via CSD, then "we're going to have to have a discussion" at MFD, and I interpreted this as as you not wanting to have those discussions at MFD (i.e., the main opportunity for editors to express their views of each alleged NOTCV instance). WhatamIdoing (talk) 17:07, 25 September 2025 (UTC)
Okay, I have a whole lot of thoughts on this topic (which isn't really in response to the proposed criteria), but first some context—I tend to watch the page creation log for new users that aren't in mainspace which has led to me seeing a lot of the really weird stuff that goes on in userspace especially, and I've also seen lots and lots of, in my opinion, bad U5s (both nominations and deletions).
The first thing I want to point out is how inherently user-hostile Wikipedia is for newcomers, especially those who are trying to figure out how to write prose and not just typo fixes. Even with the visual editor, it is an immensely confusing and unfortunately unforgiving (in terms of how easy it is to get "yelled" at if you do a small thing wrong) environment. I think this leads to a large number of people figuring that their best place to learn is by clicking the bright red button that's their name (which says "Your user page" when you hover over it) and to start trying to figure things out. They probably don't know what to try writing about, so they write about whatever. Or, they assume it's the place to put information about yourself because... well, it's "your user page"; that's what you do on any other website that has accounts.
This will then very often lead to U5s, G11s, blanking if they (heavens forbid) wrote something on their talk page, and I don't think many editors realize just how honestly intimidating (and honestly terrifying for younger editors) getting CSD'd is. Your "talk page" (whatever that is) is suddenly slammed with a page and a half of text of contradictory text where you're both being welcomed and then yelled at with a scary red danger icon that you're fundamentally violating the purpose of Wikipedia?? Welp, time to log off never to return.
Anyway all that is to say my point is this—we should care way, way, way, way less about what people do in their userspace, as long as it isn't actively detrimental to the project or to any other person (or is blatantly G11, which I don't think is misapplied too often, though there are times I disagree with its use.) No beginner editor is going to know what the fudge a "subpage" or a "draft" is, and I think carving out exceptions based on stuff like that is a bit iffy due to how there's no way most new editors will have any natural understanding of how to go about accessing their "subpages" or "drafts", as well as understanding that they need to do that to avoid getting wall-of-blue-texted out of oblivion. Perryprog (talk) 19:57, 23 September 2025 (UTC)
Yes, this is basically my thoughts on it as well. I'm not sure the issue is with WP:U5 itself so much as that it's being constantly misapplied and we're not effectively preventing those misapplications. -- asilvering (talk) 20:18, 23 September 2025 (UTC)
The bright red button since Vector2022 is worth emphasising. It is a prominent red link that is by itself in the UI. It is also one of the three tabs at the top of the Special:Homepage (the user talkpage is here as a redlink too). The design pushes new editors towards their user page, and I'm not sure what guidance is made obvious alongside this. CMD (talk) 04:45, 25 September 2025 (UTC)
@Chipmunkdavis: The echo notification an editor receives for their first edit includes a link to Help:Getting started. There's a lot of links in there, but not anything I can see that directly mentions userpages. Clovermoss🍀(talk) 07:48, 25 September 2025 (UTC)
Maybe we could have an automatic editnotice explaining (in simple words) what userpages are/aren't for? Chaotic Enby (talk · contribs) 08:55, 25 September 2025 (UTC)
Could that appear only for redlinked userpages? Seems sensible if so. As for Help:Getting started, we probably need to look at that for other reasons as well. It doesn't seem to cover the newcomer homepage the WMF has created. CMD (talk) 10:14, 25 September 2025 (UTC)
Could that appear only for redlinked userpages? Yep, that's what I had in mind, assuming that is technically possible! Chaotic Enby (talk · contribs) 11:01, 25 September 2025 (UTC)
After looking a bit deeper into it, we already have {{Base userpage editnotice}} that appears by default, but it mostly focuses on not drafting articles there, and the tone might be a bit too aggressive for new users – maybe a revamp is in order?I'm thinking a more newcomer-friendly version could be in the style of a Wikipedia:Dos and don'ts page, with clear, simple bullet points supported by visual cues and links. Chaotic Enby (talk · contribs) 11:27, 25 September 2025 (UTC)
We should rework it, but also when does it appear? It does not seem to appear when I enter the editing screen to create a userpage. CMD (talk) 12:13, 25 September 2025 (UTC)
Probably doesn't help that it seems to have been broken, the {{#if:<includeonly>{{PAGEID}}</includeonly>|| check it had at the top would never output the content since {{PAGEID}} produces 0 rather than an empty string on a nonexisting page. I don't know when that may have changed, as presumably it used to work. I've fixed that now. Anomie⚔ 15:31, 25 September 2025 (UTC)
The context behind and a yearslong history of its contents being deleted-by-script by a since-desysopped admin left most userspace patrollers under the impression that basically any page in userspace by a new user is eligible matters. Wikipedia:Administrators' noticeboard/IncidentArchive1171#Fastily was only ten months ago. Not everyone saw a problem with admins deleting pages outside the written U5 criteria. I think that if we want to change anything, we need to create an environment where challenging other admins is less toxic. The rules don't matter if people don't follow them. If people are deleting things that aren't U5s under U5, I don't see why creating a new criteria will fix that. Clovermoss🍀(talk) 20:25, 23 September 2025 (UTC)
Since WP:Database reports/Possibly out-of-process deletions exists, I think this environment change can and should be extended to non-admins asking admins for details about deletions they believe to be out-of-process (or directly challenging them, although this might not always be possible without seeing the contents of the deleted page). Chaotic Enby (talk · contribs) 20:56, 23 September 2025 (UTC)
Some scattered thoughts here.
One of the key underlying principles is that we want to discourage WikiSquatting.
However, discouraging hosting is a difficult task, even if a new system managed to catch 100% of pages that were placed here for hosting there would still be all the preexisting advice floating around the internet that would continue to direct people here for years.
Community practice in this area has not been entirely consistent, but it's been established ever since Wanli that people who abuse the project as a free host can have their content deleted and be blocked if they persist.
There's mountains of coprolite sitting in userspace but vanishingly few cases where the persistence of anything there has been an issue.
Nonetheless many people simply do not like rubbish littering userspace, and none of it is worth getting into disputes over.
Unlike G13, the proposed U6 will require increased human review when a user has edits that need to be assessed for good-faith. It's not much at an individual level but in aggregate the work will add up. To what degree I'm not certain.
The proposed U7 is easily evaded by simply formatting the stuff you want to store as an article and a script to do that could be easily implemented not really that big a deal given the low-impact on everything else userspace has, but it should be considered.
Any new CSD if applied retroactively has the potential to overwhelm the CSD process. Even something narrow that covers say only unsourced BLPs over two years old would make many thousands of pages eligible for speedy deletion.
Retroactive application of the proposed U6 is not practical for the cases of users having edits because no one is going to review that. Even for the users that have no edits retroactive application is only achievable with a bot.
Could not be remedied merely by shortening the page ensures the proposed U7 could only ever be correctly applied after someone reverts multiple attempts to blank a qualifying page because blanking will almost always remedy the issue and in the rare case it doesn't the page probably already qualifies for G10 or oversighting. So if that is the intent the CSD should be reworded in a more straightforward manner, if it isn't the wording should be rethought.
Simplifying CSD and making them more objective is helpful to everyone working in the area and this idea should be pursued. 184.152.65.118 (talk) 04:18, 25 September 2025 (UTC)
Wow, this has gotten so much great feedback! I want to step back and distill some of what's been discussed above. People seem on board with the ideas of a procedural G13-like criterion and a friendly ban on top-level-userpage drafting. The main issues raised above seem to be:
The requirement of no good-faith edits outside userspace is subjective.
It is. But the only objective criteria we could use here would be something like "fewer than X edits to namespaces A, B, and C", which won't necessarily map onto what we want it to map onto. Meanwhile, the current definition is also subjective but is broader. So my proposed solution here is just to include a footnote defining "good-faith edit" very broadly.
The requirement of no good-faith edits outside userspace can be easily gamed.
See next.
A bot can't always determine if the criteria are met.
This and #2 somewhat cancel out. If we change no edits to "few if any", then a bot can handle the no-edits contexts, where more nuanced cases (e.g. a single typo fix, or 20 edits but they were all reverted as tests) can be handled by humans. Someone could put together database reports like "User subpages of users with 1 to 20 mainspace edits" or "User subpages of users whose only mainspace edits have been reverted".
Six months with no edits wouldn't cover the case of someone who returns once every month or several to add a bit more to their off-topic userspace subpage.
Actually I don't know if anyone said this, but I thought it. See next.
Penalizing off-topic content is inherently BITEÿ, and while we can all think of things that are clearly completely off-topic, patrolling editors will naturally tend to have stricter definitions than a good-faith new user.
This is the big one, and it's the one I've been thinking the most about. What really hit me is something from IP 184's comment... It's not like we need most U5 content to be truly deleted. If something is bad in a way that needs to be admin-eyes-only, it'll fall under one of the G-series criteria. While, say, someone's Kirk/Spock fanfic has no business being on Wikipedia, it's not doing any harm sitting in a page history, and it's not doing any harm on a user subpage that'll soon be deleted. So what I've come to feel is that there doesn't need to be any direct successor to U5. What I think we can do instead, which will also address #4, is allow U6 to be a bit toothier for things that don't appear to be drafts: make it 6 months from creation, not from most recent edit. Because there's still the time requirement, this subcriterion doesn't need to be ultra-precise in the manner of a U7/rump-U5 like we were discussing, but can still be clearer than the current U5 wording. Meanwhile, for top-level userpage content, we can just allow off-topic content to be blanked under similar rules if it violates WP:UPNOT.
Thus my revised proposal is this:
CSD U5: Deleted
CSD U6:
User subpages[a] of users who have made few if any good-faith edits[b] outside of userspace, which either:
have not been edited by a human in at least six months or
were created more than six months ago, could not be interpreted as a draft article (even a very bad one), and unambiguously violate the "Excessive unrelated content" or "Advocacy or support of grossly improper behaviors with no project benefit" sections of WP:UPNOT[c]
Promising drafts may be moved to draftspace by any editor as an alternative to deletion.
For users with few if any good-faith edits outside userspace, off-topic subpages more than six months old may be deleted under speedy deletion criterion U6b; top-level userpages that would be eligible for deletion under that subcriterion if they were subpages may be blanked by any editor.
Work in progress or material that you may come back to in future (usually on subpages)
Drafts, especially where you want discussion or other users' opinions first, for example because of conflict of interest or major proposed changes
Drafts being written in your own user space because the target page itself is protected, and notes and working material for articles (Some content may not be kept indefinitely).
Drafting on a top-level userpage is confusing, and any editor may move a draft away from a top-level userpage, either to a subpage or to the draft namespace, replacing it with {{draftified userpage}}. If an editor reverts such a move, they indicate that their userpage should not be viewed as a draft for the purposes of assessing compliance with this policy.
Notes
^See WP:UPNOT regarding off-topic content on top-level userpages and WP:UPYES regarding drafts on top-level userpages.
^A "good-faith edit" here is defined as one that was not obviously vandalism, promotion, testing, harassment, or edit-warring. An edit does not need to be constructive or even competent to qualify as good-faith.
^Note that merely writing about one's personal or professional life is not a violation of WP:UPNOT unless inappropriate or excessive or extensive[ly] self-promotional.
Continuing on the logic that we don't need most U5 content to be deleted, I'm wondering whether U6a is truly needed here. The discussion that led to G13 being enacted was less than unanimous, with the two main arguments in favor of deleting abandoned drafts being the lack of promise/potential, and the risk of objectionable and WP:NOTWEBHOST content lingering around. The first is less of an issue given the wider latitude given in userspace (in fact, some supporters of G13 explicitly opposed an equivalent userspace criterion), while the second will already be covered by U6b, so an additional process to delete userpages of low-edit count users seems both redundant and WP:BITEy.I very much agree with the rest of the proposal, and especially with the change to WP:UPYES. One additional thing I would like to see would be a default editnotice that editors would see when creating their userpage, explaining in simple words what can and can't be added. Chaotic Enby (talk · contribs) 09:22, 25 September 2025 (UTC)
I don't think this could work without U6a. A big part of reducing the amount of BITE is reducing overtly confrontational interactions over userspace content. Quiet procedural deletion both means that the user is less likely to feel like they've done something wrong, and that there's less of a reason for people to go poking around in userspace ragpicking, since most of the stuff will get cleaned up on its own, and manual tagging will only be needed for edge cases. I think G13 has been very effective at avoiding over-patrolling of draftspace in the same way. If we didn't have it, we'd need to figure out what to do about drafts that are probably-net-negative for Wikipedia to host, but nonetheless don't fall under any G-series criterion... whereas with G13, when someone says "Hey this draft is horrible, what do I do?" we just tell them "Chill out, it'll be gone in 6 months". -- Tamzin[cetacean needed] (they|xe|🤷) 09:29, 25 September 2025 (UTC)
That's actually a very good point I didn't consider. I was mostly afraid that it could hurt editors taking their time with drafts – say, joining Wikipedia, starting some user sandbox, forgetting about it, and deciding to come back the next year only to see their project was deleted. But I figure the last clause ([p]romising drafts may be moved to draftspace by any editor as an alternative to deletion, which seems to only apply to U6a) will help with that?I wonder if tagging them as {{promising draft}} and leaving them indefinitely in userspace could have the same effect, as I don't see it interfering much with the anti-ragpicking effect of G13/U6a.Sorry for the nitpicking again, your proposal is certainly an improvement on the current situation, just trying to figure out the details before it goes for a full RfC. Chaotic Enby (talk · contribs) 09:40, 25 September 2025 (UTC)
No, this is idea lab, nitpicking is good! Yes, there's some space for undesirable deletions here, although it's pretty narrow: The editor would have to 1) start a worthwhile draft, 2) make ~no edits outside userspace, 3) go 6 months without editing that draft, 4) not have the tagger or deleting admin decide to draftify, and 5) come back and want to resume it. This will happen occasionally I'm sure, but probably much less often than the comparable situation happens under G13. And when it does happen, there's still a straightforward process to reverse it. (I imagine we'd just add U6 to the WP:REFUND/G13 scope.) We could make this even less likely by adding courtesy deletion notices at the 5-month mark, something I have no strong feelings on in either direction.As to how to handle these drafts, I like moving to draftspace better, on the premise that the creator probably isn't coming back, and it's better to have the content at a title where someone is more likely to notice it and adopt it. But I'm not strongly opposed either to some kind of "U6-exempt" tagging system. -- Tamzin[cetacean needed] (they|xe|🤷) 10:24, 25 September 2025 (UTC)
Thanks for the feedback on my feedback! My only worry about moving them to draftspace is that someone else might just notice them there and delete them under G13, as it would technically have been six months without edits (except maybe if moving a page there resets the counter?) Chaotic Enby (talk · contribs) 11:03, 25 September 2025 (UTC)
I believe moving would reset the timer. We could make that explicit through a footnote if needed. -- Tamzin[cetacean needed] (they|xe|🤷) 12:04, 25 September 2025 (UTC)
Continuing on the logic that we don't need most U5 content to be deleted, why don't we ...just not delete it? Like, not have U5 or anything else, and tell people if they really dislike it, they can simply blank it instead of trying to make it admin-eyes-only. If it needs to be deleted, then there are other available alternatives (e.g., {{db-attack}} or Wikipedia:Miscellany for deletion). WhatamIdoing (talk) 14:13, 25 September 2025 (UTC)
It's a little wild to me how abstract this discussion is when we have lots of examples to look at. Let's choose one from Wikipedia:Database reports/Potential U5s/5, for example User:Sussybaka228. Is anyone arguing that we need to keep pages like this? Would it be helpful to start a miscellany for discussion thread about this page to re-establish the very clear consensus that we both do not want garbage pages like this and that having a seven-day community discussion about each and every garbage page that some user creates is unnecessary? We have speedy deletion criteria to avoid wasting the editing community's time.
What about a page such as User:முனைவர். ஆ.சம்பத்குமார்? We've typically held that on the English Wikipedia, where we've intentionally siloed pages by language, pages like this should be deleted. Is anyone disagreeing here? If you'd rather we use MFD for this, we can. --MZMcBride (talk) 21:11, 25 September 2025 (UTC)
Just noting for the record that the example related to non-English language has been deleted, as it is an unambiguous copyvio. (A quick google translate revealed the copyright statement at the end of the text.) Not commenting on the general concepts here. Risker (talk) 21:25, 25 September 2025 (UTC)
I looked at entries 2000-2009 in the database link and none indicated a need for deletion under a userspace criterion:
User:AuburnAsh - student project page (courtesy ping to Stevenarntson noted as the teacher). Shouldn't be deleted. The "Article evaluation" section would ideally be dated and on a subpage but it's doing no harm where it is
User:Jessica gacus - a short extract from the I Have A Dream article as it stood in April 2022 and a link to the speech on YouTube. Shouldn't be deleted.
User:Sainhimanshu01 - A short bio in Hindi that could also funciton as a COI declaration if it were in English
user:Ngabitsinze - a short bio that could also function as a COI declaration
User:Ridaashrafia - This is not a userpage, but I see no reason why blanking wouldn't be sufficient but others would speedy it under G11
User:Gaming Euan - I think this is python code, it should be looked at by someone who knows python. If it's harmless it should at most be moved to a subpage, if it's intentionally harmful then G3, if it's accidentally harmful then MfD would suffice.
User:G-man1535 - another student in the cass run by Stevenarntson. Shouldn't be deleted.
The Python code is a wrapper for the enwiki API, seemingly written by someone who doesn't entirely understand how the API works (or maybe for a school assignment that had specific constraints). If run, it would be in violation of WMF best practices because it spoofs a user agent, so I guess there's technically a slight harm that way, but probably not enough to make MfD a good use of community resources -- Tamzin[cetacean needed] (they|xe|🤷) 09:18, 26 September 2025 (UTC)
User:Ridaashrafia is very clearly both a copyright violation and spam. What are we doing here. --MZMcBride (talk) 03:13, 29 September 2025 (UTC)
I didn't think to check for copyright violation, but the linked website is not loading for me and the copyright report linked in the G12 template returns 0% so I'm not deleting it myself. Thryduulf (talk) 03:25, 29 September 2025 (UTC)
I'm glad those userpages for my students don't seem problematic. I do want to follow up on the thought that some contents should be dated, or on subpages? Just curious what that might ideally look like. My only hesitation around modifications is that my students have pretty low digital literacy. Sometimes things that might seem easy are a heavy lift. Feel free to drop replies onto my talk page if it's off topic for what's being discussed here. Thank you! Stevenarntson (talk) 17:27, 26 September 2025 (UTC)
Sometimes, if you see some notes on what needs fixed in an article, it's handy to also see a signature/date stamp, so you can tell whether those comments are likely to be relevant. Obviously, all this information is also available in the page history to anyone who thinks to look for it.
I don't think it's a big deal, and I don't think that anyone else thinks it's a big deal, but students are used to putting their name and the date on their class assignments, so if you wanted to do this optional thing, you might be able to tell them something like "Anyone can edit, right? So you might read the article and write your evaluation of it on Octember 32nd, and the next morning, someone might change the article completely. Your evaluation might look 'wrong' if we don't know when you did it, because we might compare your evaluation against the new version of the article instead of the version you were reading. So when you write your evaluation, put the date on it." WhatamIdoing (talk) 19:24, 26 September 2025 (UTC)
We typically expect users here to contribute to the encyclopedia. Wikipedia is a free and open source encyclopedia for users to contribute to. We have user pages for encyclopedia editors to tell other users about themselves. In the case of User:G-man1535, this user has not made any edits to Wikipedia itself in over three years. Is there a reason we should indefinitely host their user page? There are so many sites where that would be happy to have content like this such as Facebook or LinkedIn. I'm struggling to see why Wikipedia is a suitable choice. --MZMcBride (talk) 03:18, 29 September 2025 (UTC)
According to WP:User pages "limited autobiographical content", "significant editing disclosures", "notes related to your Wikipedia work and activities" and "comments on Wikipedia matters" are all explicitly permitted on userpages. Everything on G-man1535's page clearly falls within one or more of those categories. Why do you think that is unsuitable for Wikipedia? What benefits would deletion bring to the encyclopaedia? Thryduulf (talk) 03:35, 29 September 2025 (UTC)
You're missing that user pages are for Wikipedia editors. If someone registers an account here and posts only their résumé on their user page, they are not an editor, they are confused and should be posting to LinkedIn or some other site instead. If you want to assume bad faith, you could say they're a spammer, which many new users are.
