Warning: file_put_contents(/opt/frankenphp/design.onmedianet.com/storage/proxy/cache/990df61ba733a5a00a652a07c4fe8909.html): Failed to open stream: No space left on device in /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php on line 36
Warning: http_response_code(): Cannot set response code - headers already sent (output started at /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php:36) in /opt/frankenphp/design.onmedianet.com/app/src/Models/Response.php on line 17
Warning: Cannot modify header information - headers already sent by (output started at /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php:36) in /opt/frankenphp/design.onmedianet.com/app/src/Models/Response.php on line 20 Wikipedia:Administrators' noticeboard - WikipediaJump to content
WP:AFD is in need of more participation. In recent weeks the number of discussions that have languished for lack of attention from !voters is considerably larger than usual in my experience as an AfD closer. Indicators of this include the number of discussions admins have allowed to remain in the "old" section in the hopes of attracting !voters; the number of discussions in which regular closers have chosen to participate instead of closing; and the proportion of discussions being relisted. I have theories as to why this is happening, but those are besides the point: the solution is clearly more engagement. I imagine other regular closers would agree with me. Vanamonde93 (talk) 17:33, 12 September 2025 (UTC)[reply]
I've been spending quite a bit of time in AfD over the last week or two (because I have been participating in the NPP backlog drive), although I'm not usually a regular. I've been struck by the need for more participation too. But also, couldn't the admins and other closers potentially make AfD a little bit more efficient just by taking a slightly more hardline approach?
What I mean is this:
If an article is nominated and nobody comments on the listing, it should always be soft-deleted as though it were an expired PROD (rather than being relisted, as sometimes happens at present: e.g. here, to take a random example).
If an article is nominated and attracts one comment in support and no objections, this consensus of two should usually be treated as consensus for a hard deletion (rather than the item being relisted or soft deleted).
If an article that has previously been tagged for PROD but has had the tag removed, or that has been soft deleted or deleted through PROD and then restored on request, is nominated by someone other than the person who nominated or tagged it before, and if nobody comments on the nomination, this should usually be treated as consensus for a hard deletion (rather than the item being relisted as ineligible for soft deletion), since this situation implies that at least two people have considered the deletion justified and no-one has set out any rationale against it.
If an article is nominated and attracts one comment in support and one in opposition, the closer should give serious consideration to treating this as consensus for deletion (rather than relisting), if on the face of it the arguments against deletion do not seem to have much weight or are not based on policy (as sometimes seems to happen, especially when the opposing editor is the page creator).
If participants in AfD knew that articles would usually be deleted under the circumstances outlined above (albeit not invariably, since of course the closer must still ultimately use their judgement!), this would not only reduce the number of relistings, but would also mean that AfD participants would not feel the need to spend time commenting on proposals that are unlikely to be opposed or that have been looked at by two editors already (as e.g. here, to take a random example), and could instead spend their limited time looking at listings where a rough consensus of two or three editors hasn't been reached yet.
Probably someone will now say that I'm not very experienced at AfD and therefore don't know what I'm talking about. But still, it seems to me that AfD is operating as though maximum participation and consensus were the priority, which would be great if there were loads of participants, but I think perhaps it would be better to consider operating slightly more in the interests of efficiency and saving editors' time, along the above lines. Dionysodorus (talk) 21:13, 13 September 2025 (UTC)[reply]
@Dionysodorus: Thoughtful comments on the process are welcome. I would at least agree that in discussions with low participation, looking at the weight comments should get can sometimes show a consensus where raw numbers might suggest a relist. However, a bare minimum of participation is necessary for the process to be meaningful. I usually require participation from three editors to find a consensus for anything. Expanding the soft deletion process as you propose may be a way around that. It would require a community consensus to implement, but you could propose such an implementation. Vanamonde93 (talk) 21:56, 13 September 2025 (UTC)[reply]
I'm not sure that a policy change ought necessarily to be needed for closers to treat unopposed nominations as soft deletes, since an unopposed nomination meets exactly the same threshold as a successful PROD (which is in itself nothing more than an unopposed tagging): indeed closers sometimes do this already, just not consistently (e.g. here and here). And is there actually anything in existing policy to prevent closers treating 2 vs 0 or 2 vs 1 as a consensus in cases where the prima facie case for deletion is strong and arguments against deletion are not expressed or transparently weak? I think that would be entirely in the spirit of WP:CLOSEAFD and WP:NHC.
So all I really meant to suggest was that you closers could be a bit bolder in closing discussions, not necessarily that we need a policy change. If you think there's any merit in what I'm saying, perhaps the ideal starting point would be for you closers to discuss it relatively informally (which could then either lead to an informal change of approach or to the proposing of more formal guidelines to the community), rather than me trying to create policy saying that you should proceed in a certain way, especially if such a change of policy might not be needed to do some of this. (If it would be useful, we could even ping the regular closers and have such a discussion here.) Dionysodorus (talk) 07:23, 14 September 2025 (UTC)[reply]
I'd argue that if people are closing 'no participation' AfDs (after one or two relists, natch) as 'no consensus' then that is arguably inappropriate. There is absolutely no need to change any policy to close an undiscussed, uncontested AfD as a softdelete. Just relist it twice, and if nobody discusses it, hit 'close'. - The BushrangerOne ping only22:27, 14 September 2025 (UTC)[reply]
But what I'm saying is that uncontested nominations should always be closed as soft delete after one week without any relisting (as only occasionally seems to happen at present). When an article is PRODded it gets deleted after a week without delay, so there's no reason why uncontested AfD nominations shouldn't be soft-deleted after a week too. In my view, the unnecessary relisting of nominations should be avoided, because commenting on these unnecessary relistings takes up the limited time of the relatively small number of AfD participants and distracts them from commenting on cases that actually require discussion, which in turn contributes to the problem that Vanamonde raised at the start of the thread. Dionysodorus (talk) 23:50, 14 September 2025 (UTC)[reply]
I would tend to agree. If you look at the various processes we need more folks to help, marginal notable articles are not the most essential part of Wikipedia. More soft deletes will lead to a higher share of 'mistakes', but they're easily reversible. Editor time is precious. That seems to be in line with current policy, where closers have this discretion: WP:NOQUORUM: Closing an unopposed XfD nomination under this procedure does not require the discussion to have been relisted any particular number of times.—Femke 🐦 (talk) 18:22, 15 September 2025 (UTC)[reply]
At it happens I disagree with this. @Dionysodorus and Femke: There's a considerable potential distinction between soft deletion and regular AfD deletion: the former is reversible by any editor requesting it in good faith, whereas the latter mandates at the very least recreation that is substantially different from the deleted version, and in practice usually requires additional sourcing in order to not face the same outcome at AfD. We softdelete articles after an AfD with no participation because it is functionally equal to a PROD. An AfD with 2 or 3 editors opining "delete" has received additional scrutiny from 1 or 2 editors, and a SOFTDELETE closure negates their participation. That said, that's a theoretical problem. If someone could show that SOFTDELETEd topics are not subject to recreation significantly often - or no more than regular AfD outcomes - that would negate the issue, and if we had such data I imagine adjusting our practice much more palatable. Vanamonde93 (talk) 18:47, 15 September 2025 (UTC)[reply]
If things are working as intended, there should be a higher recreation rate for softdeleted articles. The more interesting question is to look at the rates of recreation in either category, and estimate the amount of editor time saved / spent if we were to change practice here. I should probably start a list of 'research questions' for Wikipedia.
It seems like there is almost one request at WP:REFUND a day to restore a soft deleted article. No idea how many articles get soft deleted daily and what the time commitment is at REFUND. —Femke 🐦 (talk) 19:25, 15 September 2025 (UTC)[reply]
re: "If things are working as intended, there should be a higher recreation rate for softdeleted articles", that's precisely why expanding the scope of soft deletion to include topics that editors besides the nominator have declared to be non-notable is potentially a problem. Vanamonde93 (talk) 20:29, 15 September 2025 (UTC)[reply]
I think you may have slightly misunderstood what I am suggesting, Vanamonde93. My problem is with the fact that at present nominations that attract no participation commonly aren't being soft deleted after a week (as the equivalence with PRODs would suggest, and as your comment at 18:47 seems to suggest you agree they should be), but rather are being relisted. I'm not saying that soft deletion should be expanded to cover anything that is currently hard deleted: on the contrary, I suggested in my initial comment above that nominations that attract even only one supporting !vote (which tend to get soft deleted at present) should be hard deleted. Dionysodorus (talk) 21:00, 15 September 2025 (UTC)[reply]
@Dionysodorus: I don't believe I misunderstood: I focused on your idea of expanding SOFTDELETE because that felt like a substantive proposal, whereas AFAIK we do regularly handle nominations with no participation as we should. I know I do. Are you aware of a considerable volume of SOFTDELETE-eligible articles that aren't being so deleted? Remember that PRODs and previous AfDs preclude soft deletion. Vanamonde93 (talk) 21:13, 15 September 2025 (UTC)[reply]
@Vanamonde93: But I didn't propose any expansion of soft deletion in the first place, except inasmuch as I suggested that any nominations with no participation should be soft deleted rather than relisted (as I think does sometimes happen: I cited this example above, but I'll have a look for others).
Apart from that, all my suggestions were about reducing relisting in favour of hard deletion: that is to say, I think that nominations that result in 2 vs. 0, 2 vs. 1 where the opposing arguments are obviously weak, or even unopposed nominations that are ineligible for soft deletion because they have been soft deleted or PRODded before, should all be hard deleted after a week without relisting (not soft deleted!). Dionysodorus (talk) 21:23, 15 September 2025 (UTC)[reply]
P.S. @Vanamonde93: In today's log so far, there are three relistings of nominations with no participation:
None of these articles appears to have been previously soft-deleted or PRODded. I think perhaps you would be able to get the same result by looking at the log for most other days. Dionysodorus (talk) 21:33, 15 September 2025 (UTC)[reply]
I am one of the admins who generally relists AFDs with no participation rather than soft deleting the article. I do that for two reasons, one more philosophical and one more practical. The more philosophical one is that policy ordinarily requires a consensus to delete an article and while it's true that policy also recognises PRODs and soft deletions – which are exceptions to that rule –, I find it more in keeping with the spirit of the deletion policy to relist a discussion with no participation, in hopes of attracting more attention to it, so that a consensus can form – which does indeed happen, such as here. After all, we are deleting someone's work, I think it only fair to have a full discussion before doing it, unless there is something in the article requiring urgent attention. The more practical reason is that soft deletion can end up creating more work. Anyone can contest it at any time and, then, the article has to be recreated and, if truly unnotable, has to be nominated again. So, trying to see if a more thorough discussion can be had now, in my opinion, can save time later. However, if it turns out that the general feeling of the community is that relisting a discussion instead of simply soft deleting it is a waste of resources, I have no problem soft deleting articles. But I'm not sure AN is the best place for this discussion... — Salviogiuliano08:57, 16 September 2025 (UTC)[reply]
I think it just has to come down to admin discretion. I have no trouble soft-deleting when the situation calls for it, but sometimes it's not what makes the most sense: for instance, I just finished relisting this one (where the article was brand-new and any soft deletion would very likely be challenged) and this one (where the nominator wanted other viewpoints and had specifically chosen not to use PROD). I don't have an issue with nudging things a little closer to the soft-deletion end of the spectrum, but a bright-line rule wouldn't be a good idea. Extraordinary Writ (talk) 22:11, 16 September 2025 (UTC)[reply]
I disagree that we can't close an AfD as soft delete if the article has previously been PRODed or brought to AfD. WP:NOQUORUM says that "the discussion may be closed at the closer's discretion and best judgement". It also says that "[c]ommon options include, but are not limited to" relisting, closing as no consensus, "closing in favour of the nominator's stated proposal" (e.g., hard deletion), or "soft deleting the article". voorts (talk/contributions) 22:23, 18 September 2025 (UTC)[reply]
User:Star Mississippi and I have been talking about a decline in participation in AFDs that we've noticed over the past 2 1/2 years now. I don't think there is just one reason why the decline has happened but I know we have lost a lot of subject matter experts, unfortunately. And I think, this is a guesstimate, that we've lost a lof of inclusionists. It's hard when you tend to argue "Keep" to bust your butt looking all over for sources and the consensus STILL being to delete an article. I think those folks, after a while, just thought that their energy would be better spent elsewhere and left after months of frustration. I do know that it's a whole lot easier to be a deletionist as they usually don't have to provide a justification for their arguments. If you scan down the daily log, it's easy to come across a lot of discussions that are just a straight list of "Delete", "Delete", etc. My own perspective is in the middle, working in AFDLand for five years now has exposed me to a lot of junk articles that have been written over the past 24 years and it's good to clean this clutter out. But I can't help but notice a high burnout ratio for participants who tend towards the "Keep" end of the spectrum.