It is incredibly well established over decades that Wikipedia receives a lot of spam and other garbage. One of the editing community's roles is combating spam, vandalism, copyright violations, impersonation attempts, hoaxes, fake articles, etc.
Have you looked at a user page on a mobile device? It's incredibly difficult to distinguish a random User:Foo page and a real article. Many, many editors are creating user pages thinking that they are creating articles here. It's a very bad situation. --MZMcBride (talk) 06:55, 29 September 2025 (UTC)
Perhaps this is a more realistic question. If I wiped out everything on my userpage except for the second-last sentence (" I am fond of cheap, trashy spy novels and just about any form of chocolate."), would it have anything to do with Wikipedia? Would it still be allowed or would it be subject to deletion? Risker (talk) 21:46, 25 September 2025 (UTC)
You have over 28,000 edits. Compare with Special:Contributions/Sussybaka228. These are user pages of users with 0 edits outside of the User or User talk namespaces on this wiki. Any kind of "established" user, even one who just vandalizes an article, isn't getting reported on in this case. :-)
I will reiterate that we can use MFD but we determined it was a waste of time to have a community discussion for content like this: User:Sussybaka228. We can always go back to that. Should I nominate this page for deletion through the protracted process? --MZMcBride (talk) 03:05, 29 September 2025 (UTC)
If the community thinks its a waste of time to delete such pages, why do you think that it is worth the time finding them, assessing them and nominating them for deletion, speedily or otherwise? What benefits do you think it brings to the encyclopaedia? Thryduulf (talk) 03:23, 29 September 2025 (UTC)
You're misunderstanding. The community determined that it was a waste of time to have a weeklong discussion about these pages and so we established a process for speedily deleting these pages because it was so obvious that they were unwanted. The main idea is to prevent spam, copyright violations, and other garbage that comes along with being a wiki while also not wasting more editors' time reviewing the pages and voting to delete the obvious cases. That's what speedy deletion is, I think you already know this?
As I said above, there are plenty of sites that would be happy to host a résumé or other user-generated content. Wikipedia is not the place for that type of content, as has been very clearly established. --MZMcBride (talk) 06:50, 29 September 2025 (UTC)
Copyvios and spam don't need U5 to exist, because they ought to be deleted as {{db-copyvio}} or {{db-spam}} instead.
Your userpage contains only a silly joke. Why should we keep yours, but not Sussybaka's? Why should we impose an unequal system, in which "we" get to do things that we prohibit "them" from doing? Do you believe that a system in which some editors are more equal than others is consistent with our values?
I've no worries about the hapless mobile user. The User: namespace doesn't appear to be indexed, so they won't find it at Google, and until you mentioned it here, the page had been viewed a total of 11 times since its creation several years ago.
You ask whether we should spend a week discussing it. The options are:
Spend a week discussing it at MFD, potentially involving half a dozen editors' time.
Tag it for U5, potentially involving two editors' time.
Blank the page, using one editor's time (10 seconds?).
Do nothing, using exactly zero editors' time.
If you want the method that is most efficient, I suggest the one that uses nobody's time at all. WhatamIdoing (talk) 07:43, 29 September 2025 (UTC)
Man oh man this place sucks. Earlier today I tried to log in from my phone and got blocked by a stupid "you need to check your e-mail" prompt. I said okay, I'll just reply without logging in and then the IP address I was on was blocked from editing of course. Please, tell me more about treating users equally.
Wikipedia very obviously treats different users differently, as does every society and community. That's how trust and reputation work. You're posting as "WhatamIdoing" and you've built a reputation here and people trust what you write as a result. You also have additional user groups here based on your good standing and contributions to the encyclopedia.
We're discussing users with no contributions to the encyclopedia. They are not Wikipedia editors, they are registered users who have made user pages and have not yet contributed to building an encyclopedia. It's a distraction to suggest my user page or Risker's or your user page or anyone else commenting here's user pages are relevant. If any of these users make even a bad edit to an article, we consider them an editor. We do not consider people posting their résumé or hoax articles or other spam to be editors.
Many community discussions over years and years resulted in deletion of user pages exactly like this, which is why a speedy deletion criterion was created for this purpose.
While you may not be interested in picking up trash at the park or beach, that doesn't mean it's a waste of time. Putting aside the trite Animal Farm reference, it's rude to tell a volunteer how to spend their time. Nobody's asking you to participate in helping keep Wikipedia tidy, but doing so, in the real world and in the digital world, has a lot of benefits. --MZMcBride (talk) 01:23, 30 September 2025 (UTC)
What benefit does the world receive by deleting the user page for G-man1535 (the example you gave above), where that editor has written an evaluation of the problems with a specific Wikipedia article? How does the world benefit by us removing valid criticism of Wikipedia (e.g., that a Wikipedia article was unreferenced) from that user page?
On the flip side, what benefit does the world receive by us keeping your user page? WhatamIdoing (talk) 02:01, 30 September 2025 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
In parallel to the CSD discussion, I've been working on an editnotice to guide users creating their userpage about what should or shouldn't go there. Basically, a newcomer-friendly version of Wikipedia:Userpages, summed up in a few bullet points. It shows up as two columns ("do"s and "don't"s) at full width, but they stack up to still be readable in the more vertical format used by Visual Editor (or for mobile users).
There is already an existing editnotice that shows up for those pages, but it is harsher-looking and doesn't do a great job at summing up what user pages are actually intended for.
I'm sure this isn't a new idea but I can't find it. I believe en-Wikipedia is moving over to temporary guest accounts instead of IP editing. The first time someone edits without an account, they will be assigned an account (it will be a collection of numbers and tilde symbols), which becomes their username for 90 days, whenever they edit from the particular device they were using when they started. After 90 days, the account is automatically closed, and I assume if they edit on day 91, they're assigned a new guest account.
I was wondering, is there any plan to provide temporary guest account users an option to "upgrade" their guest account to become a real account, retaining its editing history? I could imagine a lot of people might start with a guest account, but find during their 90 days that they're enjoying editing productively. It'd be nice to let them simply carry on, rather than kicking them back to square one, and losing the connection between what they did previously and their new permanent activities. Elemimele (talk) 16:12, 26 September 2025 (UTC)
Based on my shaky understanding of the technical things involved here, I think the answer is probably no, since I think temp accounts use a different framework and you wouldn't be able to transfer edit history to a new username like when you get your account renamed. Cremastra (talk·contribs) 02:38, 27 September 2025 (UTC)
Thank you! I've asked it there, in case it can be made technically possible, but I'm sure you know a lot more about the technical basis than me (there is probably no one who knows less...) Elemimele (talk) 14:47, 27 September 2025 (UTC)
At a quick glance, it seems like prior to T333223 it might just have needed a rename and setting of authentication data (e.g. a password); now it would also need changing that flag in every wiki where the account exists too. OTOH, looking through T300273, it seems unlikely they'll ever actually implement it over concern that the temporary account for a shared public computer might have been used by multiple people and that non-CheckUsers might have saved the IP of the temporary account (and so now know an IP used by the registered account). Anomie⚔ 15:18, 27 September 2025 (UTC)
Unless things have changed radically in the last couple of years, the answer is "Legal said no". WhatamIdoing (talk) 18:13, 27 September 2025 (UTC)
Despite the technical issues making a direct "upgrade" impossible, a newly created account could link their previous temporary accounts on their userpage, if they want to keep some continuity in their editing history. I wonder if we could offer this by default to temporary account users creating a "real" account? Chaotic Enby (talk · contribs) 13:39, 28 September 2025 (UTC)
The core problem with "upgrade" is that temporary accounts are associated with devices, not people. — xaosfluxTalk 18:17, 28 September 2025 (UTC)
I checked at WMF and it definitely isn't possible and isn't going to happen, but xaosflux, that's absolutely true. It does also illustrate the futility of all this. At the moment if someone turns up editing as 123.1.2.3, creates an article about Helge Hexenhuthersteller (an expert on German hedgehog ecology), likes editing and registers an account as GermanHedgehogFan and continues to edit extensively on German hedgehogs and the works of Dr Hexenhuthersteller, it doesn't take much deduction to link the account GermanHedgehogFan to 123.1.2.3. When we get the nice new temporary accounts, if someone has managed to associate the temporary account with an IP as described above, when the temporary account holder registers as GermanHedgehogFan and continues to write about hedgehogs, we're in exactly the same situation. If, in any way, it's possible to link temporary accounts to IP addresses, there is surely no point in having the temporary accounts? Elemimele (talk) 16:41, 30 September 2025 (UTC)
The primary benefit of temporary accounts is privacy enhancement, as we don't publish the underlying network addresses to the revision metadata. — xaosfluxTalk 18:04, 30 September 2025 (UTC)
I propose a form of Captcha that could be added called Cloudflare, that when someone makes a large numbers of edits in a short span of time, the user would have to verify if they're human to make sure there not a bot. 2601:981:4401:1CC0:E92B:B306:3DDE:4055 (talk) 23:33, 23 September 2025 (UTC)
Much content on the web succumbs to regional restrictions, namely US sources not available in Europe. When I open https://www.chicagotribune.com/ from a European URL, I just get "This content is not available in your region". Template:Cite web should contain a parameter warning readers of such restrictions, ideally also name the regions affected. Would this be technically possible? --KnightMove (talk) 09:25, 29 September 2025 (UTC)
It is not all that you want, but you could use |url-access=limited to denote this: see WP:URLACCESS. Cheers, SunloungerFrog (talk) 09:38, 29 September 2025 (UTC)
Indeed that solves at least part of the problem. What if "regional" were introduced as a 4th parameter? This sounds like a much easier, if not complete, solution. --KnightMove (talk) 10:13, 29 September 2025 (UTC)
Well... does that case even exist? But in that case, I would give priority to subscription. --KnightMove (talk) 02:41, 30 September 2025 (UTC)
Is it really useful to know that a site is regionally restricted, if you don't know which region it's restricted to? WhatamIdoing (talk) 21:38, 29 September 2025 (UTC)
I would say yes. It's warning the reader to prepare for the reader for the option not to be able to read the source. --KnightMove (talk) 02:41, 30 September 2025 (UTC)
I'd say no, and think that limited would suffice in these cases, to indicate that readers may not be able to access that particular source.On an additional parameter to specify the regions, I think it's extremely unlikely that editors would be able to populate information into a {{Cite}} template parameter that is at all complete about the regions that a particular website chooses not to serve, and it would be terrifically difficult to keep this up to date. Cheers, SunloungerFrog (talk) 15:50, 30 September 2025 (UTC)
So be it, thanks. --KnightMove (talk) 14:23, 2 October 2025 (UTC)
For the 4 years I have been editing here, I have always found it weird how these warning templates say "this is your only warning", personally I have always thought that "this is your last warning, or "this is your final warning" made more sense, because basically I have thought that since this template is supposed to indicate a final warning for whatever offense or rule they broke. For example uw-vandalism4im should say. "This your final warning, if you vandalize Wikipedia again, you may be blocked from editing without further notice". This is just my opinion, and doesn't have to be done. 98.235.155.81 (talk) 16:18, 5 October 2025 (UTC)
This is just a proposal, because I can't edit the templates myself as I am an IP. 98.235.155.81 (talk) 17:11, 5 October 2025 (UTC)
The intent of the "4im" warning templates is that they're to be used for an issue that's so blatant that we want to give only one warning. If it's not that bad, start with one like {{uw-vandalism1}} or {{uw-vandalism2}}, then {{uw-vandalism3}}, then {{uw-vandalism4}} (no "im") which does not say "only warning". Anomie⚔ 00:16, 6 October 2025 (UTC)
Ok, thanks for explaining about the use of the templates. 98.235.155.81 (talk) 10:44, 6 October 2025 (UTC)
I'm coming here to reconsider how to solve (what I perceive) a problem with sidebars so filled with content and links that they are essentially navboxes. Attendants to the previous proposal discussion have given alternative proposals or (mostly) outright rejected that this is a problem.
So as I see it, this first question should be answered before trying a reform or a proposal:
Is there a problem with excessively long sidebar templates, such that they end up resembling vertical navboxes?
If there is a consensus of NO, then we can forget any proposed reform.
If there is a consensus of YES, then I think the next question has to be answered:
How do we deal with this problem?
My answer to this question is that we should only allow side-bars which essentially serve as an ordered reading list of articles, such that the reader can gain a broader context of the subject by starting at the first article, continuing through the intermediate articles, and finishing with the final article. Others have proposed that we delete sidebars entirely, and I think this is too extreme because sidebars can still be very useful for this purpose of guiding the reader directly. ―Howard • 🌽33 13:19, 11 September 2025 (UTC)
I'm going to guess that you've never heard of Wikipedia:Getting to Philosophy, a sort of "game" we talked about back in the old days. The idea was, if you didn't use the Search tool, how many articles did you have to click through, to get to the article you actually wanted (in the game, the target article was Philosophy).
One of the purposes of general sidebars is to help people to get where they want to go. They do this by typically providing links to more general articles (e.g., you are reading Si̍t-chûn Movement, and you want to get to Philosophy, which is correctly not linked in that article per WP:OL).
Removing these sidebars makes it harder for people to get to the articles that they want to read.
I think that the scenarios you should be considering are:
Some individual sidebars are too long. These should use collapsing features or be edited to remove the some items. (This will probably result in creating more sidebars, as templates get split rather than half-blanked.)
You personally dislike them. We can give you some .css code so that you won't see any of them (as long as you're logged in).
I agree with this but I don't see how GettingToPhilosophy has anything to do with anything. Aaron Liu (talk) 21:49, 13 September 2025 (UTC)
If you are at article #1, but you want to end up at article #2 (for example, Philosophy), then a sidebar can help you get there faster. WhatamIdoing (talk) 18:30, 14 September 2025 (UTC)
I have never supported deprecating side-bars altogether. I think this is an extreme position. I have only tried to figure out a way to slim down sidebars to prevent them from being excessive and navbar-like. ―Howard • 🌽33 10:13, 14 September 2025 (UTC)
Support deprecating sidebars altogether. Sapphaline (talk) 17:58, 13 September 2025 (UTC)
This is not a vote. Please present valid reasoning for this position which I disagree with. ―Howard • 🌽33 10:14, 14 September 2025 (UTC)
Sidebars are overwhelming and semantically suck. Sapphaline (talk) 16:33, 14 September 2025 (UTC)
Oppose - Is it the length of some sidebars that raises the concern here? If so, isn't the solution something like {{Sidebar with collapsible lists}}? (— 𝐬𝐝𝐒𝐝𝐬 — - talk) 13:12, 14 September 2025 (UTC)
I'd like to see more data. I usually ignore sidebars and infoboxes as a reader. As an editor, they annoy me because they make image placement harder. I don't see a difference between the two types of sidebars the OP is talking about. But I'd like to hear from sidebar users what they get out of them. —Kusma (talk) 13:41, 14 September 2025 (UTC)
IMO, sidebars are fine when they obey WP:SIDEBAR (they should only contain key, directly related, main and major items), and WP:BIDIRECTIONAL. Most do not. I try to enforce them, though. PARAKANYAA (talk) 21:19, 14 September 2025 (UTC)
Actually, you know what, they’re such a waste of time and attract so little positive use that I think it would be better to get rid of them entirely. Almost no sidebars are done properly. PARAKANYAA (talk) 22:07, 15 September 2025 (UTC)
I am coming around to this too. They are meant to collect linked articles, implying they apply to the topic of a whole article, but they are discouraged in the lead. They are meant to be BIDIRECTIONAL, but enforcing this would conflict with infoboxes and images that might fit on some of the included articles. There are some tightly knit topics where I've seen what I think is technically a sidebar which work, such as derivations of Template:Campaignbox, however the common implementation of an infobox imitator with an image and fancy presentation seems to take up valuable space that could be used for more curated linking (eg. an image with a caption which includes links to the most relevant items). CMD (talk) 12:43, 16 September 2025 (UTC)
Wait, what, they're discouraged in leads? Anyways, I like sidebars. I've never used on as an editor, but as a reader I often use ones like Template:LGBTQ sidebar or Template:Environmental law to jump between articles. Like ITN, I guess. Can't stand to work there, regularly read it. GreenLipstickLesbian💌🦋 12:55, 16 September 2025 (UTC)
See WP:LEADSIDEBAR. (Advocating inclusion in a section is not how I read the discussion, where by my reading proponents sought inclusion in the lead but below an infobox, but the text has stood for a few years now.) CMD (talk) 13:24, 16 September 2025 (UTC)
They are almost never used in accordance with guidelines; technically they aren't even supposed to be included in the leads at all. My biggest gripe is that they take up so much space and mostly include items that don't link back per WP:BIDIRECTIONAL, or even WP:SIDEBAR. And if you try to make it compliant with the guidelines, it is like pulling teeth, against a tide of people who constantly add terrible additions on pages nobody besides you watches. Navboxes sometimes have bidirectional problems too, but since they're not massive eyesores that take up a quarter of the screen or create horrible alignment issues on short articles it is much, much, much easier to fix. I have seen like 3 sidebars that comply with the guidelines. PARAKANYAA (talk) 03:02, 26 September 2025 (UTC)
I agree that efforts to reform navigation templates broadly, not just sidebars, is badly needed because navigation is a prerequisite to coordination on Wikipedia due to the limited participation in the project. I believe there should be a greater number of recommended best practices for their structure and design and that such best practices should be converted into explicit required standards to limit their use to navigation because at present many navigation templates are effectively are decorative and just contribute to clutter in articles. -- CommonKnowledgeCreator (talk) 18:25, 6 October 2025 (UTC)
Recently I've been fixing a lot of pages with Bare URLs, but soon I realised that almost 80% of the Bare URLs are actually links to PDF files, which neither Citation Bot nor reFillα can fix.
Isn't there any Bot/easier way to fix them except for manually opening each and every PDF to find their titles? Cdr. Erwin Smith (talk) 08:55, 26 September 2025 (UTC)
I think the problem is that PDFs can be structured in a bewildering number of ways, so it is difficult to code a simple script that would parse even a significant minority of PDFs to accurately determine the title. It is an interesting problem to try and solve, though, and your question made me look at Grobid and wonder whether something like that might be helpful. Cheers, SunloungerFrog (talk) 09:47, 26 September 2025 (UTC)
Ideally we shouldn't be referencing things to PDFs very much at all, as PDFs very often represent primary sources rather than reliable or WP:GNG-worthy sourcing. There will, of course, be some exceptions, but those would be in the minority — more often a PDF source would need to be removed from the article entirely rather than having its bare-URL formatting fixed. Bearcat (talk) 15:56, 26 September 2025 (UTC)
"Primary" is not a synonym for "unreliable". And once WP:GNG is satisfied by some sources, it has nothing to say about the rest of the sources used. Even if a PDF really is a primary source, it may be perfectly fine if the statements are supported by it without a need for analysis or synthesis (and, for that matter, a secondary source might be insufficient if it doesn't support the statements without analysis or synthesis). Anomie⚔ 20:21, 26 September 2025 (UTC)
I didn't say primary was synonymous with unreliable. However, primary is synonymous with non-GNG-building, and virtually the only acceptable use of primary sourcing, under any circumstances, is for basic information — such as a statement of where the person was born or where the company's head office is located — that isn't a notability claim. Bearcat (talk) 15:42, 4 October 2025 (UTC)
Reliable primary sources are acceptable for any statement of fact not about a third party living person. Thryduulf (talk) 15:50, 4 October 2025 (UTC)
You said PDFs very often represent primary sources rather than reliable [...] sourcing, which sure sounds like you're saying that primary sources are never reliable to me. And as I said that you ignored, "non-GNG-building" and WP:N in general is irrelevant once GNG is satisfied by other sources in the article.Personally I wish Wikipedia would drop the "primary" versus "secondary" distinction down to an essay, in favor of focusing on the actual matters that those are a heuristic for: GNG needs significant coverage in reliable sources that are independent of the subject (which are likely to be considered "secondary", but not by everyone in all cases); WP:OR requires not adding new analysis or synthesis of published material that reaches or implies a conclusion not clearly stated by the sources themselves ("primary" is just a heuristic for "probably doesn't have the kind of analysis we need for writing a good article on most topics"); WP:BLPPRIMARY seems to be concerned with advertent or inadvertent doxxing, difficulty in knowing that the "John Smith" in the source is the same "John Smith" the article is about without WP:SYNTHESIS, and probably WP:BALANCE; WP:DUE and WP:BALANCE, like GNG, would probably do better by focusing on reliable sources and independent of the subject over saying "secondary"; and so on. Anomie⚔ 16:03, 4 October 2025 (UTC)
I am seeing a number of academic journal articles now appearing on-line with both PDF and non-PDF versions available, with only the PDF version retaining the page numbers. Trying to cite specific passages in an article that is dozens of pages long, but does not have page numbers, requires citing section headings, which is awkward and often not very precise, so I usually link to the PDF version. Donald Albury 20:41, 26 September 2025 (UTC)
I'm not sure where you get the idea that PDFs are more likely to be primary sources? When citing historical matters, official reports, books, and the like (which can be primary, secondary or even tertiary) often PDFs are the only format available online. I'm also at a loss trying understand why you think primary sources should be removed from an article for being primary? Primary sources are often more reliable than secondary sources for certain types of information (ridership statistics for public transport for example). Not every source is required to satisfy the notability guideline. Thryduulf (talk) 21:29, 26 September 2025 (UTC)
We don't have an invariable requirement to weblink all of our sources at all, so that's not an overriding concern. PDFs are also just as often (actually in my experience more often) people's own self-published résumés, or brochures, or corporate and organizational newsletters, or transcripts of committee meeting minutes, or other unreliable sources that shouldn't be in an article at all, and are very frequently used to reference bomb a topic around a lack of any WP:GNG-worthy sourcing. Also, I did say there can be exceptions where a PDF is a more reliable source than the norm — but the norm is very much that editors use primary source PDFs instead of GNG-worthy reliable sourcing much more than they use reliable source PDFs of GNG-worthy reliable sourcing. Bearcat (talk) 15:51, 4 October 2025 (UTC)
Your experience is completely different to mine. Unreliable sources should not be used, regardless of format. What format sources are in has not relevance on their reliability. And yet again, you are implying that all sources need to be GNG-worthy, which is absolutely not the case - as long as there are sufficient sources that do demonstrate notability it's completely irrelevant whether the other sources do or do not. Thryduulf (talk) 15:55, 4 October 2025 (UTC)
No, the rule is not that as long as some of the sources in the article are GNG-worthy, the rest of them are allowed to be junk — the rule is that we have to get as close as feasibly possible to basing the article entirely on GNG-worthy sourcing while keeping any use of non-GNG-worthy sources to an absolute minimum. If I had a Wikipedia article about me, for example, then the acceptable use of primary sourcing in it would be for basic life details (my middle name, my birthdate, where I went to high school, etc.) that were desired as background information but might not have made it into any of the GNG-worthy coverage about my work — and even then, if any GNG-worthy reliable source could be found that supported the same information, it would take precedence as a better source for the fact than anything primary. It could not, however, use primary sources to support any information about my career that was meant to be understood as part of the reason why I would be eligible for a Wikipedia article: it could not claim that I was a notable journalist just because I had a staff profile on the self-published website of my own employer, it could not claim that I was a notable community figure in my own city just because I won a minor local award that could be sourced solely to the self-published press release of the award committee instead of any media coverage reporting the award presentation as a news story, it could not claim that I was a notable musician just because my music's existence was sourceable to YouTube or Spotify, and on and so forth. We're allowed to sparingly use primary sources for strictly factual background information, but not as support for anything that's meant to be a brick in the notability wall and not as anything more than the smallest "as close to zero as possible" percentage of an article's overall sourcing pool that we can get it. Bearcat (talk) 16:21, 4 October 2025 (UTC)
I'd much prefer to read about and discuss the topic that was raised: fixing bare links that are PDFs (at scale). -- GreenC 16:35, 4 October 2025 (UTC)
It requires me to download the PDF and put it in manually, defeating the purpose of not doing it manually. But then it doesn't do anything except for just showing me the PDF I just uploaded. Cdr. Erwin Smith (talk) 18:20, 4 October 2025 (UTC)
@Bearcat, where exactly is this "rule" that we have to get as close as feasibly possible to basing the article entirely on GNG-worthy sourcing while keeping any use of non-GNG-worthy sources to an absolute minimum written down? WhatamIdoing (talk) 02:32, 7 October 2025 (UTC)
That's anectodal evidence. In my experience, PDFs are usually journal articles or official reports from government agencies or respected organizations. Cremastra (talk·contribs) 16:40, 4 October 2025 (UTC)
And in mine. Donald Albury 21:30, 4 October 2025 (UTC)
PDFs are often articles in magazines or journals. Cremastra (talk·contribs) 23:02, 26 September 2025 (UTC)
Here's my latest solution in search of a problem (SSP). We use the term so much, it needs an acronym for brevity and ease of use.
Edit summaries already support wikilinks. Why stop there?
Make it possible to create clickable external links using the same coding as used on a wiki page, with the same ability to specify the linktext.
Support some wiki markup, such as bolding. Since the edit summary is in italic text, the double-apostrophe technique should create roman. ―Mandruss☎ IMO. 14:01, 30 September 2025 (UTC) Edited per discussion. 16:16, 30 September 2025 (UTC)
Belated add: The edit summary text box is one line long, and I can never see more than 90 characters of it on my system. When I can see characters 91–140, I can't see characters 1–50. Meanwhile, the system supports up to 500 characters, so I have a 90-character window into a 500-character text field. Make the box auto-wrap so none of the text is ever out of view. This is basic user interface design. ―Mandruss☎ IMO. 10:08, 2 October 2025 (UTC)
Thank you. So how do we get the developers to bump the priority of a phab ticket? Without that, this seems unlikely to happen in the foreseeable future. ―Mandruss☎ IMO. 17:10, 2 October 2025 (UTC)
Once it is reopened, you can try meta:Community Wishlist. Developers are primarily volunteers, so this is akin to How do we get editors to write new articles about some subject - you can try to recruit people with the skills and interest. — xaosfluxTalk 10:45, 3 October 2025 (UTC)
A sad state of affairs. We should be the developers' customers, not their subjects. En.wiki community consensus should mean a lot to them, by policy if necessary. Who invented this system? ―Mandruss☎ IMO. 15:12, 3 October 2025 (UTC)
The same people who invented the English Wikipedia? WP:CONEXCEPT has been our official policy since 2007. WhatamIdoing (talk) 02:36, 7 October 2025 (UTC)
Then our official policy has been wrong since 2007. Slavery was "policy" longer than 18 years. ―Mandruss☎ IMO. 04:54, 7 October 2025 (UTC)
First suggestion withdrawn. ―Mandruss☎ IMO. 16:26, 30 September 2025 (UTC)
No, do not make clickable external links. We do not need page histories and watchlists filled with spam or links to malicious websites that cannot be removed. Sure, we can delete edit summaries, but unless you have a way to double the size of the admin corps, making significantly more work for admins is unlikely to be an improvement. —Kusma (talk) 15:45, 30 September 2025 (UTC)
Then make edit summaries editable by anyone, the same as anything else. This provides benefits beyond a solution to your very valid concern, including the ability to correct an incorrect edit summary without the need for a dummy edit. Any problems introduced would be no different from what we're already successfully dealing with elsewhere. ―Mandruss☎ IMO. 16:09, 30 September 2025 (UTC)
Upon reflection, that's unworkable. We would need some way to track the changes to edit summaries, or we would lose context (see WP:REDACT). That would be an absurd Rube Goldberg contraption. So cancel the external link suggestion, leaving the markup suggestion. ―Mandruss☎ IMO. 16:15, 30 September 2025 (UTC)
What happens to all the old summaries that happen to have two or more apostrophes stuck together? What if they're specifically something like "<i>..</i> -> ''..''"? Would we have to support <nowiki> too? Plus, there are lots where the markup is unbalanced; example from the last few minutes. —Cryptic 16:51, 30 September 2025 (UTC)
I guess it ain't perfect. Off the top of my head, I can't think of anything on en.wiki that is. Never look just at the downside of something; instead, weigh the downside against the upside. The upside is that I could improve the clarity of my edit summaries, and this business is all about clear communication between editors. ―Mandruss☎ IMO. 17:01, 30 September 2025 (UTC)
It looks like most of your edit summaries are "ce", I don't think this proposal will do anything to improve them. 😀 Anomie⚔ 01:27, 1 October 2025 (UTC)
That's right, make it about me. That usually works. My use of the first person is a communication/rhetorical device only. ―Mandruss☎ IMO. 02:34, 1 October 2025 (UTC)
The enhancement I'd like to see is: Simpler edits should get automatically generated edit summaries, like Replaced 'tyop' with 'typo'. WhatamIdoing (talk) 03:23, 2 October 2025 (UTC)
The diff shows what you did, so that would be redundant in an edit summary. Edit summaries sometimes lie, so I always look at the diff. The edit summary is primarily for why you did it (for typo corrections, I would just say "ce"). If you change "Muslim" to "Hindu", I'm going to want a better edit summary than Replaced 'Muslim' with 'Hindu'. The software is not yet smart enough to identify a mere typo. ―Mandruss☎ IMO. 10:50, 2 October 2025 (UTC)
I used to think that some self-explanatory edits did not need an edit summary, but it can be worth saving others the trouble to click through to the diff when a good edit summary explains what you have done. —Kusma (talk) 14:32, 2 October 2025 (UTC)
When you say that you "always look at the diff", you presumably mean that you look at diffs when you're going through Special:Watchlist or Special:RecentChanges. An edit summary that merely duplicates the diff is still useful when you're looking at the article history and trying to figure out which edit needs to be reverted. In that scenario, Replaced 'Muslim' with 'Hindu' is very helpful, and "copyedit" or even "correction per source" is not at all helpful. WhatamIdoing (talk) 20:32, 2 October 2025 (UTC)
Again: Since edit summaries sometimes I lie, I never trust them. Or, if you prefer, trust but verify, whatever that means. ―Mandruss☎ IMO. 21:52, 2 October 2025 (UTC)
Sure. Now you need to figure out how long the words "The responsibility for achieving consensus for inclusion" have been in WP:V. There are about 7,000 revisions in the history page. Don't you think that an edit summary might be helpful in helping you figure out which diffs to check?
If you really never use them, then VPT can probably give you some .css code so they don't display at all on your screen. WhatamIdoing (talk) 00:54, 3 October 2025 (UTC)
I'll trust them if I have worked with the editor for years and know them to be both honest and competent. ―Mandruss☎ IMO. 01:43, 3 October 2025 (UTC)
Trusting some individual editors doesn't help you figure out which one of 7,000 edits to that page is the one that has the change you're looking for. An edit summary might. WhatamIdoing (talk) 02:28, 3 October 2025 (UTC)
Ok, I'll give you that if you'll give me phab:T6716. ―Mandruss☎ IMO. 03:37, 3 October 2025 (UTC)
An easier to use edit summary box would be great. I hate scrolling through the edit summary box when undoing an IP edit on my phone. —Kusma (talk) 07:57, 3 October 2025 (UTC)
@Mandruss, you know how you can start typing an edit summary, and matching previous ones pop up? Then you can pick the correct one (saving lots of typing time) and hit Enter, and it saves.
If we give you T6716, you'll lose that. WhatamIdoing (talk) 02:38, 7 October 2025 (UTC)
Thanks. I've come to grips with my dismal failure here (again), but now I have a consolation prize/silver lining. ―Mandruss☎ IMO. 03:18, 7 October 2025 (UTC)
The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.A summary of the conclusions reached follows.
This is not going to happen. Sandboxes already exist for testing formatting, and the rest are things we specifically don't want. We don't need LLM-generated proposals for things that are so very far off-mission. Anomie⚔ 21:40, 7 October 2025 (UTC)
Hello fellow Wikipedians! I would like to propose a new namespace called "Playground:", a fully separate space from the main namespace for creative, experimental, and fictional content.
This would be a safe sandbox where users can explore ideas, test new article formats, and share imaginative or original content without affecting mainspace articles.
Playground: is fully separate from mainspace, so standard Wikipedia policies for main articles (notability, verifiability, no original research) do not apply.
All content will be clearly labeled with templates such as fiction, experimental, or original research.
Editors are encouraged to maintain community-friendly standards within Playground:, including civility, respect, and compliance with rules against racism, bias, or harassment.
Content will be technically separate and not indexed by search engines, so mistakes, experiments, or unusual content remain contained.
Playground: follows the model of existing Wikipedia namespaces such as Draft: or User:, where mainspace rules are relaxed for safe experimentation.
I want this idea to grow with the Wikipedia community, so ask me anything about it! Suggestions, improvements, critiques, and additional ideas are super welcome. Think of me as your co-pilot for building the "Playground:" namespace—let's make it safe, fun, and useful for everyone.
Create clear guidelines for "Playground:" content, labeling fiction, experiments, or original research.
Establish templates for tagging and categorizing content.
Ensure content is technically separate from mainspace and configure it not to be indexed by search engines.
Encourage community input on rules, structure, and governance to maintain a friendly, productive environment.
Looking forward to your thoughts and feedback! Let's give Wikipedia a place to be playful, imaginative, and experimental—without compromising its mission of providing reliable, verifiable knowledge.
& by the way, I used ChatGPT to generate all that, because I’m too lazy and not that good at brainstorming. Swarde11 (talk) 19:49, 7 October 2025 (UTC)
Users already have personal sandboxes to experiment in. Adding fictional content or original research fundamentally conflicts with Wikipedia's mission(WP:NOTHERE). 331dot (talk) 19:54, 7 October 2025 (UTC)
Why should an encyclopedia host any of that? There are plenty of other places on the Internet, or you could create one yourself. Phil Bridger (talk) 20:12, 7 October 2025 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
There is an endemic problem on Wikipedia where articles are just timelined with dates followed by prose. I try to correct these when I can, but it would take eons. Quoting from my userpage, "One of the most frequent problems on Wikipedia is listing dates of occurrences in articles. As an example, you will see "on MMMM DD YYYY[i], X occurred." This makes for choppy and bad writing. In many cases, the dates or full dates are unnecessary. However, they should be integrated into the flow of the text. The above example would be better written as "X occurred on MMMM DD YYYY[i]". The specific date is only important when its directly relevant, otherwise only the month need be mentioned, and only then if it seems worse with just the year. Mentioning the date is usually redundant when a link to a specific dated event is in the sentence."
When I have changed this on articles, I dont have a ready example, it goes from a timeline of events to an actual prosaic narrative. I think this needs to become part of the manual of style and encouraged more. It also would make editors be more involved than just sniping updates on articles every time an event happens.
Awesome!!! Thanks Metallurgist (talk) 07:30, 8 October 2025 (UTC)
I don't think anyone does this on purpose. It seems to happen as a consequence of each person adding a separate bit, and nobody looking after the whole article. If you'd like to see a long list of 'bad examples', then most of the "COVID-19 pandemic in <place>" articles have very severe cases of this problem. WhatamIdoing (talk) 03:18, 8 October 2025 (UTC)
Oh dont even get me started on coronacruft. I joyfully participated in a mass AFD on that awhile ago. Metallurgist (talk) 07:29, 8 October 2025 (UTC)
Anyway, I dont think its on purpose. Its just lazy and unencyclopedic. And really a pet peeve of mine. This isnt always adding current events. There are lots of old articles where itll say In 1945, this. In 1948, that. Metallurgist (talk) 07:31, 8 October 2025 (UTC)
There is a perception that because an encyclopedia must be neutrally, impersonally written, there is no need for the skill of the creative writer. I'm sure we're all guilty of this; I'm sure I've done it. But we're here to tell a (well-sourced, unbiased) story, and we should tell it well. We live in an era where the skill of good writing is undermined (Grammarly: "you can't write well, you aren't even aware of the stylistic weaknesses that will cause you to fail your degree and lose your job, you haven't the time: you need me to write for you") or undervalued (chatGPT: "who cares whether you can write! I can do it for you!"). Humans, arise: and write good articles instead of date-lists! Elemimele (talk) 14:16, 8 October 2025 (UTC)
I wonder to what extent this sort of thing may be to differences in the way people organize information. What some see as "lazy" or "no need for skill" may be someone else's "clear organization of facts in chronological order". Anomie⚔ 14:38, 8 October 2025 (UTC)
I think it's mostly what WhatamIdoing says, it's a natural consequence of people making small additions. I myself recently updated an article, not realising that my update made an earlier "As of..." now outdated. Generally these things should work themselves out over time, it's just that over time can approach infinity given the size of the encyclopaedia. CMD (talk) 15:08, 8 October 2025 (UTC)
I'm a writing snob, so there are times when I wish all English Wikipedia contributors could write effectively. But the reality is that if my standards had been followed, English Wikipedia would be a lot smaller and probably a failure. English Wikipedia relies on editors taking the raw updates made by less skilled writers and distilling them into well-written prose. Unfortunately, sometimes editors think having exact dates is really important, and resist attempts to rewrite proseline passages. isaacl (talk) 21:55, 8 October 2025 (UTC)
The "exact dates" thing is one of my minor pet peeves. Do we really need to know that _____ happened on 32 Octember 1962? Unless there's another even happening a few days before or after that, "Octember" or even just "1962" is enough. And sometimes it goes beyond the full date. There was a guy arguing earlier this year that the exact time of death when a celebrity died had to be included because it had symbolic meaning (to him, anyway). WhatamIdoing (talk) 19:04, 9 October 2025 (UTC)
To what degree of precision we need to specify dates (and times) is entirely dependent on context, and there are occasions when every precision between century and fractional-second are appropriate. The appropriate precision for the same event can also differ depending on context, for example History of Test cricket from 1877 to 1883#The first Test starts At 1.05 p.m. on 15 March 1877, the first Test began. while History of cricket states formal Test cricket matches are considered to date from 1877.Thryduulf (talk) 20:03, 9 October 2025 (UTC)
Here's the background data, which you can read or ignore as you choose:
In any given week, Wikipedia:Articles for deletion has between 250 and 700 separate nominations, with an average around 400–450.
During the last five years, nobody has ever created more than 135 nominations in the same week (averaging 20/day). Only 9 editors ever created more than 70 nominations in the same week (10–19/day; representing 20% of noms in those weeks), and most of them only did that once. Only 46 editors created between 36 and 69 nominations in the same week (5–9/day; representing 10% of noms in those weeks), and most of them only did that once.
About 15 editors created 10+ nominations about once every 10 weeks. For example, during the last five years, @Bearcat has created 10+ nominations in the same week on 28 weeks [out of 265 weeks] (range 10–29; average 14), and @Piotrus has created 10+ nominations in the same week on 34 weeks (range 10–36; average 15).
My question is:
What number of nominations would you personally consider to be someone "flooding" AFD, or otherwise taking up more than their fair share of the community's attention at AFD at a given point in time?
Note particularly the words "at a given point in time", because having one person create a thousand separate AFDs on the same day is a big problem, but if the nom divided those same thousand AFDs up so that only three were posted each day, that would probably go unnoticed, and still get all thousand through AFD this year.
I'm looking for practical, common-sense guesses about what's functional and what's too much. Helpful responses probably sound something like "A hundred on the same day is too much, in my personal opinion" or "It should be measured over a week, and I think 7 in a week is plenty for anyone – just save the rest for next week" or "Newbies should be limited to two nominations a week, but WP:XCON editors should get 20 a week, and admins and WP:NPP should have no limits".