Looking at the participants has always been my approach when discussing the situation at AFDLand which has been a problem since about 2022-23. I have never thought of approaching this problem by changing our threshold of what qualifies as a Delete, or Soft delete. I disagree with some of the opinions brought up but it's great for us to be having this discussion. LizRead!Talk!01:59, 17 September 2025 (UTC)[reply]
This is all very interesting, and I am grateful to Salvio Giuliano and Liz for explaining their approaches. At the expense of possibly testing everyone's patience, it does occur to me that there are also other more radical ways in which deletion processes could be reformed to reduce the number of listings that required discussion, if low participation were felt to make this desirable (although perhaps the problem isn't so severe as to require this at all).
For instance, what if we got rid of PROD and replaced the two-tier system of PROD and AfD with a two-stage process? It could be the case that all nominations for deletion (except for speedy deletion) took the form of AfD-style listings in a "Preliminary AfD": unopposed nominations, or nominations opposed only on clearly insubstantial grounds, would generally be soft deleted after a week; but any listing that any editor on reasonable grounds opposes (or thinks needs discussing more fully: this could include the nominator) would immediately be moved to regular AfD for fuller discussion; and any listing that the deleting admin doesn't think should be deleted (or thinks requires fuller discussion) could also be moved by them to regular AfD at the end of the week. Also, requests to undelete soft-deleted articles could be reformed in such a way that they would require a reasoned response to the original nomination, and in such a way that the original nomination and the objection would then immediately be posted at regular AfD for discussion, where consensus could then emerge to keep the article permanently or to delete it permanently.
If we had a system like that, uncontroversial listings wouldn't end up in regular AfD at all, and participants there would spend their time looking at actually controversial cases rather than just adding a third, fourth or fifth "delete" !vote to an article on a subject that clearly doesn't satisfy notability. Also, it would become impossible to contest a PROD or to undelete soft-deleted articles without reasoned grounds, which would eliminate the problem that Salvio Giuliano mentioned of people reviving articles that have been just deleted and thereby necessitating a whole new nomination: everything would be streamlined, because the discussion would always be kept in one place rather than sometimes being spread across a PROD tag, an undeletion request, and sometimes more than one AfD thread. Obviously anything of this kind would require community consensus in a more suitable venue than AN, and I put forward this suggestion simply for the sake of throwing ideas around and in case anyone finds it interesting.
I do agree with Extraordinary Writ that it is essential that any approach should allow for the closer to use their discretion: when I said "always", I should have said "always except if there is a good reason why not". I also kind of agree with Salvio Giuliano, in that ideally everything should be discussed in detail and on the basis of consensus: but nothing I'm suggesting would prevent even a single reasonable objection from prompting a fuller discussion. Dionysodorus (talk) 22:33, 17 September 2025 (UTC)[reply]
but any listing that any editor on reasonable grounds opposes (or thinks needs discussing more fully: this could include the nominator) would immediately be moved to regular AfD for fuller discussion This would increase the number of AfD discussions. I've made PRODs that have been declined that I didn't bother taking to AfD after because I was satisfied with the dePRODing editors' response. voorts (talk/contributions) 22:27, 18 September 2025 (UTC)[reply]
Also, requests to undelete soft-deleted articles could be reformed in such a way that they would require a reasoned response to the original nomination, and in such a way that the original nomination and the objection would then immediately be posted at regular AfD for discussion, where consensus could then emerge to keep the article permanently or to delete it permanently. I'm not opposed to this. voorts (talk/contributions) 22:28, 18 September 2025 (UTC)[reply]
@Voorts: I would have thought the two things (requiring justification for contesting any proposed deletion, and requiring justification for requesting any undeletion) logically ought to go together. Also, I'm sure editors do occasionally PROD an article and then find themselves convinced by the rationale of the person who removes the tag, but surely that can't happen so often as to make a significant difference to the numbers that would end up in AfD? I would have thought any such increase would be more than offset by the fact that many pages currently end up in AfD because a PROD tag has been removed without sufficient justification, whereas on my scheme only pages where the objector can provide a justification would ever end up in AfD.
In any case, the nominator could easily withdraw the nomination if convinced by the objection (as sometimes happens at AfD as it is), and so such cases need not actually take up AfD participants' time to speak of. Dionysodorus (talk) 07:39, 19 September 2025 (UTC)[reply]
Comment - I have been participating in AfD in recent weeks, I'm not hugely experienced. But I do see there is a problem here. There are just too many deletes and not enough regular participants. What I'm seeing is one or two deletionist editors getting into a somewhat one-sided debate with a new editor, and the cards get stacked in favour of delete. The danger here is that this process is supposed to give due diligence, and yet that isn't happening, almost the opposite. My vote of sympathy to those who do Closing, it must be dispiriting to see the same nomination on a revolving door. I certainly agree with the point that if you start with an ATD mindset then it can be pretty soul destroying, as Liz has said, and indeed you think "why waste my time on this?" Yes we need more participants, but in the absence of that there is a logical necessity: reduce the number of delete nominations to ensure due diligence actually happens.
Incidentally there is a critical difference between BLP and historical articles. If a BLP gets scrubbed there is a very good chance it will be (correctly) reinstated if GNG eventually comes in place. TOOSOON duly sorts itself out. But deletions on historic items are much. uuch more troubling in my view. The chances are the material will be permanently lost, off the back of changing emphasis on sourcing. ChrysGalley (talk) 09:49, 26 September 2025 (UTC)[reply]
New editor here who takes an interest in article winnowing.
I'm not qualified to comment on how admins close AfD's (though, for what it's worth, no objections to encouraging admins to make more use of soft deletes), but I can fully see how AfD could become a deletionist echo chamber. I also unfortunately don't see a way to stop this due to the very nature of the process, where the burden of proof is on Keep !voters. In an ideal world, the burden of proof would be on the editor who creates the article as part of verifiability requirements, because I see the GNG as simply a practical matter of "if no reliable sources discuss this topic, how are you going to write a verifiable article?"