For clarity: I'm not trying to reach a settled consensus at this stage. I'd just like to get some idea of the general range most people have in mind. If we tend to cluster in a given range, then I'll probably turn that into a proposal later. If we're all over the map, then I might drop it. WhatamIdoing (talk) 19:52, 14 September 2025 (UTC)
I would think it would have to be seen as contextual. The "limit" would undoubtedly be higher for an established editor who's got a solid record of knowing what they're doing, and lower for a newer editor who appears to be making frivolous or tendentious nominations. There are also sometimes projects that clean up significant batches of problematic articles in a collective sweep (e.g. tourist information radio stations that failed WP:NMEDIA a few years back), which would undoubtedly generate a significant number of AFD discussions in a relatively short timeframe. So I don't necessarily have a specific number in mind, because it would depend to an extent on each individual editor's history and record. Bearcat (talk) 20:01, 14 September 2025 (UTC)
To reduce the number of AfD nominations, maybe we could prohibit non-Extended Confirmed editors from nominating articles for deletion? (An exception could be made for WP:BLPREQUESTDELETE-type AfDs, but the BLP subject could always post a request on WP:BLPN and an experienced, extended-confirmed user could nominate the article on their behalf). Some1 (talk) 20:13, 14 September 2025 (UTC)
This would force newer editors to try and use the PROD, CSD, or draftification system, even when it doesn't apply or isn't the best solution. GreenLipstickLesbian💌🦋 20:31, 14 September 2025 (UTC)
I have no interest in reducing the total number of AFD nominations from its current level. Think of the story this way: We normally get 50–75 nominations each day. Imagine that someone decided to go on an article-deleting spree. They have a list of thousands of articles they believe should be deleted. How many can they post today/this week/this month/this year, without the AFD process getting overloaded due to sheer volume? WhatamIdoing (talk) 20:48, 14 September 2025 (UTC)
If someone has 5000 artisanal cage-free AI-free pristine-BEFORE AFD noms ready to go, they should post them. We should at that point thank them for their tireless work and work out a schedule/triage for getting them through AFD, and or a new x series criterion.
If someone did that same 5000 articles with a hallucinating AI or skimping on BEFORE, then it's disruptive due to the need for others to rigorously check the work. At that point we might call it a good defense of editor time to kick it all out and not waste the resources.
If there are clusters of similar articles in either case, well-managed bundling should be used to get the pages through. This seems likely in a lot of worlds where someone has a glut of AfDs they'd like to run. If you have 3000 stubs on Canary Islands footballers, and the first dozen afds come back with unanimous not notable, deletes, then bundling the last 2988 into a single discussion should work fine. Tazerdadog (talk) 07:51, 16 September 2025 (UTC)
I completely disagree. Nobody should be stacking up that much BEFORE effort before posting AfDs, because every nomination needs to be rigorously checked regardless of whether it is AI-generated or not, and thousands are massively beyond editor's capabilities to do that and just because you were right about 12 articles does not imply you are right about the next 1 let alone the next 3000. If you want to delete that many articles about Canary Island footballers, start and RFC and get consensus that Canary Island footballers are inherently non-notable and articles about them should be speedily deleted. Thryduulf (talk) 10:27, 16 September 2025 (UTC)
Important to note that the data we have is only of the number of afd subpages created. If someone made, say, 9 multi-article afd nominations each proposing fifty articles for deletion, they wouldn't even show up on the list. Data for that is much, much harder to collect. —Cryptic 20:33, 14 September 2025 (UTC)
Yes, that's correct. But searches such as this one make me believe that's not very common (probably less than 1% of noms). WhatamIdoing (talk) 20:54, 14 September 2025 (UTC)
Two thoughts:
Ideally, I'd want to tie any variability into, not whether someone's an admin or a NPP or whatever, but how consistent their record is. If you make 40 nominations in a week and 30 of them are kept or moved or merged, I'm more inclined to think you flooded AFD than someone who made 80 of which 60 are deleted or redirected. (I see now Bearcat made much the same point already. —Cryptic 20:52, 14 September 2025 (UTC))
What you're nominating matters. 30 nominations in a week by the same person, all of Algerian football players, is more of a flood than if they'd nominated 10 Algerian football players, 10 American companies, 10 Armenian villages, and 10 fictional characters.
If we assume that the ideal is for every article nominated at AfD to be given sufficient review to accurately determine whether
the nomination is correct
whether we should have content about this subject (primarily but not exclusively notability), and
if so, whether the content we do have is good enough (quantity and quality) to be worth keeping and/or merging.
Then there needs to be sufficient editor time to carry out that review and assess the article. Editor time is finite so we have to ensure that nominations are too. Unfortunately this is not easy because neither nominations nor editors are homogenous; as Cryptic notes the pool of editors reviewing nominations related to Algerian footballers is different to the pool of editors reviewing those related to American companies (pools do not overlap, but both editors and articles can be in multiple pools but the more different they are the less likely this is). Different pools are different sizes giving different capacities - there are far more editors interested in contemporary American baseball players than there are editors interested in indigenous religions in Africa.
Additionally the amount of time and effort a review takes varies immensely - for example it's almost inconceivable that a contemporary American rock band could be simultaneously notable and unknown to Google, however it is highly plausible that an 19th century Korean poet writing in Hmong could be widely regarded as the greatest of his generation but essentially unknown to the English-language internet.
Finally what matters is not how many articles per day are nominated by a single editor, but how many nominations related to a single topic are open concurrently - and relists make calculating that harder. Putting this all together means that 100 concurrent nominations related to contemporary US politicians in a week is significantly less likely to be problematic than 10 concurrent nominations related to 1920s Uruguayan authors. Thryduulf (talk) 22:11, 14 September 2025 (UTC)
IMO the number of articles per nom is always relevant in the sense that there are limits to how many articles a human editor can do a decent BEFORE check on in a given time period.
It's also relevant in that no one editor should demand too much of the community's time. AFD is for everyone more or less equally; it shouldn't become "WhatamIdoing's AFD playground", with a large percentage of noms only from me.
The number of articles per nom obviously becomes relevant if the nom is User:My_Secret_ChatGPT_Deletion_Bot, with a new nom every three minutes round the clock.
It sounds like you're leaning towards 100/week as potentially okay, but not 100/week on a narrow/difficult subject. Maybe the community would like to encourage editors to limit both total volume and also subject concentration (technically forming two severable pieces of advice). WhatamIdoing (talk) 23:52, 14 September 2025 (UTC)
there are limits to how many articles a human editor can do a decent BEFORE check on in a given time period. this is entirely true, but that's not an issue of flooding AfD. In a topic area with a large capacity, say contemporary US television, the limit of one person's ability to perform a proper BEFORE is lower than the community's capacity to properly review nominations. In a topic area with a low capacity, the community's capacity to review could be lower than an individual's capacity to do BEFORE checks (especially given the repeated objection to mandating a meaningful BEFORE search).
It's also relevant in that no one editor should demand too much of the community's time. yes, but that's not really core to the issue - what matters in terms of flooding is the total number of concurrent AfDs related to a topic area. If the community capacity for AfDs related to say Spanish places is 30 concurrent discussions it doesn't make a difference whether they're initiated by 1 person in one day or 30 people over 7 days. Ideally no single person should use the whole capacity, but unless someone is doing that regularly it's not really a major problem.
It sounds like you're leaning towards 100/week as potentially okay, but not 100/week on a narrow/difficult subject. No. Any limit needs to be expressed in terms of concurrently open nominations in a given topic area (remember that some discussions are relisted multiple times and others are speedily closed). 100 concurrently open nominations is an extremely large number and will only be viable in topic areas where AfDs are frequented by large numbers of experienced editors. 100 concurrent nominations related to UK railways would completely overwhelm the available editors and that's hardly a narrow or difficult subject. Remember also that the effort required to determine whether sources exist can vary considerably between nominations, even in the same topic area - a nomination consisting of a single article about a contemporary UK local councillor takes much less effort to review than a nomination of 5 articles about cotemporary UK local councillors, which takes more effort than a nomination of 5 articles about UK local councillors active in the 1980s for example. Thryduulf (talk) 16:58, 15 September 2025 (UTC)
As an editor who has stumbled across an article about an (IMO) non-notable UK local politician, how am I supposed to:
(a) know what the community's overall capacity for reviewing this subject is and
(b) find out how many other articles are currently at AFD in that narrow area?
I think that a limit per nom makes sense, because I know how many I've nominated. I have no idea what other editors have been doing. There is no WP:DELSORT for minor politicians in the UK during the 1980s. WhatamIdoing (talk) 17:31, 15 September 2025 (UTC)
(a) You can't. Although it's safe to presume that mainstream topics that are easy to google have the highest capacity and obscure topics where the majority of reliable sources are offline and not in English have the lowest.
(b) Sometimes this is easy by e.g. looking at deletion sorting lists, sometimes it's near impossible.
Together this is why it is not possible to put any firm numbers on what constitutes flooding. A limit per nominator is easy to measure, but just because something is easy to measure doesn't mean that measurement is useful. We want to avoid things like "It says I can start 30 nominations in a week, I've only started 29 therefore my behaviour isn't problematic", whatever limit we set becoming a target and rules lawyering about whether X is a different topic to Y.
If anything is needed, it is to explicitly state that editors must not overwhelm the capacity of AfD to fully and properly review their nominations. If other editors are complaining that they can't keep up, slow down. If your nominations are not getting any attention or are only getting perfunctory comments, slow down. If someone asks you to slow down, slow down. Editor capacity varies, the amount of capacity a single nomination uses varies. Thryduulf (talk) 18:02, 15 September 2025 (UTC)
I am not sure if there can be such a thing as too many AfDs. They can wait for few weeks to backlog to clear. I'd say creating over a 100 noms for more than one week in a row would be too much. Anyway, ironically, while I have not been creating many AfDs recently, I may breach 10+ this week, as someone just reverted my PRODs and redirects on a number of unreferenced novels... (and I am unaware of how to use a gadget to move noms into a single one, so they'll have to go separately to AfD). Piotr Konieczny aka Prokonsul Piotrus| reply here 00:20, 15 September 2025 (UTC)
As someone who's been a fairly prolific nominator you really can't bludgeon your way through AfD. Since the default outcome is Keep, you have to convince other editors to spend their time assessing the articles and they're unlikely to do so if they feel there's too many to give a good look at or the effort put into the nom is too low. If I nominated 40 or 100 or 1000 articles at once, we'd get well-thought-out responses to the first 5 or 10 and the rest would be procedurally kept or, worse, kept based on low-effort responses like "There must be sources out there somewhere, I just can't be arsed to find them." And there's a good chance that if I came back later with small batches, they'd all end up being deleted because folks had enough time to give them sufficient attention. When someone tries this I think they learn pretty quickly that it doesn't work.
I'm curious whether "flooding" is a documented problem that's causing ongoing issues due to editors who do it repeatedly, or if we're just talking hypotheticals here. We had to scramble the Top 10 list of prolific article creators because 'all' of them were huge time sinks for the community - Can we say the same about the list of deletion nominators that's taking shape? And if there is actually an overwhelming number, can't we just relist and work through the backlog like we do at AfC? –dlthewave☎ 01:40, 15 September 2025 (UTC)
Your record (in the last five years, various caveats apply) appear to be 29 in the same (calendar) week. But that could have been four a day/otherwise spread out.
I linked to the 'origin' of this question in the first sentence, which started with one (1) AFD in which the newbie nom used an LLM to generate the nom statement. However, this led to discussions about a situation that is AFAIK still hypothetical, but foreseeable: a AI-based bot/semi-automated script that finds articles, writes nom statements, and submits them to AFD. There was some interest in an anti-flooding rule. We've also previously had conversations about indiscriminate AFDs of poorly sourced WP:LUGSTUBS articles about Olympic athletes, in which editors expressed a desire not to have hundreds of AFDs at once. WhatamIdoing (talk) 01:59, 15 September 2025 (UTC)
Perhaps 10 nominations per day would be acceptable. Then the nominator may have had enough time to properly evaluate whether there adequate sources. If LLMs are used to find or nominate pages, then a BOT approval would be required, and I would say do not approve such a bot. Also bulk nominations of large numbers of pages in one nomination is ill-advised as the nominator probably did not check properly, and later participants probably won't have time either. Graeme Bartlett (talk) 08:20, 15 September 2025 (UTC)
Solution in search of a problem This isn't something that needs to be legislated by WP:PAGs. Headbomb {t · c · p · b} 15:58, 15 September 2025 (UTC)
Sure, I can agree that the community doesn't "need" to have this rule (yet). But if the community did decide to create such a rule anyway, approximately where do you think the limit should be? WhatamIdoing (talk) 16:07, 15 September 2025 (UTC)
As someone who is consistently cranking out the AfDs I think when a real problem shows up I think we can invoke the inverse of invoking WP:IAR and just tell them to stop without having a specific number in mind. I see the peak I ever did was 42 in a week, which is probably in excess of six on one day; I think someone might have complained. The problem already shows up when we get a big group nomination, and those tend to get closed and resubmitted one at a time.The big mass edit problem continues to be creation, precisely because it takes too much work to run up all the deletions. Mangoe (talk) 17:19, 15 September 2025 (UTC)
I don't think we can count on everyone pausing their noms just because someone asks/complains/tells them to stop. That was not the case, e.g., a few years back when a couple of editors were starting 10 RFCs at a time. Mere "requests" are not enough for a small fraction of editors, especially when they take a rigid view of rules and seem to be perseverating on the wrongness of these articles existing. "Hey, the rules say no more than X noms in a week" or "Hey, the rules say no more than Y articles in the same subject area in one week" usually works for that group.
(I think all of RFC's frequent flyers eventually ended up blocked. Since then, though, the "rule", which is really just advice to leave a note on the talk page before starting more than two or so simultaneous RFCs, really has prevented overuse by others. "Enforcement" generally requires nothing more than a chat on the user's talk page.) WhatamIdoing (talk) 17:41, 15 September 2025 (UTC)
If an editor tells you, in good faith, to slow down, slow down. If they tell you to stop doing something, stop doing it. If necessary, start a discussion about it.
If you ask an editor to slow down or stop, and they do not, get the opinion of others. Take it to a relevant noticeboard if necessary. Thryduulf (talk) 18:05, 15 September 2025 (UTC)
Yes, we can take it to a relevant noticeboard, where another rules lawyer will inevitably say "There's no written rule against flooding AFD with a thousand noms per day, so therefore flooding AFD is permitted."
I wonder if you could imagine a level of AFD activity that was at least unlikely to be productive, even if it wasn't "a rule". WhatamIdoing (talk) 20:53, 15 September 2025 (UTC)
There's no written rule against flooding AFD with a thousand noms per day, so therefore flooding AFD is permitted." so what we do is make it a rule that flooding AfD is not permitted. We don't put a number on it (because that's not possible, per my previous comments) but note (with clearer wording than here) that there should be sufficiently few nominations that the editors interested in AfDs in a given topic area have enough time to do a proper review of all nominations within 7 days of a nomination; and what factors impact what the capacity of a topic area for new nominations is (popularity, discoverability of sources, ease of access to sources, number and complexity of concurrent nominations, etc), and examples of signs that a topic area is saturated with nominations (e.g. calls to slow down/stop, comments that time is lacking to review all nominations, rushed engagement, etc). Thryduulf (talk) 04:35, 16 September 2025 (UTC)
I think that it is possible to put some numbers on it. Above, you say that thousands at once is unacceptable. Another time, you say that 100 per week is "an extremely large number" that will usually not be viable. Elsewhere, you say that 5 or 10 on a narrow subject could be an upper limit for a niche subject.
This leads me to guess that you would normally accept more than 10 noms per week (just not necessarily on the same niche subject) and be concerned about more than 100 noms per week (especially if on the same subject). WhatamIdoing (talk) 16:44, 16 September 2025 (UTC)
That's correct if you're just cherry-picking the numbers out of context, but when you read what I'm writing in full it should become clear that you're grossly simplifying things. The context is necessary and unavoidable because what would be grossly overwhelming in one topic area would be a barely a ripple in others. I shouldn't (but apparently do) need to stress, again, that the limiting factor is the combination of editor availability and the cumulative total complexity of all concurrently open nominations in a given subject area. Both of these factors are constantly varying, and both are the sum of multiple other constantly varying factors. In the same subject area 10 nominations might overwhelm editors if all nominated today but nominating 20 in a day in a fortnight's time could be fine.
An editor posting circa 100 nominations today spread over 25 very different topic areas is less likely to be problematic than the same editor posting the same number of nominations but in only 4 topic areas. However if those 4 topic areas are say contemporary US politics, 20th Century British history, 2010s Australian sportspeople and mathematics then it's again less likely to be problematic than if the topic areas were Chilean, Argentinian, Peruvian and Bolivian soccer teams. We cannot speak in absolute numbers, only generalities and relative probabilities of disruption. Thryduulf (talk) 19:04, 16 September 2025 (UTC)
I don't think that whether "User:Expert" is on holiday for a fortnight should affect whether an article should be sent to AFD, or even ten such articles. I think that we can speak in absolute numbers: It's always okay to send 10 articles a week to AFD. It's never okay to set up a thousand separate noms in one week. WhatamIdoing (talk) 04:14, 17 September 2025 (UTC)
It's always okay to send 10 articles a week to AFD. It's never okay to set up a thousand separate noms in one week. except neither of those are true. If there are 10 open, low-quality nominations about an obscure topic that takes a lot of effort to properly review then another 10 nominations is not unlikely to overwhelm to available capacity, especially if this is done week after week. On the other hand a single instances of 1000 well-researched nominations across a very broad range of easily-researched topic areas following extensive pre-AfD discussion is possibly manageable, especially if the concern is something other than notability. Thryduulf (talk) 10:41, 17 September 2025 (UTC)
There shouldn't be anything at AFD that for which the primary concern is "something other than notability".
Nobody is ever going to get sanctioned for opening 10 AFDs in the same week. They may get yelled at for "low-quality" part (that can, and does, happen even for single AFDs, and therefore has nothing to do with the volume), but not for the "10" part.
The average day at AFD has about 60 nominations. I assure you that even if you post 1,000 absolutely perfect nominations in the same week, you will get complaints about the volume. This would more than triple the usual volume (your 1000 plus the usual ~400) and it would do the opposite of winning friends and influencing people. WhatamIdoing (talk) 16:43, 17 September 2025 (UTC)
At this point I'm going to stop because it's clear you're determined to try and put a single number on something that absolutely cannot be reduced to a single number, regardless of what I or anyone else says, so further discussion is not going to be productive. Thryduulf (talk) 19:05, 17 September 2025 (UTC)
Much of the time, serial AfD noms have not done appropriate levels of WP:BEFORE eg. searching inside books at archive.org and Google Books. Signing up for Wikipedia Library, doing broad searches across commercial databases (GALE etc). And of course Google Search, though that is usually the least reliable method - what's active on the web on any given moment is random and changes. These noms prefer to leave the work for others. They nominate quickly by gut instinct and glancing at the article itself. I say this based on years of experience at AfD observing behaviors. There is no consequence for BEFORE, the policy might as well not exist. The reason is plausible deniability. -- GreenC 19:23, 15 September 2025 (UTC)
Mass creation of under/badly-sourced stubs is far more a presenting problem, and at least one editor was brought to ARBCOM and quit rather than face a case. If someone's bad noms were a real problem, I expect AN/I and ARBCOM could be brought to bear on the problem editor. Mangoe (talk) 20:58, 15 September 2025 (UTC)
But even if we all agreed that we needed to run a million articles through AFD, the fact is that AFD can't handle a million articles at once – or even ten thousand. What do you think AFD's carrying capacity is? WhatamIdoing (talk) 21:02, 15 September 2025 (UTC)
I imagine that someone would make a fuss at AN/I, the whole nine yards would be reverted, and if the person tried again, they would be blocked. But the worse we ever have had has been a bunch of combined nominations, and those just get closed procedurally and people renom the articles one at a time. We don't need anything extra to deal with a problem that hasn't happened and which we can think of a straightfomward solution. Mangoe (talk) 23:27, 15 September 2025 (UTC)
If the issue is just that you feel AFD's carrying capacity can't support dealing with individual AFDs for mass-created articles, the correct thing to do is to suggest alternative mass-deletion options to reduce that pressure; then we can point people to those instead. Just banning mass-nominations at AFD without providing an alternative is obviously not going to work (otherwise, what exactly should people do if articles are inappropriately mass-created?) I think it's obvious that situations exist in which we may need to delete hundreds, thousands or even tens of thousands of nearly-identical automatically or semi-automatically-created articles at once, so the real question is what the best way to do that is. --Aquillion (talk) 04:05, 16 September 2025 (UTC)
I'm more interested in the situation in which the articles aren't mass created.
@Tazerdadog says "If someone has 5000 artisanal cage-free AI-free pristine-BEFORE AFD noms ready to go, they should post them. We should at that point thank them for their tireless work and work out a schedule".
I think:
What if we can't figure out whether they're AI-free? Most nom statements are too short for AI-detector websites to be reliable. We've got an outline of a fully autonomous AFD nom bot higher up on this page. If it had fairly good accuracy and posted one every 15 minutes or so, for eight hours in a row, with short nom statements, you'd never know that it was a bot instead of a dedicated human.
Why should we "work out a schedule" after the fact, instead of telling everyone up front that 5000 AFD noms is too much at once, and to encourage them to work out a human-scale schedule before hand?
Re: AI. Distinguishing AI generated text from human generated text is in general a HARD thing to do. If one of us had a reliable way to do it, we should patent it, build a small wrapper website, and rake in the millions. Realistically the best we can do is ask the prolific posters what their workflow is, and then assume they're honest about it until/unless we get significant evidence to the contrary, such as a hallucinated source or a "as a large language model" that slipped into a nom.
Obviously best practice would be to post the discussions as you finish with them instead of in a batch. That said, if someone is offering us high quality work, we should not decline it because it's too much, but rather adjust our processes so that it can all be addressed.
One big point is that a lot of the times when someone is proposing dozens or more articles for deletion at about the same time, it's because they are substantially similar. Having a functioning process for such batched nominations is important. No opinion on if that should be the current process, a RFC, or a hybrid or something else entirely. Tazerdadog (talk) 17:14, 17 September 2025 (UTC)
Meh… Doing just about anythingin bulk is considered disruptive behavior. Even when we are following (or enforcing) “the rules”.
We all have different thresholds before we call something “disruption”, but ultimately we know it when we see it.
And we have warnings and procedures to deal with disruptive behavior once the line has been crossed. I don’t see a need for yet more “rules” to deal with disruptive behavior. Blueboar (talk) 21:17, 15 September 2025 (UTC)
I believe that "no consensus" is still supposed to result in keeping the article, so I'd call keeping the article "the default outcome". WhatamIdoing (talk) 16:06, 16 September 2025 (UTC)
It depends. If there are zero responses then it's treated as an expired PROD because they could've just PRODed the article in the first place. If there are responses but no consensus then the article is kept.
In practice, editors who feel like they can't keep up with the number of AfDs will usually voice their objection at each one rather than let them be deleted due to lack of responses. That's what I meant when I said you can't bludgeon your way through AfD.
I'm fairly certain that even the AfDs that get zero responses are at least being glanced at by other editors, and hopefully closing admins would notice any red flags or odd patterns. –dlthewave☎ 17:02, 16 September 2025 (UTC)
Since the default outcome is Keep – what??? In sports we routinely have editors mass copy-paste simple rationales like "Fails SPORTCRIT" on dozens of articles on accomplished foreign and pre-internet athletes – without any indication of WP:BEFORE whatsoever – and they almost always get deleted/redirected unless me or a handful of other users put in ridiculous levels of research to prove that the obviously-notable subject is indeed notable. BeanieFan11 (talk) 15:42, 17 September 2025 (UTC)
What happens at the end of an AFD discussion? If there's a consensus, we do whatever the consensus says. If there's no participation, we treat it like an expired prod. And if there's participation but no consensus, we keep the article. I call that last one "the default outcome", since it is the outcome that applies when editors can't agree on what to do. WhatamIdoing (talk) 16:47, 17 September 2025 (UTC)
This is imv a rare issue that I am only marginally concerned about. Consensus seems to be that it crops up rarely enough that it's not a huge issue and we don't need special rules for it. You may want to look at previous discussions like [1][2][3]. I think that for certain kinds of editors there should essentially be no limits to nominations. An NPPer patrolling the queue without filtering by topic area should be allowed to nominate very many AfDs per day/week since they're spread out over a many topic areas. (Maybe 50 and 200 respectively? Those numbers are definitely on the high end; I'm curious what the record of someone like Onel is.) The real issue is when a lot of noms are in one topic area, which overwhelms the editors in that area, especially when the noms could be bundled. Good bundling (first doing a "pilot" or three, then bundling the rest) could reduce a lot of pressure on AfD. Toadspike[Talk] 16:34, 17 September 2025 (UTC)
More generally, a decline in quality participation is a real issue. I previously suggested expanding the circumstances under which soft deletion is allowed as a potential solution. I also think we should come down harder on editors who have a history of erroneous AfD noms, even if seemingly made in good faith, with strongly worded messages directing them to other areas of the project or tbans. Toadspike[Talk] 16:35, 17 September 2025 (UTC)
In the last five years, @Onel5969 has created 10+ AFDs 53 times, so about once every five weeks. Half were 10–14 nominations in the same (calendar) week. Only five times were there more than 20 noms in the same week. Those weeks had 22, 31, 40, 54, and 63 noms in them. At a glance, all of these happened in the second quarter of 2023. WhatamIdoing (talk) 16:53, 17 September 2025 (UTC)
This is looking at the issue the wrong way. The real issue is editors not doing the work to make sure only articles that should probably be deleted have AfDs. If the system is working correctly most articles that have AfDs should be deleted, as only articles likely to be deleted should be nominated. If an editor's AfDs can be shown to be problematic and they won't learn from there mistakes, then they should be taken to ANI and tbanned for deletion discussions. The solution to "to many AfDs" is more editor participation. That editors who work in areas that will result in more AfDs raise more AfDs is just a truism. -- LCU ActivelyDisinterested«@» °∆t° 17:03, 17 September 2025 (UTC)
+1 as the kids say today. Given the overall number of articles this project has, the number of articles that are nominated in a given week, month, or year is exceedingly small. Astronomically small as a ratio, even. And the overall volume is certainly far, far beyond the threshold of all the articles that are fit for deletion. We could lose a hundred non-GNG complaint articles on footy players alone each month and still not see the end of that problem for decades, for example. So the question of optimization of AfD as a whole should not be seen in terms of what a reasonable bottleneck should be, but the identification of the right articles for deletion. Now, to some extent that latter inquiry is consistent with what WAID is getting at. But I also sense here (as with many broad strokes discussions about AfD) that there is a sense that the main reason we should be concerned that due diligence is observed in the nomination process is that we presently have to high a potential for the wrong articles (and for too many articles in general) to be deleted, or for the process to be exercised inefficiently. And I'm not entirely convinced that has ever been the case: at least, not at a level which accords with the amount of agitation that contention engenders. Basically what I am saying is that we could stand to have 25x the deletions we currently have in a given week and it would probably not be a problem as a per se matter. And anyone who feels instrincly that this statement is wrong should probably stop to question themselves about whether they are really concerned that "just the right articles get nominated" or if they are using that concern as a pretext to justify trying to slow down the whole machinery because they have broad concerns about a "deletionist" agenda. None of which is to say that we don't have editors who habitually do take an overly liberal, injudicious or lacking in caution approach to nominations. But as AD says, those can typically be identified and handled through individual scrutiny, rather than trying to form a one-size fits all set of metrics to judge all editors who nominate far above the average by. SnowRise let's rap 07:09, 18 September 2025 (UTC)
If we could guarantee that every article nominated at AfD should be deleted (not merged, redirected or any other ATD, actually deleted) then I would agree that the number of nominations at AfD would not be problem. However we cannot guarantee that, or even anything close to that - only 48% of discussions on 2nd September resulted in a straight delete, speedy delete or soft delete outcome (32 of 67); 33% (22) were withdrawn or had a straight keep or speedy keep outcome. If we assume this is typical (and I have no reason to suspect it isn't) then unless we are happy to delete around a third of notable articles (and if you are then you really need to reassess whether you actually have the project's interests at heart) then we must ensure that there is an opportunity for every article nominated to be fully reviewed. Unless and until there is a consensus that AI can reliably do this job (to say the chances of that happening are infinitesimal is probably overstating the likelihood) then this requires human editors and human editors have a finite capacity. So there needs to be some limit on nominations. Thryduulf (talk) 12:02, 18 September 2025 (UTC)
I don't know what the current numbers are, but we used to delete about three-quarters of articles at AFD, which would suggest that September 2nd might be an atypical day. WhatamIdoing (talk) 15:36, 18 September 2025 (UTC)
Even three quarters is way too high a rate of outcomes other than deletion to allow for unfettered deletion without adequate review. Even if only 1% were non-delete outcomes, that would still equate to around 5 wrongly-deleted pages every week (calculated based on the number of discussions currently listed at Wikipedia:Articles for deletion#Old discussions (open) which works out at around 75 nominations per day or 525 per week on average; I've not attempted to determine how representative this period is). 25% non-delete outcomes would be around 131 wrongly deleted articles per week, 33% would be circa 173 per week. Thryduulf (talk) 15:49, 18 September 2025 (UTC)
@Snow Rise, I don't think we actually can stand to have 25x the deletions we currently have in a given week. I say this not because I've any particular love of keeping articles, but because 25x the nominations would need 20x–30x the editor participation, and we can't magic up several thousand extra editors.
For context, in 2023, about 26,500 registered editors participated in at least one AFD (as contrasted with 7,100 editors at all village pumps combined). There were about 21,000 AFD noms that year, or 400 a week. 25x would be 10,000 noms in one week. The "regulars" wouldn't be able to expand their participation, because there aren't enough hours in the day. We could predict that the total number of editors would go up (if the typical AFD participation is one nom, one article creator, and two other editors, we'd still expect the article creator to appear), but we wouldn't realistically expect enough extra editors to appear. A lot of those AFDs would probably end up being closed as "no quorum" (usually soft delete) or "no consensus" (straight keep). WhatamIdoing (talk) 15:53, 18 September 2025 (UTC)
You also can't necessarily extrapolate that each new article nominated would result in one new unique editor, because some will have created more than one article that has been nominated and not all of them will be both aware of the nomination and available to comment on the nomination in the given week. The latter is true even for very regular editors - back in early 2005 when I was still a new editor I created an article to make a point (see this comment). That was nominated for deletion) while I was on holiday (and had apparently even noted this on my userpage) so was entirely unaware of the nomination until after it had been closed. Thryduulf (talk) 16:27, 18 September 2025 (UTC)
Ok, I don't disagree on the manpower logistics here; my previous comments were about the propriety and due care of nominations in the abstract, and the advisability of putting our thumb on the scale because of perception of potential harm from the deletions themselves, rather than the feasibility of handling a given volume. But let's just be clear about what we are saying. Are we suggesting that there should be a limit on nominations of even those articles which are unambiguously proper deletion candidates, for purely workload throughput reasons? Because that's a different discussion from most of the previous. And a position I would have various reservations about, though I admit it is hard to argue with the numbers under our current volunteer hour scarcity crunch. SnowRise let's rap 19:48, 18 September 2025 (UTC)
Are we suggesting that there should be a limit on nominations of even those articles which are unambiguously proper deletion candidates, for purely workload throughput reasons? Yes. It is of topmost importance that deletion is done only after proper review and is seen to be done only after proper review. If that means articles that are not actively harmful (and all articles that are actively harmful can and should be speedily deleted) hang around longer than some people think is ideal then that is a tiny cost to pay. Thryduulf (talk) 22:40, 18 September 2025 (UTC)
Yes, I do think that there should be a limit on nomination volume, regardless of the righteousness of any particular nomination.
I think this for workload reasons (my primary reason; there is a maximum number of AFDs that is practical at any given point in time, and having AFD collapse helps nobody) and also for social/community reasons (no fair taking up more than your fair share of the community's time. In the case of AFD, I think 'your fair share' might occasionally be ~20% of noms in any given week, or 10% over the long haul, but it's probably not >50%). WhatamIdoing (talk) 22:48, 18 September 2025 (UTC)
But then how do we build into any guidance we develop the flexibility to match the upper limit to the availability of volunteers, given the fluctuations we should expect over time? It's one thing to suggest a percentile value as you do above (which I agree makes sense), but that's probably too awkward a system to implement for a policy. And if we fix a specific number, it stands to become too onerous if volunteer numbers go up, or too permissive if they go down. SnowRise let's rap 07:56, 19 September 2025 (UTC)
One option is to offer a fairly wide range: "10 a week is okay, but 100 a week is usually too much."
(Pros: It's concrete, and everyone reading it will come to the same conclusion. Cons: Perhaps it will encourage someone to make exactly 99 noms every single week. The numbers might need to be changed in the future.)
Another option is to offer a vague principle: "Don't flood AFD. If you believe a large number of articles should be deleted, especially if they are on similar subjects or subjects that are more difficult for editors to research (e.g., people from non-English-speaking countries; subjects more likely to be discussed in offline sources), then spread them out over time."
(Pros: The principle is the "real" rule. Cons: Someone will cry "flooding" over two or three articles, and others will claim that 200 is just fine.)
There are other things that can be done to support AFD's health. For example, we could check stats for the regular noms to identify people with below-average deletion rates to be more selective (or to stop using AFD for merge/split/redirect/content-dispute purposes). WhatamIdoing (talk) 16:29, 19 September 2025 (UTC)
The vague principle is only one that can work. If there is disagreement about whether something is or is not flooding, stop nominating and discuss it with other editors until there is a consensus about whether it is or isn't. 2 and 200 are both going to be OK in some circumstances and not OK in others. Thryduulf (talk) 17:31, 19 September 2025 (UTC)
I think that makes sense for another reason as well: there has historically been a number of scenarios where a problem editor mass-created a large number of articles which all had similar issues which militated for the presumption that they were mostly policy non-compliant, but they didn't qualify for speedy deletion under some technicality or another. Keeping an emphasis on flexibility and contextual considerations would prevent any limitations put in place to control nomination volume on the average day for AfD volunteers from becoming a needless WP:BURO issue whenever the community has to clean up some sort of expansive mess. SnowRise let's rap 22:53, 19 September 2025 (UTC)
I see AfD as a community effort to cull bad articles. An editor who makes 100 nominations in a day isn't creating work for others or monopolizing our time; they're doing the good and thankless task of finding content that we don't need and bringing it to our attention. If one editor is taking on 1/4 of that work, good on them!
That being said, if we find ourselves unable to keep up with the workload, I think that setting up some sort of queue would be a far better option than limiting input. I'm not exactly sure what that would look like but perhaps efforts like Articles for Creation or Page Curation would be good models, focusing on working through the queue and bringing in help to address the backlog. –dlthewave☎ 00:10, 20 September 2025 (UTC)
It would be better if we could find ways to discourage people creating articles that need to go to AfD. Every academic nowadays writes themselves a Wikipedia article as part of their standard social media management, along with their LinkedIn profile. The major theme of the Teahouse is "writing an article without secondary sources won't work", "do you really want to write about yourself?" and suchlike. I can't see how we can limit AfD without the flip-side of making it harder to get rid of articles that are basically COI promotion. Elemimele (talk) 16:07, 26 September 2025 (UTC)
I don't see anti-flooding concerns as a way of "limiting" AFD. It's more of a way of "smoothing out the rate" at AFD. 50 non-notable BLPs a week for 10 weeks is better for the AFD process than 500 this week and nothing for the next 9 weeks – even though in both cases, we're talking about sending 500 BLP articles through AFD. WhatamIdoing (talk) 16:46, 26 September 2025 (UTC)
I specify "better for the AFD process", because if someone's real goal is deleting the articles (e.g., articles they've judged to be "basically COI promotion"), then they might not care about whether AFD gets the right answer, so long as AFD deletes the articles they dislike. Even notable subjects can engage in "COI promotion", and AFD's job is to determine the subject's notability. AFD isn't actually supposed to be protecting Wikipedia's purity against COI promotionalism. WhatamIdoing (talk) 16:49, 26 September 2025 (UTC)
Yes, that's fair comment. I definitely think it's unhelpful when people create too many AfD's simultaneously. For example, on all the associate profs who make a wikipedialinkedin page for themselves, you're quite right: the correct thing is for someone to check whether they meet NPROF based on how well-known and well-cited their work is, but that requires good judgement and a bit of research from someone willing to look into typical citation rates in their field etc., which takes time. Maybe I'd support the idea of an informal guideline. Elemimele (talk) 14:35, 27 September 2025 (UTC)
Yes, that's the kind of problem that's concerning. Delete them all, if they're all non-notable, but send them through AFD at a pace that isn't going to strain the process. I'm not skilled at NPROF assessments myself, and I think I'm typical of Wikipedia editors in that regard. So if we dump a huge number of associate profs in AFD at once, then what we're really doing is dumping a stressfully large number of AFDs on a very small number of competent respondents, or we're indicating that we don't actually care if they're evaluated fairly. Just spreading them out would help a lot.
In this discussion, I feel like we've identified two main themes. One is that too many at once is too many. AFD usually gets ~400 AFDs in a week, so you shouldn't nom hundreds or thousands in a week. The other is that in niche areas (e.g., athletes from the same developing country), the supply of competent respondents is low, so even if you're going to start 100 AFDs this week (←not a recommended practice), spread things out and nom 10 Olympic athletes from Ruritania, 10 associate profs from Canada, 10 plastic surgeons from Mexico, etc., instead of 100 professional cricketers from Pakistan.
Do these sound reasonable to you?
As for a process, I was thinking this morning about something I've seen a couple of editors do, mostly for new articles. So lets say that you want to kill 500 articles on associate professors. It's not fair to nom them all at once, so what do you do? One option is just to make a hit list (in your sandbox or on your own computer; it doesn't matter) and nom 5 or 10 articles a day until you get to the end.
Another, when you're dealing with really large numbers of articles with a really high likelihood of deletion, is to try escalating measures. For example:
Make a list of subjects you believe are non-notable. (Optional: Tag them all with {{notability}}.)
Use Twinkle to {{subst:prod}} some of the articles in your list.
Check back later to see how many prod tags have been removed. Send those contested prods to AFD (or remove them from your list, if it looks like they're probably notable after all).
Continue trickling your list through the PROD process, and then, as necessary, sending the survivors to AFD.
I'd suggest this process to anyone who wants to get rid of articles about, say, minor local non-profit organizations. For something like an athlete, you could do the same general kind of thing, except with a proposal to merge/redirect to their team instead of WP:PROD. WhatamIdoing (talk) 17:29, 27 September 2025 (UTC)
I want to reiterate the argument I have made at the mass-creation RfC: while the rate of nominations can sometimes be a problem, the defining characteristics of a spate of AfD nominations that are bad are not usually rate alone: it's usually a lack of a valid reason for deletion, a misunderstanding of notability, or a bad-faith attempt to remove an article. None of these problems will be addressed by limiting the rate of nominations. There are cases in which we may wish to limit the rate: when we change one of our notability guidelines, for instance, a large number of articles may need to go to AfD, and need to be reasonably spread out in time. But I see no evidence that that hypothetical scenario has been a problem, and I see no other scenario in which this is a problem that needs solving. Vanamonde93 (talk) 17:57, 27 September 2025 (UTC)
There has been a recent issue where dozens of PRODs and AfDs (about 80 of them last month) of pre-Internet-era track and field Olympians were all created in a short timespan. For comparison, the usual rate that these get created is one or two per week. The rate is of particular importance here because unlike most processes on Wikipedia, there is a one-week deadline for most PRODs and AfDs, so when many are created all at once it can be difficult to properly address them in time.