Unfortunately, we don't live in this ideal world, and I don't see any way to fix it apart from superficial fixes, like giving out awards for demonstrating the notability of topics, stuff in the scope of WikiProject Notability. The 'damage' from large amounts of unreferenced articles has been done, and the best course of action might just be to introduce non-systemic changes through a process like this one for the GA backlog to keep AfD useful for as long as possible.JustARandomSquid (talk) 13:03, 1 October 2025 (UTC)[reply]
Have you considered making an attempt to articulate what actually is wrong with the IP edits? It doesn't seem that you've done that either here or at AIV. 128.164.177.55 (talk) 13:59, 17 September 2025 (UTC)[reply]
I came here to request an independent assessment by an Administrator. I expect you've looked at the examples, given by other editors, in the warning templates they have placed on the IP's Talk page. I would have thought that edits such as this one would be pretty self explanatory. Furthermore, in the meantime, I see that User: Criticize seems to have rolled back the majority of the IP's edits. But if everyone else thinks there is no problem with edits like this, and no further action is required, I don't intend to waste any more of anyone's time on this. Martinevans123 (talk) 14:25, 17 September 2025 (UTC)[reply]
I guess that's a "no" then. Let me suggest that, in the future, if you want someone to do something about a problem, it is helpful if you explain what the problem is instead of expecting people to agree with your unarticulated views. Lots of people know nothing about Jimmy Saville and whether his sexual abuse scandal would merit categories like Category:Royal scandals in the United Kingdom and Category:Margaret Thatcher; two or three sentences of context and explanation would vastly increase the clarity and potential audience of your message. 128.164.177.55 (talk) 14:41, 17 September 2025 (UTC)[reply]
I glanced at a few edits — adding Category:Jimmy Savile to articles about shows and events in which he participated — and I'm tempted to agree with the IP's action. The only potential ground for disagreeing with them, in my mind, is whether his involvement were significant enough to warrant the category; it's definitely not a matter of vandalism. Nyttend (talk) 20:12, 17 September 2025 (UTC)[reply]
I didn't look at those ones; my comment was limited to the ones I glanced at, which were shows and events in which Savile participated. Putting the Thatcher template on the Savile scandal article goes against the idea (sorry, no link; I can't find it) that you shouldn't generally put navboxes on pages that aren't linked by the navbox, but such an edit isn't vandalism if nothing else happens. Nyttend (talk) 22:54, 17 September 2025 (UTC)[reply]
On the basis of the Admin advice above, I have now removed the level 4 warning for vandalism that I posted to the IP's Talk page. It seems that all of the warnings have had completely no effect anyway. Martinevans123 (talk) 16:29, 18 September 2025 (UTC)[reply]
The user shows no interest in heeding the warnings given to them, has not responded in any way, nor even left any comments in their edit summaries, and is therefore clearly being disruptive. Xexerss (talk) 15:51, 18 September 2025 (UTC)[reply]
I have concerns over this recent edit [1]. They added categories for a manga/anime and an internet meme to a short article about food - neither are mentioned at all in the article itself and I fail to see the connection.
Yes, I'm the same person, sometimes I edit from one place and sometimes I edit from another. Shocking, that! Meanwhile, your astonishingly thin skin and complete lack of concern about generating a positive outcome (to the extent of actively driving the thread off-topic when someone else tries to do what you should have done in the first place) are a wonderful source of entertainment, so thanks for that! 173.79.19.248 (talk) 00:35, 30 September 2025 (UTC)[reply]
This venue seems to be a great source of entertainment for you. The contents of your Talk pages are enlightening. I wonder how many other IP addresses you might use, just to keep yourself entertained. Martinevans123 (talk) 08:23, 30 September 2025 (UTC)[reply]
Still 100% committed to undermining any attempt by anyone to actually make anything happen. Plus baseless aspersions, I see! And some incoherent attempts at wit, I guess not surprising. 173.79.19.248 (talk) 11:46, 7 October 2025 (UTC)[reply]
BFDI and Object Shows have been blocked on Wikipedia for a long time, due to BFDI not having "correct sources or something like that." Battle For Dream Island and many other Object Shows have been on news sources, and BFDI is also getting an episode in theaters.
Until those news sources are provided and clearly demonstrate that the subject meets WP:GNG, it is highly unlikely that the article will be unblocked. I recommend you review WP:BFDI, as this has been a surprisingly contentious topic. ThadeusOfNazereth(he/him)Talk to Me!00:19, 26 September 2025 (UTC)[reply]
There is nothing preventing you from writing a draft, if you believe that the sources exist to make the subject now. The fact that people have gone to the troule of writing WP:BFDI should give you pause, and you should read that first, but if you genuinely think that the sourcing has improved then go for it - write a draft and submit it to AfC. Just be aware that you might be wasting your time, and other people's, if you haven't first established that there are sources that are independent, secondary, reliable and which give the subject significant depth of coverage. GirthSummit (blether)04:13, 26 September 2025 (UTC)[reply]
Oh, wow - sorry, I didn't realise that even the draft was salted. I guess that someone could draft something in a sandbox, but absent any new and significantly improved sources they might be better advised to spend their time... well, doing anything at all. GirthSummit (blether)04:30, 26 September 2025 (UTC)[reply]
It really does baffle me. One of those cases of everyone just pressing the bigger, scarier button each time the previous button doesn't work, without any consideration of whether that still accomplishes our goals. What exactly happens if we let people write the draft? A crappy draft exists? Most drafts are crappy. Most have a lower chance of becoming notable than BFDI. Most have a higher risk of hosting spam or BLP violations. Versus the status quo of constantly having to field questions from people who, reasonably, want to know why they can't even do the initial step toward creating an article. We could always have a big notice at the top saying like, "Current consensus is that BFDI is not notable. This draft may not be submitted to AfC without a consensus on the talk page that that has changed" or something like that. -- Tamzin[cetacean needed] (they|xe|🤷) 04:59, 26 September 2025 (UTC)[reply]
That seems an uncharitable assessment of previous admin actions. Looking through the history, what seems to have happened was that the draft existing led to disruption elsewhere, amid wider regular BFDI-related issues which if I recall at one point included admin conduct issues. Looking into it, apparently salting can be temporary like other protections, which could be another tool in the box, but the concept is rare enough that I can't recall ever seeing it discussed anywhere. CMD (talk) 05:30, 26 September 2025 (UTC)[reply]
I don't mean to criticize the admin actions in themselves. The deletion and salting were reasonable implementations of community consensus, and other things stemmed from there. It's the community consensus that I think has been wrong and should change. -- Tamzin[cetacean needed] (they|xe|🤷) 05:34, 26 September 2025 (UTC)[reply]
I'm not actually sure how community consensus works here. There was consensus against whatever draft repeatedly got recreated, but that is not necessarily a consensus against any draft ever, and non-admins can't really assess that at any rate. WP:BFDI says only that "drafts about BFDI have often been deleted", not that there is consensus against any existing. Half a decade on, I suspect an admin could boldly unsalt the draftspace unilaterally, so long as they were prepared to resalt if "drafts being submitted for review and declined over and over again" issues emerge again. CMD (talk) 05:42, 26 September 2025 (UTC)[reply]
Looking at WP:BFDI, it appears to me that what happend was that the community completely ran out of patience after repeated pushing over and over and over completely poisoned the well for anything related to BFDI, largely because of...well, what started this thread was a very mild example, but it's an example nontheless - "the sources totally exist bro for this totally notable thing that Wikipedia must have an article about". Was salting the draft overkill? Maybe, but take a look at Wikipedia:Why is BFDI not on Wikipedia?#List of deletion nominations - especially that footnote "d". If there really is new sourcing that evidences notability, somebody can request the draft page be unsalted - but Wikipedia:Source assessment/Battle for Dream Island, which includes references to this movie-theater screened episode, is telling reading. - The BushrangerOne ping only07:55, 26 September 2025 (UTC)[reply]
Realistically, all the blacklistings do is prevent creation at a page named after "Battle for Dream Island" or "BFDI" or "Object Show" or several other variants. It doesn't do anything about individual users' sandboxes like User:Example/sandbox.Probably the way to go forward, if we want to unsalt something, would be to have an administrator create a single page like Draft:Battle for Dream Island leaving all the other protections and blacklistings in place, and have people work on it there. It's not worth the disruption to try unprotection or unblacklisting in general. I do think it likely that anyone who does take us up on it and work on it there is wasting their time because of the continued sourcing issue, but so long as it's confined to one page, it's wasting only the minimum amount possible of everyone else's. —Cryptic00:55, 28 September 2025 (UTC)[reply]
Vinthepro7v, writing a new Wikipedia article isn't easy. You need solid sources, and the lack of those is the reason why the article (and draft, which is very rare) was salted. What I'm about to say is not going to be nice, but I think it's important that someone says this to you before you waste a whole load of your own time, and a little of other people's. Further up this thread, you suggest that a random Twitter/X account is a possible source for this new article - that suggestion leads me to think that you have not read WP:RS at all. You also mention dubious sources like Business Insider and Bubbleblabber. I don't really know what the other sources you mention are. If you are serious about this undertaking, you are going to have to do some proper research into the sources you hope to cite, and convince people that they comply with the reliable sourcing requirements. That needs to happen before you start work on the draft. So far, you have made only 13 edits, most of which have been deleted for various reasons. I hope you don't think it would be unkind or unfair of me to suggest that you don't really know how to go about writing an article. Starting your editing career with a subject like this is, perhaps, foolhardy. If you want to contribute here, kick it off with a few articles about obviously notable subjects first, learn the lay of the land, and once you're up to speed you will be in a better position to judge whether or not this article is one that you want to try to write. GirthSummit (blether)03:49, 28 September 2025 (UTC)[reply]
I'll also point out that every Wikipedia article is more work for the community to maintain when the creator inevitably stops editing it, and it will most likely start out as an orphan article. Orphan articles attract a lot of attention from people like me, and there are lots of WikiHunters there looking for an easy target. GrinningIodize (talk) 22:30, 29 September 2025 (UTC)[reply]
I strongly oppose loosening any of the protections or blacklistings here; I think the current situation, in which we simply don't allow attempts to create an article or draft on this subject, is correct, and am astonished that this AN thread was given any daylight at all rather than being shot down on sight. * Pppery *it has begun...03:22, 2 October 2025 (UTC)[reply]
I have a few thoughts that are not entirely consistent. First, the community did not salt the title in draft space. The community deleted the draft article, because the proponents of the draft were being disruptive. There were two more disruptive recreations of the draft, at which point an administrator protected the title to prevent continued disruption.
Second, now that ECP-Protect is available, I am not sure that admin-salting a title in draft space is ever a good idea. Admin-salting a title in draft space almost always leads to changing the spelling or capitalization of the title in draft space, requiring more vigilance to prevent versions of the article from being sneaked into article space. Maybe in retrospect the title should have been admin-salted in draft space for two or three months, rather than indefinitely.
Third, I have two conflicting opinions as to whether Battle for Dream Island should have its protection downgraded to ECP at this time. On the one hand, I think it has been fully protected for about five years too long. On the other hand, I am not sure that this is the time for a downgrade. A request to unlock the title, but with the usual unreliable sources, makes me wonder whether there is off-wiki coordination at this time. So maybe it would be in order to fully protect the title in draft space for a month, so that it will expire in a month, and that then it can be allowed to be used as a draft.
Fourth, the state of BFDI in Wikipedia (no article, no draft, multiple rings of protection to prevent article) is very much the fault of its ultras, overly enthusiastic fans. In their eagerness to get an article into the encyclopedia, they tried various sorts of gaming of titles and other sorts of gaming. This has also happened with other subjects, especially actors and entertainers with cult followings.
Robert McClenon (talk) 01:55, 2 October 2025 (UTC)[reply]
A request to unlock the title, but with the usual unreliable sources, makes me wonder whether there is off-wiki coordination at this time.
I don't think so, just some new editors excited about the in-theater screening and relatively unfamiliar with what makes something notable here. While I'm not sure Vin is the best person to start this draft, I think they're here in good faith. StarMississippi03:12, 2 October 2025 (UTC)[reply]
Admin-salting a title in draft space almost always leads to changing the spelling or capitalization of the title in draft space, requiring more vigilance to prevent versions of the article from being sneaked into article This is why we have title blacklisting, which was used here. And as a general principle I find that style of argument ludicrous - for the rare situations where we really want to even prevent a draft from being created, of course we should signal it by admin salting - "but it might not work" isn't an argument at all, just a futile plea. * Pppery *it has begun...03:20, 2 October 2025 (UTC)[reply]
Calling it "admin salting" gives the implication that non-admins can salt pages, and that when a non-admin salts a page, the salting has a different effect because the salter isn't an administrator. – MrPersonHumanGuy (talk) 13:08, 2 October 2025 (UTC)[reply]
I could probably source an article on this topic so as to be facially over WP:GNG, at least in the sense that "there are newspapers that aren't deprecated on WP:RSN that have stories about it". I'm thinking particularly of this.