Only the idea of an AI-type AFD-nominating bot is "hypothetical". WhatamIdoing (talk) 19:43, 27 September 2025 (UTC)
If one user submits a 109 PRODs in an hour, and persists in similar behavior after a warning, we ought to TBAN that user from PRODs, because that is disruptive behavior. I don't see how a community-wide rate limit will help in any way. I am also fundamentally opposed to a numerical rate limit on deletion unless a numerical rate limit on creation is instituted. Vanamonde93 (talk) 20:23, 27 September 2025 (UTC)
That discussion does not establish a rate limit on creation, only a requirement of community support for "large-scale" creation, and does not limit manual creation in any way. No rate limit exists for manual creation besides the very generous ones tied to account flags. Vanamonde93 (talk) 17:13, 30 September 2025 (UTC)
Note that that's in part due to historical accident. Mass creation was originally considered in the context of automation, and therefore, until fairly recently, WP:MASSCREATE was part of Wikipedia:Bot policy. Attempts to institute a limit on manual creations failed in part due to the fact that Wikipedia:Bot policy is not suitable for regulating the actions of human editors. Part of the reason we finally evicted WP:MASSCREATE from the bot policy was because we got tired of people trying to bend the definitions of "bot" to cover some kinds of fully manual editing so they could apply WP:MASSCREATE (and others bending it back wrongly to try to counter those arguments). I don't think there has been a discussion since that eviction to see whether consensus has changed with respect to restricting the rate of human mass creations, now that "WP:Bot policy can't regulate human behavior" is no longer a relevant concern. Anomie⚔ 18:04, 30 September 2025 (UTC)
Why would anyone AI generate noms when there are easy tools to use? I think the bigger problem is LLM generated articles, which is the real reason for more noms. LLM use needs to be a blockable offense, altho difficult to prove. Many people admit to it now, but they wouldnt if it were blockable. Metallurgist (talk) 02:18, 8 October 2025 (UTC)
Why would anyone use AI to generate noms? Maybe because they don't feel confident in their ability to express their thoughts, but maybe also because they wanted to get a lot of articles deleted. A human probably can't write a nomination reason for 500 articles on the same day, but an AI tool could. WhatamIdoing (talk) 02:26, 8 October 2025 (UTC)
I havent noticed there to be 500 in a day. Metallurgist (talk) 08:33, 8 October 2025 (UTC)
So far, we're getting very few AI noms – probably less than 1%, by my spot check (which would translate to less than one per day). But that could change without warning. WhatamIdoing (talk) 16:53, 8 October 2025 (UTC)
I think this is a "we'll cross that bridge when we get to it" situation. Hopefully we never will. Bremps... 01:03, 12 October 2025 (UTC)
AfDs like Wikipedia:Articles for deletion/Erika Kirk tend to be quite messy, with both !keep and !delete votes making comments about what future sourcing may or may not exist (it's hard to talk about lasting significance when anything is breaking news). What if we had a moratorium for proposing an article for deletion for a week until after a {{breaking news}} template is removed? Clovermoss🍀(talk) 16:52, 22 September 2025 (UTC)
Or a moratorium on creating an article until a week after {{breaking news}} would have been removed? ScottishFinnishRadish (talk) 17:05, 22 September 2025 (UTC)
I feel like that's a good idea in theory but that would cause its own issues whenever we get another Donna Strickland-type situation. Clovermoss🍀(talk) 17:07, 22 September 2025 (UTC)
I had a similar suggestion to that of SFR on Discord, as a way to reduce BLP violations as editors create dozens of articles using non-vetted information that is prone to be wrong on multiple levels. As a compromise, maybe we could have a moratorium focused solely on BLPs, so that articles focused on the victims or suspects would only be allowed after 7~14 days had passed. Isabelle Belato🏴☠️ 18:01, 22 September 2025 (UTC)
I'm trying to think how to make that work... the vast majority of current events type articles (killings, airplane and traffic accidents, assaults, kidnappings) are going to mostly have of content related to BLPs, just by design - the shooting and the Jan 6 attack are actually pretty good examples of that. Even articles about natural disasters (the other brand of current events articles) still sometimes have substantial BLP-content relating to disaster management, political blunders, and missing persons. Circling back to the main proposal, while I get the idea behind it, the issue really isn't that it's an article about a current event - it's that it's an AfD of an article that was in the news, getting viewed a lot, and which people, in their normal, non-Wikipedian lives, have very strong opinions of. (See also Wikipedia:Articles for deletion/Killing of Austin Metcalf and Wikipedia:Articles for deletion/Killing of Iryna Zarutska) I'm going to explicitly disagree with JPXG's idea that here that like if you look at any of the breaking-news afds they are just like this, as it's been my experience that most AfDs about current events, even if a bit unclear, are a lot calmer. (Wikipedia:Articles for deletion/2025 shootings of Tremonton police officers and Wikipedia:Articles for deletion/2025 Vienna Township Cessna 441 crash for a couple of AfDs that have an average amount of PVP, and still ended up being pretty uncontentious)If nobody minds me bringing up two more examples, I'd strongly opposed any rule that stopped somebody from starting Wikipedia:Articles for deletion/Cristo Rey Jesuit High School shooting, as it was both a little gross and spread misinformation. Ditto, Wikipedia:Articles for deletion/Syrian Air Flight 9218 was about as in the news and current-eventy as you could get, created in good faith, and it was snow-deleted within about 18 hours of its creation due to not being, well, real. But neither of these examples were clear, they needed an AfD to solve. GreenLipstickLesbian💌🦋 20:02, 22 September 2025 (UTC)
"PVP"? -- Beland (talk) 22:37, 22 September 2025 (UTC)
@Beland Oh sorry, it's a video game term that stands for Player versus player, - basically when, in a game, the part of it where you fight other people instead of working together to solve an issue, a boss, or some environment hazard. GreenLipstickLesbian💌🦋 23:27, 22 September 2025 (UTC)
Aha, thanks for clarifying. -- Beland (talk) 01:50, 23 September 2025 (UTC)
Reposting my incipiating go-off, for those who were not scorded enough to see the origin of this:
i really do think there should be some special case breaking-news notability process thing that isn't afd, because afd is basically totally worthless.
the closer for an afd about a breaking-news thing often has to just disregard the first four days of votes, because they are arguing about stuff that became completely untrue
like "nobody knows whether he was even really the guy!" or "this is a one-day news item that will be never be talked about again and instantly forgotten!" and then it isn't, or "this is a big deal that will have lasting significance" and then it doesn't
ostensibly, the purpose of an afd existing at all (since it uses a gigantic amount of editor time to run it versus doing something else) is so we can have on record a high-quality discussion for the notability of a topic
which you just can't do for something that is happening right now
like if you look at any of the breaking-news afds they are just like this, maybe by the end of the week people are making useful comments based on having enough info that it might be useful in 5 months time
so i don't know if there is a concrete suggestion lying underneath this. afaict there are a couple things, one being a "the afd has a 7-day preliminary where there's a notice and maybe a discussion section but either way you can't cast keep/deletes"
or maybe no discussion section at all, just like, "this article will be assessed for notability in one week" and then a blank afd
the only thing i can think is like, maybe there are some cases when blp (or just common decency) demands we not have a page at all, and then due to this we have to have it up for 14 days and not 7
but idk i think that if it's genuinely super bad people can speedy it or sth
I've noticed that these types of AfDs also tend to have an influx of newbies who have no idea how Wikipedia works. I've never seen a breaking news AfD not be chaotic. For example, Wikipedia:Articles for deletion/Steven van de Velde was a headache because it was getting 40,000 daily pageviews and everyone reading it saw the AfD notice. Clovermoss🍀(talk) 17:39, 22 September 2025 (UTC)
Does putting a low level of protection on the AFD page resolve chaos sufficiently? -- Beland (talk) 19:24, 22 September 2025 (UTC)
To some extent it does and that's why it's often done, but it doesn't resolve the fundamental problem of it's hard for everyone to come to a consensus while the dust is still settling IRL. No one can accurately predict the future. Forcing everyone to wait would save a lot of editor time, prevent no consensus closes, and prevent "Wikipedia is trying to censor X!" fearmongering from people looking for the latest scandal to grasp onto. Clovermoss🍀(talk) 19:39, 22 September 2025 (UTC)
Wouldn't this prevent the deletion of blatant BLP violations if they were related to breaking news? -- Beland (talk) 19:47, 22 September 2025 (UTC)
If the goal is to avoid wasting editor time, quick deletion of an article that is going to be deleted may be better, because it prevents editors from wasting a lot of time working on the article itself. I wonder what the ratio is of time spent on article vs. AFD in these situations, and how often these articles are kept vs. deleted. -- Beland (talk) 19:53, 22 September 2025 (UTC)
If there is no consensus and no stable version (as in the case of breaking news), the default would absolutely be non-inclusion. Chaotic Enby (talk · contribs) 19:53, 22 September 2025 (UTC)
Though AFD consensus takes 7 days unless there's a SNOW decision, right? Which puts the end of the process into a timeframe where the long-term significance of a topic is a lot clearer than the beginning of the process. Beland (talk) 19:56, 22 September 2025 (UTC)
True, but if the article being a BLP violation is blatant enough, it might just as well be SNOW closed (even if the numbers are not overwhelmingly in favor, but the policy-based arguments are). And, conversely, if the outcome is no consensus (again, considering the strength of the arguments – useful to keep in mind given the influx of !votes from new editors), then the content likely wasn't of the "blatant BLP violation" type.Speaking of, I'm not sure how "no consensus" closes for BLP articles whose existence is a potential BLP violation operate – usually, no consensus defaults to keep in AfDs (but to delete when considering inclusion of material inside an article), and WP:BIODEL only mentions it defaulting to delete on request of the subject, which surprises me given how conservatively WP:BLP is written. Chaotic Enby (talk · contribs) 20:27, 22 September 2025 (UTC)
Yeah, it sounds like maybe just letting the normal AFD process play out results in the right thing happening? I am listening right now to the On Point hour about the Erika Kirk AFD [4]. I was also quite alarmed at people on Fox News and podcasts and whatnot screaming things to their audiences about Wikipedia editors like "you are reprehensible, disgusting, evil people who are trying to cancel her and get rid of her because you are leftist" because one person dared to nominate a biography of one conservative they thought might not be well known. But then the host clicks through to the talk page and reads through the discussion and sees that people are mostly having a non-political conversation about reliable sources and Wikipedia notability policy. And sees that the overwhelming number of editors here were in favor of keeping the article, and that was the final decision. I can only expect that many of the conservatives who came running to the AFD because of angry partisan commentators experienced a similar teachable moment where they learned more about how Wikipedia actually works and that it is generally not evil as certain clickbait commentators make it out to be. Maybe we don't need to change the way we do things because certain loud people are disconnected from reality? Maybe people in general will get tired of running to Wikipedia only to find someone was crying wolf yet again? Maybe simply providing people neutral, accurate articles and having transparent editorial discussions will engender trust from anyone interested in learning more about reality? Looking at survey results like this 2019 survey, it seems more people already trust us than trust e.g. US conservative opinion-shouters. (And I do mean shout - one of the shows quoted is called "Louder with Crowder" which I just find hilarious.) And comparing to Gallup polls, a lot more than nearly every other institution in society. -- Beland (talk) 23:07, 22 September 2025 (UTC)
@Beland Blatant BLP violations (something like "X is a poopy-face" or "Y is horrible murdering scum and should be put down like a dog") can be revision deleted, just like in any article where it appears. ~ ONUnicorn(Talk|Contribs)problem solving 20:26, 22 September 2025 (UTC)
I think the question is about the existence of the article as a whole being a BLP violation? Say, an article about a person connected to a suspect/alleged criminal, whose existence might give the impression of guilt by implication. Chaotic Enby (talk · contribs) 20:30, 22 September 2025 (UTC)
Okay, and what about a (now deleted) paragraph-long, sourced, article about a teenager allegedly in a gang who shot somebody, and a large percentage of the article is a sourced quote from said teenager and a claim that he laughed while admitting to shooting them? But looking at the sources, some present it exactly as the Wikipedia article did, but other draw a distinction between the two events and make it clear that they got the quote in a "An official representative of the police said that somebody in the police heard the suspect say X"-type manner? While I clearly view that as problematic, I also have to concede that it wasn't blatant or revision deletable. GreenLipstickLesbian💌🦋 20:48, 22 September 2025 (UTC)
It would be a bit weird and not a great look to say "administrators can completely delete revisions or articles with no discussion if they feel justified by policy, but nobody is allowed to suggest deleting or object to deleting based on policy for several days". -- Beland (talk) 23:12, 22 September 2025 (UTC)
Something I have been thinking about recently was how to interpret WP:LASTING. We've seen editors arguing to delete recent articles on the basis that lasting coverage isn't present yet, as well as editors arguing to keep recent articles to give them time to see whether coverage lasts or not. Currently, this situation falls into a policy grey area, and I believe holding a broader discussion to clarify it would be more than welcome. Chaotic Enby (talk · contribs) 19:47, 22 September 2025 (UTC)
@GreenLipstickLesbian: I mixed up the shortcuts, what I was looking for was WP:SUSTAINED (although WP:CONTINUEDCOVERAGE does mention it). The only warning it gives is this may be difficult or impossible to determine shortly after the event occurs, as editors cannot know whether an event will receive further coverage or not, which doesn't read that clearly as warning against applying it strictly, but only about it being hard to ascertain – which could be a reason for either "keep and wait to renominate it" or "delete and wait to recreate it". Both of these frequently happen as !votes citing WP:SUSTAINED on AfDs about recent events, and having more explicit guidance in the guidelines could help avoid holding this recurring debate every AfD. Chaotic Enby (talk · contribs) 22:09, 22 September 2025 (UTC)
@Chaotic Enby Ah, that makes sense. And really? I've always read that as a warning - not a strict prohibition again invoking it, sure, but a comment at that something will be difficult/impossible to figure out is a sign that it can't be applied strictly - but it's interesting to hear a different perspective, and you're quite right that SUSTAINED doesn't seem mention it. And yes, trust me, I'm familiar with event AfDs [5] and how inconsistently SUSTAINED/CONTINUEDCOVERAGE gets applied at them. I don't find it to be that bad outside of high-profile cases, honestly. Sometimes I end up rolling my eyes ("no, a paper you found on research gate that somebody said they were submitting to a journal is not a good source") but I find the regular closers to be quite good at disregarding "it's sad/I saw this on the news/this was in a rural area, who cares/no European sources talk about it"-type !votes. And even when they don't, there's often a valid reason (sheer number of !keep votes, poor attendance, or genuine disagreement over whether the event had a LASTING impact). Again, I get that the people here who mostly only participate in the high-profile event AfDs are going to have the complete opposite experience to mine. GreenLipstickLesbian💌🦋 22:29, 22 September 2025 (UTC)
@GreenLipstickLesbian, closers might be reasonably good at ignoring various unhelpful !votes, but I do think there's a real tension between "we're not a newspaper, we're an encyclopedia" and "one of wikipedia's key contributions to the public good is neutral coverage of ongoing events". In my own opinion as a closer, I would say that I find most participation in event-related AfDs to be rather poor, and I think it leads to keeping many more events than we actually "should" according to our stated guidelines. But in a media ecosystem where Wikipedia is the place to go to get real and complete information about ongoing pandemics, disasters, and hot-button political events... -- asilvering (talk) 20:27, 23 September 2025 (UTC)
@Asilvering Yes, there's a genuine tension - but I don't entirely think it's always between the notnews crowd and the "we need good coverage of ongoing events" crowd; it's deeper than that. For example, the main reason I don't like event articles reliant on sources published within a few months of the event itself is because it's impossible to know you're writing a balanced article about them, right? The significance of events is something that's bounded by time, no amount of careful writing is able to overcome the fact that we can't see the future. And it's a really bad idea to write an article mostly based off your own interpretation of primary sources, for like... so many reasons. And yet we have situations where several very experienced editors look at event articles based on breaking news sources and conclude that that the actual article is of such a high quality that it's one of our best.
Tropical Storm Carlotta (2018) relies entirely on one government report, lots of daily weather reports, and newspapers published while the event was ongoing. Passed a modern FAC (I haven't done a proper BEFORE, no clue if it's notable)
Pan Am Flight 214 - 29 sources, 14 of which were published within a week or two of the event, mostly the 72 hours following the crash. (These are what most of the article is nearly entirely based on). 2 were published before the crash. 2 are government reports/musing from the FAA about what the FAA learnt from the crash. 10 sources remaining; [6] is a self-published blog by amateur history buff. [7][8] are fluffy, very local stories published on the event's anniversary, and can't support much in the article.[9][10][11][12][13] are primary news reports on the CAB hearings; you find these for every single airplane crash under the sun and don't really give a good perspective on the impact of the an event [14] is what I consider to be an okay source; it's a book about several crashes, this one being an example, but it was published a year out from the event - that's not much distance at all, and it was written/published before the actual ramifications of the crash (none) were known and the conclusion of the CAB hearings. With no lasting impacts, under NEVENT this is borderline notable at best, and certainly shouldn't survive a good NOPAGE argument - passed a modern era FAC easily.
The issue doesn't become any better when you look at Category:GA-Class Thunderstorm and tornado articles or Category:GA-Class Tropical cyclone articles. And yes, technically GA and FA are different from the notability criteria - but as the too-often invoked WP:FAOWN reminds us, these articles have been checked to make sure they rely on high-quality sources, which is also what NEVENT is meant to be checking for. FA criteria 1B also says that it's meant to place the subject in context; that's what the NEVENT continued coverage requirement is about (though we make an exception for recent events that look like they're going to have a lasting impact). If anything, they should be stricter - but then again, an experienced editor I like and trust told me once they thought an old revision of Victoria Roshchyna was almost ready for GA, well before her body had been returned, investigations started, charges filed, or actual facts known. I don't dispute that the !votes at event afds tend to not be all that useful (have I mentioned how often I end up rolling my eyes at some !votes?); I just believe that that the root cause is notnews versus must document current events, I think the disconnect is over the notability of events, period. GreenLipstickLesbian💌🦋 23:05, 23 September 2025 (UTC)
The claim that Pan Am Flight 214 had no ramifications is quite false - see Pan Am Flight 214#Legacy for a summary of improvements to aircraft design that resulted from this crash. -- Beland (talk) 23:13, 23 September 2025 (UTC)
Hurricane Hector wasn't deleted; it was merged and redirected to 2018 Pacific hurricane season. It looks like editors decided it had a few noteworthy characteristics, but there was enough fluff that it could be trimmed to fit into its parent article.
It seems like 1.) perfectly notable things are being complained about as if they are non-notable, and 2.) complaints are being made about the normal editing process not limited to current events. Enthusiastic contributors are always adding too much detail to one article or another, and there's a constant trimming somewhere on the project to suit a general audience. I don't really see that as a problem; it may generate stronger sourcing or more accurate summaries. -- Beland (talk) 23:27, 23 September 2025 (UTC)
@Beland And there's another slight disconnect - are very selective merges and delete different? The article went from 42k btyes to just over a thousand. That's not technically deleted, but the vast majority of the content was deleted and far greater merges get fought at AfD because they resemble a pure deletion too much. Similarly, you said it was all fluff - but if it's so objectively fluff, then how does it make it through quality controls that should be much higher than AfD? Agree with you that regular editing can deal with these; some of the dislike of current event articles seems (to me) poorly targeted, yet also the proposed moratorium on AfDing them would be damaging. I still like the idea of a more prominent "wait and see" option for recent events, but I think we have to trust other editors if they voice a PAG-based reason why we shouldn't. GreenLipstickLesbian💌🦋 23:45, 23 September 2025 (UTC)
Okay, not many ramifications currently documented in secondary sources, you're quite right. The vast majority of that entire section is the author's summary of the CAB/FAA discoveries, about the CAB/FAA discoveries, supplemented with a the odd newspaper article about the hearing. (The FAA is better than many investigations; but as a rule, I try to avoid trusting government agencies for accurate summaries of their own impacts. It doesn't really allow you to place to events in context, and when I want to read an FAA report, I'll just go read the report) The final paragraph is better sourced- which might give it notability, but isn't really passing a NOPAGE argument. GreenLipstickLesbian💌🦋 23:30, 23 September 2025 (UTC)
I'm an engineer with an interest in system failure; case studies are invaluable for learning about how to design safely. Wikipedia articles on plane crashes are generally a lot more digestible and discoverable than government reports, and link together to identify trends and contrasts. We don't need an article on every one, but the idea of deleting this one seems harmful, given that this sort of thing is of interest both across engineering disciplines and as a matter of public policy and good governance. How the government handled this crash, for example, is a good case study to think about when considering self-driving cars. -- Beland (talk) 00:02, 24 September 2025 (UTC)
That's true - but, no offense Beland, but I don't know you, I don't know how qualified you are, and I don't know how qualified any other editor on Wikipedia is. So yes, I'd like the article, as plane crash articles as what I read... probably the most of, and I actually do like reading about how governments respond to disasters - but I'd like the article to be based on better sources, and I'd like Wikipedia to not be how we identify trends. (My area's of interest isn't quite as fancy as yours, and I often end up reading an article written by a non-expert based on news reports, going "well that's a lot of crap/that's a myth/ffs you've conflated two very different groups of people here in a way that's very offensive", then moving on because, well, the article went through GA/FA process and fixing it is too much of a headache; I can only assume it's similar in other disciplines) GreenLipstickLesbian💌🦋 00:10, 24 September 2025 (UTC)
Whenever fixing it, or even a little piece of it, is too much of a headache to face, then that's what we have maintenance tags for. WhatamIdoing (talk) 00:51, 24 September 2025 (UTC)
Definitely agreed about maintenance tags.