The issue here is appears that the BFDI fanbase managed to majorly annoy the Wikipedia community such that any new article is going to get the extra-special-audit approach that it may well not pass. Lots of things that we presently have articles on would not withstand this approach, which is of course not a reason not to disagree with that assessment. Had they not done this, we would probably have a stub article about it somewhere and no-one would care.
Writing entire 2.5k-word essays about why we don't have an article on this and tables of how only one source yet reaches GNG (TBH I think another might be argued, but don't care enough to do so) is probably not the best use of editor time.
I favour unsalting this and just letting people get on with it. I think we're eventually going to have an article on it, once another link on the sources table turns green all along. FOARP (talk) 10:07, 6 October 2025 (UTC)[reply]
Sorry. I completely missed that while I was reading the source assessment. It's probably right, but it would be even better if someone could confirm it with a source. Phil Bridger (talk) 12:56, 6 October 2025 (UTC)[reply]
It's independent. It's not great as a source in general as it's for a niche purpose (advising Christian parents on what they should let their kids watch). Like I said, if I cared a stub article could probably be written that would be a WP:GNG pass. The extra-special-audit approach might nuke it, but then it would nuke a lot of things, and why would we apply it?
My view is as documented at WP:DEEPER: " If somebody finds reliable sources that establish notability for this topic, they should request unsalting a draft page at WP:RFPP and provide these sources there." It seems to me from Wikipedia:Source assessment/Battle for Dream Island that the GNG requrements are not yet met, but I have not examined the sources myself. We should not unsalt anything unless there are experienced Wikipedians capable and willing to write a properly sourced stub at least. Sandstein 09:24, 7 October 2025 (UTC)[reply]
Now, again I'm going to emphasise that I'm not invested in the existence of this article and if you want to speedy it right now, you can. I think it's only a squeaking-pass of our guidelines, but I think there is nothing wrong with having an EC-protected draft page that people can work on. FOARP (talk) 10:47, 7 October 2025 (UTC)[reply]
Looking at history, one would never have expected a rail tunnel run by France and England, or a space station run by Russia and America, or a notability proposal endorsed by FOARP and JPxG, and yet here we are.
I'm not greatly opposed to allowing one BFDI draft to exist so interested people can work on it, similar to Draft:CaseOh and for several years Raegan Revord. Drafts beyond that can be nuked IMO. There might be a lot of submissions, but a "no changes since last submission" decline is easy to do, I think. Gråbergs Gråa Sång (talk) 09:31, 7 October 2025 (UTC)[reply]
Thanks for the ping, I somehow turned off notifications on the thread.
@FOARP you're an established editor, I trust your judgement to do whatever feels appropriate especially together with the source conversation from @Gråbergs Gråa Sång. I'm mostly offline these coming weeks, feel free to undo any of my protections related to BFDI. StarMississippi12:13, 7 October 2025 (UTC)[reply]
Bah. I'm not going to try to stand up against the consensus people came to here, but I still think allowing complaints of this sort to succeed even as much as this one has makes a mockery of the way Wikipedia's decision-making processes are supposed to work, and we should have just said "no". * Pppery *it has begun...15:36, 7 October 2025 (UTC)[reply]
I don't see why the draft should continue to be salted. What's the point of draft space if the same standards as main space are used? Of course, if there is disruption, salting can be reinstated. Any draft written would still have to get through AfC before it gets into main space and once there, like any other article, it could be taken to AfD. Phil Bridger (talk) 09:44, 7 October 2025 (UTC)[reply]
"Removing admin-only content as user is not an admin" - updating(sic) user page archives for no-longer admins?
Are these: [2][3][4] good edits? (Happy to accept them as GF)
This is not something I've ever seen done before. Yes, these are users who are no longer admins (for whatever reason), but these are archive pages, of editors who were admins at the time. I find this particularly concerning where it's done to a deceased admin of good reputation. I see no reason to do this, I don't think we should be doing it (and if we did, there'd be a 'bot).
The subcategories of Category:Wikipedia_administrators (and other privileged users) get cleaned up occasionally. People use categories to find admins, so having a non-admin there is not a positive. You might see edits similar to these, periodically, to remove the category. It's not really helpful to do it otherwise. No one checks the backlinks of a topicon to find an admin. -- zzuuzz(talk)20:23, 28 September 2025 (UTC)[reply]
I've just looked at Gonnym's talk page, where he assumes that you are making a fuss about this because you were a friend of RHaworth. I certainly was not a friend of RHaworth, but I still found that edit grossly inappropriate. Phil Bridger (talk) 21:24, 28 September 2025 (UTC)[reply]
A top icon, a userbox or a category that says "This user is an administrator", when a user is not an admin for any reason, is bad for the project. These tools identify users and it can clearly confuse other users, who most don't understand the minute details, why someone isn't an admin but has the tag. These templates or categories didn't only appear on past admins, but also on editors with a few dozens of edits. As I've commented on my talk page, if somekind of tag is still wanted, a new template, that clearly distinguishes the editor as a past admin, can be created. Gonnym (talk) 21:36, 28 September 2025 (UTC)[reply]
Yes, if those subpages ought to be removed (questionable), then surely that's a deletion not blanking? We basically never simply blank stuff, it's just not a useful end result. Andy Dingley (talk) 22:46, 28 September 2025 (UTC)[reply]
Removing the admin topicons is entirely cromulent. Admin topicons should not be on any page that does not belong to an active administrator, full stop. (I'd go so far as to say they shouldn't have been on the archive pages in the first place at all even while the admin was active - for, among other reasons, precisely this one). However the blanking of subpages is arguably against guidelines - In general, one should avoid substantially editing another's user and user talk pages, except when it is likely edits are expected and/or will be helpful. - and is absolutelyan end-around of MfD, which is where these pages should be taken if they should no longer exist. Given that, @Gonnym: needs to receive a {{trout}} and pledge to never make these sort of blanking edits of others' userspace pages again, lest they be sanctioned for it. - The BushrangerOne ping only23:14, 28 September 2025 (UTC)[reply]
"Admin topicons should not be on any page that does not belong to an active administrator, full stop. "
Why? We can control any 'side effects' separately. But labelling archive pages as "This was the page of an admin at the time it was archived." seems an entirely valid (and I'd say the best) interpretation for it. It's also (very clearly) our long-established practice. So you and Gonnym need to go through the regular consensus discussion [sic] processes if you want to change that. Andy Dingley (talk) 10:14, 29 September 2025 (UTC)[reply]
Why? The fact a Wikipedian of your stature and tenure is asking 'why is it a problem to have something that says 'this user is an admin' on the page of a user who is not an admin' is troubling at best. - The BushrangerOne ping only01:42, 30 September 2025 (UTC)[reply]
I don't see the reason to have any topicon on an archive page. As far as I've known, archived talk pages should just consist of archived talk page messages. There is no need for them to contain any "extras" to indicate some sort of status for the editor. I didn't know that there were some exceptions. LizRead!Talk!04:36, 30 September 2025 (UTC)[reply]
Exactly as per voorts. Even if this was an admin defrocked in shame, if it was valid at the time of the archiving, I'm fine with keeping it. Like all the best archives, it's a robust snapshot in time. Andy Dingley (talk) 15:36, 30 September 2025 (UTC)[reply]
There are several separate issues being brought up here, which have different solutions. As I see it:
Categories implying the user is an admin should always be removed from people who aren't admins. I would be shocked if that, by itself, were controversial - this was the underlying task Gonnym was trying to accomplish, and it only blew up because of the things below.