I see a lot of badly written articles, but generally they have not gone through GA nor FA. Articles that are the subject of high-profile news events tend to be some of the best, simply because there are so many eyes on them and active participation in improvement. We have a backlog of over half a million articles missing citations, going back to 2007, so articles that actually have verified citations to reliable news sources for all substantial content, are doing very well compared to that backlog or the large mass of low-traffic articles that haven't even been tagged for missing sources, or have sources that haven't been checked. -- Beland (talk) 01:22, 24 September 2025 (UTC)
I agree with this, one thing I've noticed a lot with AFD votes in this area is that votes often seem to come down to editors' gut feelings on whether there will be more coverage in the future, which is not a super helpful way to determine if an article should stay. 🌸wasianpower🌸 (talk • contribs) 00:11, 23 September 2025 (UTC)
It would be much better if we waited a few weeks or months for the dust to settle before creating such articles in the first place, but that's not gonna happen, so I just don't know what to do about it. Phil Bridger (talk) 20:01, 22 September 2025 (UTC)
Maybe if we require new articles about breaking news topics to go through AFC? And put a 7 or 10 day holding requirement before they can be accepted? ~ ONUnicorn(Talk|Contribs)problem solving 20:29, 22 September 2025 (UTC)
Deleting an article after creation is a wastage of time. Curbs should be placed right at the beginning.
For 'Breaking-related' living beings :
Wait for 14 days before creating an article.
For 'Breaking' events/'Breaking-related' deceased beings :
No need to wait for article creation. Cdr. Erwin Smith (talk) 20:28, 22 September 2025 (UTC)
I'll note that WP:RSBREAKING is already a guideline, and it advises "wait a day or two after an event" for better information. I think this is a better time window than 14 days, during which time potentially millions of people have checked Wikipedia for in-depth information about a current event. I find Wikipedia often does a very good job of compiling news sources into a more comprehensive narrative than any individual news source does. This works on short notice simply because so many editors are paying attention and are highly motivated to update. I think it's harmful if articles are created with bad information because of poor sourcing or addition of spurious information from otherwise good sources in the minutes and hours after an event, which is why we have RS:BREAKING. But for a current-events article to be created that's simply not notable to kick around for a month or two...that seems fine. I think we should probably actually err on the side of inclusion because if we make a mistake in the other direction we'll simply miss the opportunity to inform the vast majority of an article's potential readers. And to recruit new editors who are brought in by a subject of interest they can improve right away. -- Beland (talk) 23:20, 22 September 2025 (UTC)
I said to maintain the existing law for 'Breaking events'/'Breaking-related deceased beings' which was perhaps lost in the words -
"No need to wait (for 14 days) for article creation".
But Wikipedia suggests to take some extra caution for living beings, while at the same time a large portion of these living people related to the 'Breaking events' might also loose significance down the line. Hence the suggestion in line with WP:NOTWHOSWHO -
"Wait 14 days for 'Breaking-related' living beings."
I think this will be the most balanced approach which would cater to the opinions of both the sides. Cdr. Erwin Smith (talk) 07:20, 23 September 2025 (UTC)
Yeah, 14 days is far longer than the 1-2 days the news media needs to debunk spurious allegations against a living person, and far shorter than the months needed for the legal system to exonerate someone. -- Beland (talk) 19:47, 23 September 2025 (UTC)
Refer to WP:BNS. We don't need to rush to publish everything. There are some things for which news websites are present.
Also, if the said allegations are so 'spurious' that they got debunked within 1-2 days, I don't think the event would be notable enough for an entire article to be created around it.
As for the person involved, he might or might not already have an article writen on him in wikipedia, which won't be built just based on/because of the 'spurious allegations' on him. Cdr. Erwin Smith (talk) 20:41, 23 September 2025 (UTC)
We already don't "rush to publish everything". Most news events aren't ever mentioned in Wikipedia at all.
Debunking an individual allegation doesn't mean the whole event is non-notable. You would hardly advocate that we delete September 11 attacks, right? And yet there were all kinds of allegations about 9/11 that were debunked within one or two days.
And, yes, sometimes we do (and should) have articles about individuals merely because of the spurious allegations against them. See also Category:Journalistic scandals. WhatamIdoing (talk) 21:49, 23 September 2025 (UTC)
Exactly. I remember some events when people were confused about the number of shooters because it's unclear if what was seen by two different witnesses was the same person or two different people. Sometimes people are confused about the number of bullets or explosions because of echoes or other loud noises that happen. Often the number of people killed is over-reported because people are double-counted. Once authorities can get on scene and take photos and assign names to victims and properly aggregate witness accounts, mistakes like these get eliminated. -- Beland (talk) 22:38, 23 September 2025 (UTC)
I've already suggested to include breaking events instantaneously.
For 'Breaking' events/'Breaking-related' deceased beings :
No need to wait for article creation.
As for the individuals having spurious allegations against them, if such allegations don't get debunked within 14 days, then we can surely consider including the person. However, baseless allegations on a person which gets debunked within merely 1-2 days shouldn't and doesn't warrant an entire article on him/her in Wikipedia. It's to stop such low value articles we're having this discussion at the very first place.
For 'Breaking-related' living beings :
Wait for 14 days before creating an article.Cdr. Erwin Smith (talk) 06:02, 24 September 2025 (UTC)
This discussion actually started with the opposite problem: people nominating high-profile articles that should be kept immediately after being created. The "wait 14 days" idea not only doesn't solve that problem, it makes it much worse, inviting more accusations of censorship. -- Beland (talk) 07:09, 24 September 2025 (UTC)
So deleting an article after the expiration of the moratorium isn't censorship, but waiting for 14 days for its creation is?
WP:NOTCENSORED : Content will be removed if it is judged to violate Wikipedia's policies (especially those on WP:BLP).
WP:GNG : "Presumed" means that significant coverage in reliable sources creates an assumption, not a guarantee, that a subject merits its own article. A more in-depth discussion might conclude that the topic actually should not have a stand-alone article—perhaps because it violates WP:Not.
WP:Sustained : Brief bursts of news coverage may not sufficiently demonstrate notability. However, sustained coverage is an indicator of notability, as described by notability of events.. If reliable sources cover a person only in the context of a single event, and if that person otherwise remains, and is likely to remain, a low-profile individual, we should generally avoid having a biographical article on that individual.Cdr. Erwin Smith (talk) 08:44, 24 September 2025 (UTC)
The situation which generated cries of censorship was the threat of deleting an article that is widely considered appropriate. The "speedy keep" proposal would quickly end that threat for future articles editors think appropriate. I expect deleting an article that is widely seen as unnecessary would not generate much complaint. Delaying the deletion of a non-notable article is also unlikely to generate complaints. In contrast, I would expect preventing the creation of an article that is widely seen as appropriate would generate complaints of censorship. -- Beland (talk) 16:56, 24 September 2025 (UTC)
More importantly, nobody would follow that rule.
The English Wikipedia's rules are derived from actual practice. If the page has a {{policy}} or {{guideline}} tag at the top, then that page is meant to be documenting existing reality – not imposing new things on the community. Waiting 14 days is never the community's practice for anything. We should not create a rule that tells editors (all three-quarter million registered editors each year) to do something completely different from what they're already doing. And if we're stupid enough to create such a rule, we should expect the community to ignore the rule and do what they think is best.
Editors who have rigid thinking styles often find this difficult to grasp, but most Wikipedia editors don't care what the rules say. WP:Nobody reads the directions. Writing down novel directions doesn't have the desired effect because (almost) nobody will read or care about your made-up rules. WhatamIdoing (talk) 17:24, 24 September 2025 (UTC)
I think a simple *mandatory tab* which asks whether the new 'Article Being Created' is related to any 'Existing Wikipedia Article' , would make enforcing the 2w waiting period a lot easier.
The existing 'Existing Wikipedia Article' should definitely have the 'date of occurance' of the mother event.
Deleting articles after creation risks demotivating and driving away a lot of new good-faith editors, who would've otherwise voluntarily decided to wait out the 2w stay period. Cdr. Erwin Smith (talk) 10:33, 25 September 2025 (UTC)
No. People don't work that way. WhatamIdoing (talk) 13:59, 25 September 2025 (UTC)
Many editors often rush to create articles on living beings related to 'breaking news' , a lot of which lose relevance down the line, and get deleted.
A waiting period of 2w will allow the dust to settle down, and seperate important people from the unimportant based on sustained coverage, causing a lot of these editors to lose interest on such unimportant people through a more level-headed assessment, who would then refrain from creating such articles. Cdr. Erwin Smith (talk) 15:58, 25 September 2025 (UTC)
If "many editors" create articles about "unimportant" BLPs related to breaking news, then how many of those happened in the last 24 hours? Have a look at Special:RecentChanges. You can filter for page creations. Come back with a list. If this is happening "often", then you should have more than a few links to share. WhatamIdoing (talk) 16:01, 25 September 2025 (UTC)
I think you should ask this question to @Clovermoss since she added the topic.
I'm just giving a suggestion that's all. Cdr. Erwin Smith (talk) 16:57, 25 September 2025 (UTC)
Clovermoss isn't the person who says this is a frequent occurrence. You are the person saying that. I don't think it's a frequent occurrence. WhatamIdoing (talk) 16:59, 25 September 2025 (UTC)
The primary purpose of my reply was clearly not to depict how frequently this problem occurs, but to give a potential solution to a problem. So my wording was loose there.
Even then I said "Many editors often rush to create" not "Often many editors rush to create".
I'm very confused by this requirement. Isn't every new Wikipedia article related to some other Wikipedia article in some way? Is this trying to prevent creation of articles about events that aren't mentioned in other articles? Is it trying to prevent creation of multiple articles about the same event? -- Beland (talk) 14:05, 25 September 2025 (UTC)
Only preventing articles on 'Breaking-related' living beings.
OK, but could you explain why identifying existing articles is useful in furthering that goal? -- Beland (talk) 15:51, 25 September 2025 (UTC)
If there's any other way to do this, I'm okay with that as well ! Cdr. Erwin Smith (talk) 16:01, 25 September 2025 (UTC)
You don't have any particular reason why you suggested asking editors whether there's a related existing article? I'm opposed to the waiting period in general, so I'm not advocating an alternative implementation of this idea. I'm just trying to understand it in case it's getting at problems I wasn't aware of or something? -- Beland (talk) 17:55, 25 September 2025 (UTC)
Already covered under the 3 policies I mentioned earlier.
Your suggestion would open the floodgates for hundreds of low value articles, a large chunk of which would inevitably get deleted down the line. This will demotivate and drive away a lot of new good-faith editors, who would've otherwise waited out the the 2 weeks period if notified prior. Cdr. Erwin Smith (talk) 10:01, 25 September 2025 (UTC)
Wouldn't being forced to wait 2 weeks (or realistically, not waiting 2 weeks and having their article deleted immediately without debate) also drive away a lot of potentially valuable new editors? -- Beland (talk) 14:00, 25 September 2025 (UTC)
I believe they'll be able to make a more level-headed assessment after the dust settles down a bit in that 2 week period.
A lot of them would refrain from creating such low value articles, which they would've created otherwise based on hype.
Ofcourse, some of them might still go on with their articles. But the number would be much lower. Cdr. Erwin Smith (talk) 16:10, 25 September 2025 (UTC)
Regardless of what's done in the short term, AFDs of events that are long since over (like, a month or so) where there have been no significant updates to the article nor non primary sources to support updates, should be handled far better. Those not-votes arguing the burst of initial coverage meets notability should be ignored, because in the long run we are looking at enduring coverage, not a flash of news. Too many AFDs of events with no long term impact are kept because of this argument. Masem (t) 21:07, 22 September 2025 (UTC)
Is your complaint that enough editors don't go around nominating articles for deletion a few months later when events turn out to be non-notable, or that a few months later other editors simply disagree with you about notability, or something else? -- Beland (talk) 23:22, 22 September 2025 (UTC)
I'm not Masem, but I think current event articles are largely based on primary source coverage, so they're not notable in the first instance, and not particularly encyclopedic. I think the issue with deletion discussions is that AfD closers give equal weight to !votes that are "keep, this received coverage in the two weeks after the event" with !votes that point out that GNG actually requires secondary coverage, not a bunch of press about a particular event in the two weeks after it. voorts (talk/contributions) 00:03, 23 September 2025 (UTC)
Something two weeks after can be secondary coverage; whether it's reliable or shows lasting notability is a different matter. (It would be lovely if the FA and GA reviewers didn't say articles were quality when they only relied on coverage from within a couple weeks of events, but hey, I think that proves it's not just an AFD closer problem). GreenLipstickLesbian💌🦋 00:12, 23 September 2025 (UTC)
The optimal solution is: stop editors from creating event articles that have no clear indication of long-term significance in the first hours or days of the event breaking, since the bulk of the news at that time is going to be primary sources, which will not satisfy long-term notability (per WP:N/NEVENT). We do want articles like 9/11 and Jan 6 to be created in the hours that they happen, they clearly had significance near immediately, but many local crimes, traffic accidents, and other things get created too quickly before significance is well known.
That optimal solution is near impossible, we cannot easily stop article creation (even with AFC and drafts that only slows it down). And from a AGF aspect, once created, we shouldn't rush to delete, unless there is clearly a problem (such as its based on unreliable sourcing, is a major BLP issue, etc), and give the creators and editors time to expand. Sometimes, as with the Erika Kirk or the Killing of Iryna Zarutska articles, more context may reveal its notable. However, I'd say at least (if not more) 50% of the time, that just doesn't happen.
So a suboptimal but reasonable solution to weed out event articles created in the rush of a flurry of news that ends up going nowhere is via AFD some time after the event, like at least a month, with a good BEFORE search to verify the event has not had any further coverage nor is likely to have further coverage. What ends up happening moreso right now is that during the AFD for these cases is that editors argue the rush of priamry news coverage at the time of the event satisfies notability, and these articles are kept, making it near impossible to delete. We need to fix that part of the process. Masem (t) 00:20, 23 September 2025 (UTC)
Gotcha. Personally, I think 90% of my edits are either cleanup or adding facts based on news sources. They kind of straddle the boundary between primary - when a reporter is on scene and describing what is happening - and secondary - when the NYT does a roundup of the known facts the next day based on both original reporting and talking to witnesses and government officials and synthesizing everything with context.
I was curious how much our existing articles rely on news sources, so I hit "random article" ten times and looked at references 4 were all news sources, 4 had a lot of secondary sources (books or specialty web sites), 1 was referenced to a species wiki, and 1 was referenced to YouTube. This implies to me that rigorous enforcement of "secondary sources define notability" and defining news sources as primary would cut out ~40% of our content (larger sample size needed), and the majority of AFD participants in practice don't agree with this rule. I would prefer our rules be changed to match what people actually think should be in Wikipedia; changing our content to match the above interpretation of our rules would ditch a lot of valuable material. -- Beland (talk) 02:18, 23 September 2025 (UTC)
On one hand, I don't really think this is feasible to implement — and on the other hand, I don't think it would be a good idea even if it were. I am sort of an optimist about these things: I think that, even if a Wikipedia article is total ass by our own standards, it almost always manages to outpace basically all live news coverage for current events (for free). Typically, when something is in the news, you have two options: a) to read everything about it on Twitter the second it happens, which is time-consuming and involves wading through giant piles of crap, or b) to read the newspaper, which will generally be hours to days behind your unemployed friends who do a), and costs a ton of money, gives a very limited perspective (e.g. what a single reporter/editor care about) and generally isn't very comprehensive. The Wikipedia article on any big event is, in almost every case, vastly superior.
I think this is a genuinely useful thing we do for society. Even if these articles do attract the worst and most pugnacious editors, and even if editing them is a cruel and unusual punishment, I personally benefit a lot from it, and I think it does a lot to help people online figure out what the hell is going on in an increasingly confusing world.
I kind of feel like destroying this would be like when AT&T had the opportunity to buy the entire ARPANET in 1971 and they just said "this is new and weird and we don't get it so nah". jp×g🗯️ 23:47, 1 October 2025 (UTC)
We need some guidance for AFD nominators so that they can assess whether to leave the article alone or to try to delete it. They should consider international reporting, time scale of event as well as political response. Graeme Bartlett (talk) 22:26, 22 September 2025 (UTC)
I think our articles that are created as events unfold are, on balance, a strength of Wikipedia. Overall I'd be inclined to add a speedy keep criterion for articles that are within X days of a major event around the subject of the article, so that we can have the discussion with some knowledge of the discussion or lasting impact of the event. X should probably be small, maybe 7, with no prejudice against a renom once it's up. Tazerdadog (talk) 23:00, 22 September 2025 (UTC)
I can't view the article, but from context I would evaluate CSD criterion G3. Depending on the speed of the sources coming out and how clear what really happened came out we could also keep the article for the week and edit it to reflect the situation as the sources are reporting it at the time. Tazerdadog (talk) 00:19, 23 September 2025 (UTC)
I was watching that article at the time and when it went to AfD, and even edited it; it wasn't G3, it accurately reflected what was written in reliable sources. But it should never have been an article. GreenLipstickLesbian💌🦋 01:11, 23 September 2025 (UTC)
I think I see...AFDs would be allowed to be opened, but "Speedy Keep: Event less than 7 days in the past" would be an option that editors would actually have to advocate for. If people looking at the article actually say "Speedy Delete: G3", bad articles less than 7 days old could still be deleted. Presumably if there was a strong dispute early in the process and no consensus for either speedy keep or speedy delete, a full AFD would run 7 days and we're no worse off in terms of wasted time and reputational risk than we are now. I like this idea. -- Beland (talk) 02:23, 23 September 2025 (UTC)
I definitely like the idea of requiring all articles on breaking news events to be started in DRAFTSPACE (and automatically draftifying those that are prematurely started in Mainspace). This would allow editors to contribute, and yet also avoid NOTNEWS situations. It would give us time to sort the wheat from the chaff of news coverage. Blueboar (talk) 23:47, 22 September 2025 (UTC)
Which leads to a possible specifically new space, News: where current event articles should be created, which should follow all of the core policies outside of the NOTNEWS/NEVENT factor, and which then would have approval processes to move such articles into mainspace after they have been prepared properly for an encyclopedic treated. Maybe require articles to have no more than 14 days to remain in this News: space after which they are deleted if not moved to mainspace by some approval process. There's more gotchas there, but that's a possible route. Masem (t) 00:23, 23 September 2025 (UTC)
So, a sort of intermediate between draftspace and mainspace? In my mind, the question is really whether this space should be reader-facing (and part of the encyclopedia itself) or focus on being a drafting area. The first one might require a (much less strict) notability criterion to avoid it being flooded with local blips in the news, but, in that case having content be deleted after a few weeks wouldn't be ideal.This News namespace actually reminds me, not of Wikinews, but of Wikispore's Event Spore, which could be a good staging area before making them full-fledged encyclopedia articles (or leaving them there if there isn't enough lasting coverage). Chaotic Enby (talk · contribs) 00:34, 23 September 2025 (UTC)
If the worry about negative press is driving this, then for draftification I would worry about: "I'm a conservative and I went to Wikipedia to start an article and was told any article about current events had to go to a special hidden place for liberal editors to decide if readers can see it". And also about the general negative burden bureaucratic procedures put on editor retention. -- Beland (talk) 02:29, 23 September 2025 (UTC)
I’ve said a couple times that there should be some WikiProject or something where articles w only primary sources are draftified, listed, and worked on while they wait for secondary coverage, like an incubator. Drafts are usually just written by one person, part of the motivation to publish early is to put it on someone else's plate Kowal2701 (talk) 13:36, 23 September 2025 (UTC)
The idea of requiring waiting periods and special procedures just won't work. This is a de-centralized wiki. We have a culture that encourages editors to WP:Be bold. People can and will do whatever they think helps Wikipedia, even if a bunch of us sit over in a Village pump – a set of back pages that 83% of WP:XCON editors have never posted to, BTW, which suggests that most editors will never see or care about the views expressed in this conversation – and say "Well, how dare anyone be interested in plebeian things like the biggest news story of this month. Wikipedia editors should have high-minded, refined, academic interests, like art and mathematics, instead of all this vulgar money and grubby politics".
It just won't work. We can issue all the edicts we want, on purple vellum with gold leaf, but the community won't bear it. They'll WP:IAR their way around any restriction on creating articles on subjects they believe to be appropriate. The only thing we have a chance of affecting is whether we keep an AFD open during the first days. WhatamIdoing (talk) 03:13, 23 September 2025 (UTC)
I don't think telling people to put stuff in draftspace would work. My own experience with this, having created a few of these articles myself, is that a lot of people want to be the one to do it. The minute something happens, a bunch of people will rush to create drafts, and then the second news coverage ticks over into an actual published article that you can cite for GNG, one of them will get moved into mainspace. But, in reality, people will tend to just move a premature draft into mainspace to get the jump on all the other drafts: it's just human nature. Even if you think this is a dumb situation, the only reward you get for diligently waiting is that you lose!
All of that to say, I think the only way that this would actually work — and this would be an actual good idea — is to redirect the mainspace title back to the draft, or salt it, or something. This way, it would simply not be possible for anybody to "jump the gun", and meanwhile the article would be clearly identified as a draft during the hectic first hours or days when not much is known about an event.
I think this might even go some way to improving the public perception of Wikipedia, since a clearly-identified draft article doesn't carry with it the implicit guarantee that most of our pages do (and it conveys pretty clearly the idea that it's not perfect and will change a lot prior to things being settled). jp×g🗯️ 23:55, 1 October 2025 (UTC)
I agree that the problem this is highlighting is in need of solving, but GreenLipstickLesbian's comment above does a good job highlighting situations where AFD is necessary for breaking news articles. In general, the proposed solution would probably be difficult to implement without a flood of WP:NOTNEWS articles that get created and then can't be deleted for a week. Draftifying for a set amount of time does seem like a good solution for current events of mid-to-low importance, but a bad solution for major news articles. Assassination of Charlie Kirk is a recent example of an article where it would've been detrimental for us not to have an article before a week after the event. I think it would help to have a stronger emphasis on the "wait" option at AFD. The best current event AFDs are snow closes, so we don't have a ton of eyes on an article with a deletion tag, and in an ideal world, we could just snow "wait" or "delete" every borderline current event article. That's more of a cultural change though, not something easily changed through policy. 🌸wasianpower🌸 (talk • contribs) 00:36, 23 September 2025 (UTC)
I really like that idea of more prominent a "procedural close/wait" option, actually; if a borderline article goes through AfD while the event is ongoing as is kept, then that makes second AfDs more complicated. My main worry with a "consensus to move out of draftspace" idea is that the poor-quality arguments will still be made, but in greater number and with more fervor as people try to read about events they've heard about (tiny earthquakes felt by major American cities, small airplane crashes in South East Alaska, North American weather), and most closers will not ignore them. Then it will become much harder to delete the article in the next 18 months when we discover that coverage didn't persist after all. GreenLipstickLesbian💌🦋 02:08, 23 September 2025 (UTC)
I think a "procedural close/wait" option should exist, and go in the direction of keeping the article. Neutral, well-sourced articles on current events are one of the things that our readers value. IMO we shouldn't be trying to delete them in the first three or so days. WhatamIdoing (talk) 02:53, 23 September 2025 (UTC)
I agree with the idea of a procedural close (especially for non BLPs). - Enos733 (talk) 16:11, 25 September 2025 (UTC)
Well, we shouldn't be writing "news articles", in the sense of original reporting based on first-hand reports from Wikipedia editors, Wikipedians interviewing witnesses, presenting information that has never been published anywhere else ("news", as traditionally and humorously contrasted with "olds"). But also, we don't do that.
We instead write encyclopedia articles about historical events, some of which are still developing events. I think it's important to look past the WP:UPPERCASE and read what that policy says: "all Wikipedia articles should contain up-to-date information. Editors are also encouraged to develop stand-alone articles on significant current events...Wikipedia does have many encyclopedia articles on topics of historical significance that are currently in the news...breaking news should not be...treated differently from other information". Excluding "breaking news articles" would be "treating it differently", which the policy forbids. WhatamIdoing (talk) 02:59, 23 September 2025 (UTC)
Most of the articles about "historical events" are of events that no one will care about in 2 weeks. LilianaUwU(talk / contributions) 03:03, 23 September 2025 (UTC)
I have the same experience. Special:Homepage lists some stats on page views, and it's discouraging. I fix a template somewhere, or blank some trivia, and suddenly my "impact" is all about some celebrity I've never heard of. But work that I think will actually matter, like telling readers whether "We're at the hospital with the baby, and the doctor says it's scaryitis" is good news vs bad news, that doesn't get picked up by the metrics. WhatamIdoing (talk) 03:40, 23 September 2025 (UTC)
You can set filters in Special:NewPagesFeed to get all the articles from one day. It lists 601 non-redirect articles that were created on 8 September 2025. Are you aware of any in that list that's about a current event that no one cares about today? I only saw articles about past events when I scrolled through. (Which reminds me: Erika Kirk is an article about a person, not about an event.) WhatamIdoing (talk) 03:36, 23 September 2025 (UTC)
Anyways, y'all don't tell me that the Charlie Kirk shooting is worth seven articles. Maybe one for the event and one for the shooter AT MOST. LilianaUwU(talk / contributions) 18:46, 23 September 2025 (UTC)
Yeah. And people won't do shit about it because consensus is required to get rid of the most egregious violations of WP:NOT. LilianaUwU(talk / contributions) 19:16, 23 September 2025 (UTC)
There's no need to use four-letter words in what's supposed to be a civil discourse here. I do see a number of articles in Category:Assassination of Charlie Kirk, though some of them are about the major attack on free speech in the United States that happened afterwords. I see you have already objected to Memorial service of Charlie Kirk; Reactions to the assassination of Charlie Kirk also seems like it might have excessive detail. Merging into Assassination of Charlie Kirk may require trimming details to avoid an overly-long article, though many are redundant to what's already there. If you want action to be taken on this content, my recommendation would be to start a formal merge discussion following the procedure at Wikipedia:Proposed article mergers to get a wider sampling of opinion. -- Beland (talk) 20:14, 23 September 2025 (UTC)
And my recommendation would be to wait at least a few months, when cooler heads have a better chance of prevailing. WhatamIdoing (talk) 21:52, 23 September 2025 (UTC)
Agreed. These things stop being news after a while. Secretlondon (talk) 22:12, 23 September 2025 (UTC)
Reaction sections are perhaps the worst offenses when it comes to most event articles today. Most reactions that are included are, for all purposes, empty words said in the short term that give no indication of any possible action or response to the event at hand (that is, stuff that is akin to the "thoughts and prayers" after a shooting in the US). They are easy fruit to pick from the news sources and fluff out the article to make it appear longer than needed (and then we often get MOS:FLAGS spam when international reactions are added). With the Kirk shooting, there are several valid reactions that actually have created actions and responses (Trump's antifa EO for example). Reactions based on long-term analysis of an event to place it in context are also fine, but we're not going to know those in the first few days/months of an event; those are usually going to come from academics far down the line. Masem (t) 01:01, 24 September 2025 (UTC)
Are you proposing a new guideline for reaction sections, similar to MOS:POPCULT? -- Beland (talk) 01:25, 24 September 2025 (UTC)
I'm not sure whether we need a new guideline or to best better enforce current ones, but some exceedingly trivial reactions do get added. As an example, following a plane crash (possibly Med Jets Flight 056) the local high school sports team posted a generic condolences message on what used to be Twitter, and that got added to the article. That was one of the most extreme I recall, but generic "thoughts and prayers" type messages from random politicians and celebrities with no connection to the event beyond being alive when it happened accumulate in reaction sections more often than not. Thryduulf (talk) 01:56, 24 September 2025 (UTC)
Yeah, then maybe the thing to do is for anyone who is unhappy about the detailed reactions is to trim on sight, or better yet search for "Reactions" in section headers and trim systematically. -- Beland (talk) 02:01, 24 September 2025 (UTC)
I think that a whole guideline would be overkill, but we should have a paragraph somewhere that discourages congratulations/condolences/thoughts and prayers/nice sounding stuff said by politicians and celebrities.
After a disaster, what we want is "Paul President deployed the military" or "Near Neighbor sent a team of experts". What we don't want is "Sal Slacktivist said something saccharine on social media". WhatamIdoing (talk) 03:47, 24 September 2025 (UTC)
Something like this might fit at NEVENTS, that sections that catalog short-term reactions that lack any actual action (planned or completed) should be avoided/minimized as such info doesn't contribute towards the long-term notability of an event. But I also can envision a better guideline on how to write event/breaking news articles that doesn't focus on notability. Masem (t) 03:54, 24 September 2025 (UTC)
We have Wikipedia:Reactions to... articles which is an essay that focuses on articles, but much of what it says also applies to sections. WP:QUOTEFARM would also be good place for a bullet or specific to reactions. Thryduulf (talk) 11:17, 24 September 2025 (UTC)
Maybe if we make sure there's something about this (with an example?) in each of these three places, then eventually the good advice will get found by people working on these articles. WhatamIdoing (talk) 16:40, 24 September 2025 (UTC)
I hate to be the Negative Nancy who shoots down an idea and doesn't offer a better one, but I don't see this being implemented. I think the same issues you mentioned will probably be present after a week (as few news stories are single-day). Articles that do not deserve to be created will be immunized from deletion when the most eyes are on it. When the event passes, perhaps it doesn't have enough readers to even AfD, leaving a non-notable subject on the encyclopedia.
Again, maybe I can be sold with some revision to the idea, but I just don't see it working in its current form. Bremps... 00:53, 12 October 2025 (UTC)
And the use of the {{Breaking news}} template itself may become a tug of war between deletionists and inclusionists, which would be very bad. Bremps... 00:54, 12 October 2025 (UTC)
I don't see how applying {{breaking news}} would prejudice an article in either the "keep" or "delete" direction? -- Beland (talk) 06:22, 12 October 2025 (UTC)
If we say "don't AFD newly created breaking news articles for X days", then someone is going to get the idea that 'a breaking news article' is defined by the presence of the {{breaking news}} template instead of by real-world events. So if I want to AFD it, then the first thing I do is edit-war the template off the top of the page, and now that it's no longer tagged as breaking news, that means I can AFD it without following any pesky rules about not AFDing breaking news articles. WhatamIdoing (talk) 06:32, 12 October 2025 (UTC)
I don't see how anyone would arrive at that illogical conclusion, nor if attempted would it work, so I don't see why they would attempt it a second time. It looks like the other discussion is leaning toward allowing editors to vote to pause the discussion for 14 days, and the presence or absence of a template isn't going to prevent anyone from voting to pause. -- Beland (talk) 06:40, 12 October 2025 (UTC)
Prefacing this by saying I don't think this has a serious chance of being enacted and that is for the better. Nonetheless, discussing this might help contribute to other, better ideas in the future. The idea rests on two assumptions which I'm unconvinced by, but trust the community's got a better finger on the pulse of those matters than I do. Firstly, that there is a significant need for more aggressive admin recruitment. The recent admin elections driving more candidatures (from what I've seen) and my own work with User:Ixtal/Analysis on administrators' activity does not really support that, but it might be the case that a temporary uptick in candidatures thanks to the election system is just a transitive phenomenon. Secondly, that exposure to the back-end of administrator activity in general as opposed to exposure to exclusively enforcement or sanctioning administrator activity is a productive way of engaging future recruits. A lot of new editors that stay involved in the Wikipedia space are exposed to administrators and our PAGs through editor conflict resolution noticeboards like WP:3RRN and WP:ANI. This helps drive the idea that admins are more cops than they are janitors, which I think is a highly unattractive proposition.
Since I involved myself in Wikipedia:Closure requests, I've been subscribed to the Admin's newsletter in case any major PAGs are changed that might affect future closures. It has been very helpful even as a non-admin. I was wondering how good of an idea it would be to have a similar newsletter to, for example:
Administrator elections will take place this month. Administrator elections are an alternative to RFA that is a gentler process for candidates due to secret voting and multiple people running together. The call for candidates is July 9–15, the discussion phase is July 18–22, and the voting phase is July 23–29. Get ready to submit your candidacy, or (with their consent) to nominate a talented candidate!
Note the difference from the admin's newsletter is the permission change announcements have been removed. When an editor gains the EC permission, they'd get an automated message explaining the newsletter and encouraging them to explore the Wikipedia space to involve themselves more with the community. What do y'all think? I know it's a very wild idea but I think it's neighboring a good idea. — ♠ Ixtal( T / C ) ⁂ Non nobis solum ♠ 09:30, 9 October 2025 (UTC)
I really like this, and, while it makes sense to give an automated message to editors at one point (which could be the EC threshold), maybe we could have other ways for editors to find this newsletter before becoming EC? Say, having a subscription link on WP:CENT, or on the noticeboard template. I like the idea of new editors getting a feel of how the project works "in the back-end" without needing a specific edit threshold or user group. Chaotic Enby (talk · contribs) 09:37, 9 October 2025 (UTC)
I like the EC threshold because it matches our minimal standard to edit protected articles. If I cannot trust an editor with a protected article, I'd be hesitant to encourage them to participate actively in the backend of the project. Not that all ECs are trustworthy and all trustworthy newcomers are ECs, but I don't see EC perms as a high bar to reach. I do think having the newsletter be easily discoverable is a good idea and that hiding it away would be self-defeating, but since it'll be in Wikipedia space anyhow, any non-EC editors that discover it are likely becoming ECs soon enough. You raise a good point on the automation, Chaotic Enby, which I hadn't considered but do agree having it work with the EC threshold would be a good way to go about it. — ♠ Ixtal( T / C ) ⁂ Non nobis solum ♠ 09:58, 9 October 2025 (UTC)
This is actually a really good idea. If this goes to an RfC, I'd support making it opt-out rather than opt-in to maximize the people it reaches. We already make it hard for people to find information, we don't need to make it harder with a contrived subscription process. Toadspike[Talk] 09:51, 9 October 2025 (UTC)
Opt-out seems the only way for something like this to succeed, imho. — ♠ Ixtal( T / C ) ⁂ Non nobis solum ♠ 09:58, 9 October 2025 (UTC)
I'd prefer a one-time opt-in system, similar to the way the first administrators' newsletter was, where the user receives such a newsletter the first time they reach their edit/time threshold but their "first-time" newsletter has a notice saying "to continue receiving this newsletter, please subscribe here", etc. My angle is that (a) I strongly dislike talk pages that are filled with random newsletters for barely active editors ... and it just simply wouldn't scale if the newsletter runs for say twenty years like the Signpost has, (b) it'd make it more difficult to assess talk pages of, say, unresponsive/disruptive editors who've gotten many warnings but have managed to make it to extendedconfirmed anyway if there talk page was filled with newsletters they never read, (c) some people would see regular emails/notifications like this as spam, and (d) I sometimes unsubscribe deceased editors from newsletters to avoid errant log entries like this, and that'd be one more newsletter to check. Also, I think it's worth noting that, at the time of writing, we have 76,816 editors in the extendedconfirmed user group, per the statistics special page, compared to 1,561 users on the Signpost subscription list (my screen reader tells me the number of items in HTML lists like that one). Graham87 (talk) 16:38, 9 October 2025 (UTC)
Is it possible to generate a list of "ECP with >1 edit in past 12 months" (or something like that) to post the newsletter to? Schazjmd(talk) 16:42, 9 October 2025 (UTC)
Yes, that'd be quite possible. Details would have to be hashed out and a database query would need to be run to get the relevant info. Graham87 (talk) 17:13, 9 October 2025 (UTC)
Before any deployment of a signup process, I suggest doing a test run of a few issues to work out the kinks of production, establish a regular cadence, and determine scope. For instance, I think the appropriate degree of overlap with Wikipedia:Community portal should be considered. If an opt-out model is going to be adopted, I think it's important to demonstrate that there are enough people involved in production to sustain the initiative for the foreseeable future. Good luck! isaacl (talk) 18:47, 9 October 2025 (UTC)
Instead of creating a newsletter (most people find creating newsletters to be a boring slog), why not put this in Wikipedia:Wikipedia Signpost, or (as has been successful for the Signpost) put each item individually on the MediaWiki:Watchlist-messages as they come up? WhatamIdoing (talk) 19:07, 9 October 2025 (UTC)
I think attaching this to the Signpost would be both to the Signpost's detriment and to this idea's. I do think that whatever solution we'll come up with will struggle with the required buy-in to get it up and running. If none volunteer to run it as a newsletter, we'll have to come up with some better alternative. What I proposed is similar enough to the admin newsletter's current formulation that it might be possible to ask whoever runs that if they'd be willing to do this too. — ♠ Ixtal( T / C ) ⁂ Non nobis solum ♠ 20:59, 9 October 2025 (UTC)
For admin news, the main person who runs it is DreamRimmer. 45dogs (they/them) (talk page) 06:47, 12 October 2025 (UTC)
I'm not sure what scope is being considered for the items, but I can't imagine it's so narrow that watchlist notices are an appropriate communication method. It could be a Signpost column, though the overlap with the News and Notes column and the village pump roundup would have to resolved. I agree with Ixtal that it should be carefully considered if the audience for this idea would be willing to subscribe to the Signpost (maybe they just want the editing ecosystem changes and nothing else). isaacl (talk) 22:20, 9 October 2025 (UTC)
This sounds really useful. If I miss a P/G RFC when it's on CENT it's hard to know it ever happened. —Rutebega (talk) 13:27, 14 October 2025 (UTC)
I think the monthly newsletter idea is fantastic, and should be automatically opted in for all autoconfirmed users. It might not be in that form, but any kind of newsletter for Wikipedia would be great (that isn't the Signpost).
Not sure how much this would dip into WP:CANVASSING, but another section for very important RFCs that have site-wide implications is worth thinking about. That could be for extended confirmed users, because tbh autoconfirmed typically don't have enough experience for important RFCs EatingCarBatteries(contributions, talk) 01:45, 15 October 2025 (UTC)
I recently came across a very cool project - Rotate the World by Jason Davies. I believe these interactive globe maps would be well-suited to replace the existing globe projections currently in use on various infoboxes throughout Wikipedia, in articles such as United States. In addition to allowing desktop users to move the globe around, the zoom function would work well with the existing pattern used on articles such as Germany, where the user is presented with a range of zoom options (in Germany's case, global > European Union). Potentially, it could even be used to expand articles such as Lower Saxony by allowing the user to zoom *out* from the default view. DeklinCaban (talk) 14:01, 14 October 2025 (UTC)