User basepages of non-admins should not normally have userboxes or topicons saying that the user is an admin. I can think of some exceptions to this - for example I use a long series of topicons on my userpage as a history of everything that ever happened to me so would probably want to keep the topicon (without a category) even if I were to resign adminship (although when I did exactly that OwenX removed the topicons and I can't remember whether I didn't notice or noticed and didn't complain). Likewise I wouldn't mind memorializing a deceased admin's user page with {{deceased}} while leaving any userboxes or topicons it has unchanged, as the {{deceased}} template provides needed context.
I'm more in line with Andy Dingley re user archive subpages, though, I see no problem with letting those reflect what was true at the time.
Wikipedia:User pages implies that it is acceptable under certain circumstances to blank user pages of inactive users (WP:STALEDRAFT and WP:User pages#On others' user pages, among other things), so I don't fault Gonnym for extending the principle embedded there to this case, although I'm inclined to agree he extended it slightly too far.
@Pppery agreed re:archive pages, but WP:STALEDRAFT says if the draft is not problematic (e.g. no BLP, reliability, promotional issues) but not ready for mainspace, let it be and WP:User pages#On others' user pages says If the material must be addressed urgently (for example, unambiguous copyright, attack, defamation, or BLP reasons, etc.), the user appears inactive, your edit appears unlikely to cause problems, and you are quite sure the material is inappropriate, then remove or fix the problem material minimally (emphasis added). Based on that the edits that @LakesideMiners pointed out look like pretty cut and dry bad edits that should be reversed. CambrianCrab (talk) please ping me in replies!01:15, 2 October 2025 (UTC)[reply]
Isn't there be some way to add a param to the templates that makes them not emit the category? This should be, like, a thirty-second edit to the topicon template. jp×g🗯️05:20, 5 October 2025 (UTC)[reply]
I have come across a large number of Myanmar village stubs of questionable quality. Two examples I have already sent to PROD before I noticed the scale of the issue are Rgangkum, Chipw and Rgangkum, Hsawlaw. These articles all have almost identical content of the form "X is a village in Y Township, Z State, Burma." and give very imprecise coordinates (degrees and minutes, but no seconds) and they are all citing one or two sources (some combination of GEOnet Names Server or Maplandia and Bing Maps). I spot-checked a few articles and it appears that they have been produced by 3 different editors, though it is hard to tell who produced which article without looking at the history.
Some of the articles are linked in the following Kachin State and Sagaing Region township and district pages Chipwi Township, Hsawlaw Township, Shwegu Township, Bhamo Township, Homalin District, Kale Township, Banmauk Township. By my estimate we are already approaching ~1000 articles here, there are likely more that are not listed in a township page. Hopefully all of them are in the Populated places by region categories under the Populated places in Myanmar category. The problem seems mostly confined to Populated places in Sagaing Region and Populated places in Kachin State categories as the number of pages in the other categories seems more reasonable though probably worth having a look there as well.
The question is what to do about this? For examples of the articles listed in Chipwi Township for somewhere like the aforementioned Rgangkum, Chipw I have not been able to find any mention in any reliable sources, Sawnkyawn could conceivably be an alternate translation of Shankyaw (25.760588682378096, 97.9916423226629) and Chipwi is clearly a notable place.
The volume of the articles seems just too much to do a thorough WP:BEFORE for each one of them to see if they are verifiable and even so I'm not sure if flooding PROD with hundreds of these articles is a productive use of anyone's time. In my view the best option would be to filter out the articles that are clearly pointing to an actual village and WP:TNT rest. Giuliotf (talk) 17:35, 29 September 2025 (UTC)[reply]
Worth pointing out I have tried to have such geo stubs deleted enmasse and it was rejected, given that verifiable populated settlements are considered notable. Agreed with Giuliot but not sure this is the right place to discuss this. The idea of course 16-17 years back was to try to start developing great coverage of Myanmar geographically, but most are small villages with little more than a database ref. Nuking the villages and then developing the district and township articles is the way to go. I sorted out Hkamti Township a while back, redirecting most of the stubs, but it's just too much work needed to go through it. The notable villages will likely be created eventually with proper content and sources.♦ Dr. Blofeld18:58, 29 September 2025 (UTC)[reply]
As chief deleter-of-dubious-settlements, I must second Dr. B's proposal. Right now I am going through Indiana's "unincorporated communities", and I estimate I'm maybe a bit over half done, and the sourcing on these is mostly better and the materials for verification are likely far more available. And that has taken a year and a half, involving 355 deletion nominations; I've probably looked at and not nominated at least twice that many. In terms of the work involved, never minding who ought to be doing it, it makes much more sense to delete the lot and start over, particularly since GEOnet and the like are known to have reliability issues. Mangoe (talk) 19:37, 29 September 2025 (UTC)[reply]
If this is not the right place then I'm happy to move this discussion somewhere else, as for the outcome, I'm fine with whatever the outcome, my preferred choice would be deletion after some basic filtering, but the important thing for me is to have a consensus about how to handle this as its one thing to WP:BEBOLD with one or two articles, but when dealing with this many it would clog up whatever avenue is chosen to deal with them and I think the community would need to be on board for that. Giuliotf (talk) 20:10, 29 September 2025 (UTC)[reply]
Your bit about imprecise coords isn't generally a problem with settlements; as long as the article is correct that it's a village, minutes are routinely sufficient, because seconds are for things more on the scale of individual buildings. Sure, it's good to get seconds when we can — might as well focus on the centre of the community — but it's not something we should expect with typical towns. Nyttend (talk) 21:01, 29 September 2025 (UTC)[reply]
I guess I maybe over-estimated the difference the lack of seconds would make, in which case a lot of the coordinate are either outright wrong or simply aren't populated places, for example Kyokha picked at random doesn't point anywhere near a discernible settlement Giuliotf (talk) 21:28, 29 September 2025 (UTC)[reply]
In densely populated tropical areas it's fairly common for a location specified in minutes to indicate a point in the midst of a cluster of villages, where it can be difficult to work out which one is meant. The other issue, particularly in this region, is that the actual sources are very frequently old maps of quite dubious accuracy. I seem to recall a case of a Myanmar location that was sourced (eventually) to GNS, where their authority was a military map from WW II! As far as Kyokha is concerned, GNS gives 26°40′59″N98°19′53″E / 26.68306°N 98.33139°E / 26.68306; 98.33139 for a location, which turns out to be more or less nowhere and is consistent with what the article says, within a few seconds each way. The listings don't give sources, so no idea where they got that from. But there's plainly not a settlement there. Mangoe (talk) 01:59, 30 September 2025 (UTC)[reply]
A minute of latitude is equivalent to one nautical mile at the poles (by definition of that measurement), or a little more than 1800 metres, and a minute of longitude is roughly the same at the equator; it decreases to nothingness at the poles, but since Myanmar is mostly tropical (its extreme points are roughly 9°54'N and 28°30'N), that shouldn't matter hugely; you can assume that a minute in either direction is just a little less than a nautical mile, which is generally sufficient for a town. I agree with Mangoe's comments, but if the village really does exist nearby and is surrounded by countryside, at worst it shouldn't be hard to find on Google Maps. If you can't find anything at provided coords, and if it's not in the midst of a cluster of villages, I think you can conclude that it's disappeared, that it never existed, or that the coords are completely wrong. Ghost towns are notable — an existing town is, and a notable entity doesn't lose notability merely because it ceases to exist — but I'd be more open to deleting articles about ghost towns in this situation, simply because we can't tell whether such an article (1) covers a ghost town or (2) is outright wrong somehow. Nyttend (talk) 02:30, 30 September 2025 (UTC)[reply]
Nowhere and nothing is automatically notable, and this goes for towns/villages also. Even the GEOLAND standard requires legal regonition, not simply that the location has a name and that people lived there at some point. FOARP (talk) 17:26, 3 October 2025 (UTC)[reply]
Yes, in most cases you can find the precise coordinates and see they are legit villages on Google Earth. You can add some location observation data to flesh it out but if there is no info online about it aside from a database listing I think we should redirect to a list of settlements with coordinates and tabled summary if verifiable. When the sources are available then separate articles can be created. Would advise the same on Carlos's Iran short stubs. Hkamti is an example of what we should do. We shouldn't be deleting or redirecting village tracts IMO, Mepok was once an xx is a village database stub. Some can be expanded, but just too time consuming going through all the townships.♦ Dr. Blofeld08:14, 30 September 2025 (UTC)[reply]
@Giuliotf: The xxx is a village and database ref isn't just contained to Myanmar either, it's long been a problem on here for many developing world countries. What I would suggest is starting a new proposal to change the WP:GEOLAND guidelines on settlements. Suggested revision. "While most human populated settlements are generally considered notable, in many cases where there is no information about the place online aside from a GEONames or database ref and the article has been a short stub for many years, consider redirecting the article into a higher administrative division article or a tabled list of settlements by division with coordinates until more information about the place can be found online to justify a full article". No admin is going to blanket delete a thousand Burmese villages. Gain a new consensus in adjusting the notability guidelines would be my suggestion. If that fails, you are still free to take whatever neccessary action you want in sorting out Myanmar. ♦ Dr. Blofeld09:22, 30 September 2025 (UTC)[reply]
Actually in looking it already says " If a Wikipedia article cannot be developed using known sources, information on the informal place should be included in the more general article on the legally recognized populated place or administrative subdivision that contains it." But I think this should be elaborated on with my suggestion to give editors a better idea.♦ Dr. Blofeld09:29, 30 September 2025 (UTC)[reply]
These articles are really very problematic. I too have spent quite some time trying to clean up these mass-created articles. They are based mostly on GNS, which is an unreliable source (see the RSN consensus on this). Many of them patently do not exist, or at least not as villages. FOARP (talk) 17:24, 3 October 2025 (UTC)[reply]
Revoking autopatrolled right from Sarefo for undisclosed LLM-generated articles
Apologies if I am in the wrong place - this is my first (hopefully last) time coming to AN. I am writing here to request that @Sarefo's autopatrolled right (granted in 2011 by @Acalamari) in light of discussion at their talk page and the NSPECIES guideline talk page (the latter is a long discussion that only partially pertains to this user). Sarefo has been creating articles on South African spiders (see xtools) using Claude Sonnet 4.1 without disclosing that the content was generated by a LLM and, being autopatrolled, these articles bypassed new page patrol. This was seemingly only discovered because obvious LLM tells and hallucinated statements were noticed in the articles. I don't believe that any admin would grant autopatrolled to a user knowing that they intended to use an LLM to automate the mass creation of articles, hence my request that the right be revoked. AI generated text requires extra scrutiny from editors, not less, and these articles should not be automatically reviewed, and significant use of LLMs should be disclosed.
If you read the linked talk page discussion, Sarefo has been asking Claude to write articles, based on specific sources; he saves the AI-written content, reviews it to ensure that it's accurate, and then mass-creates articles when convenient. The basic workflow is not problematic. The rate of upload does not necessarily represent the rate of creating. I'm sometimes batching uploads that were prepared earlier, so doing them in batches is fine. Also note a comment at the bottom of the guideline talk page: I wouldn't be using claude 4 as part of my toolkit for this. Perhaps the quality of reviewing Claude's content is sufficient to revoke autopatrolled, but since Sarefo won't continue to use Claude here, there's no reason to revoke autopatrolled. Nyttend (talk) 02:43, 30 September 2025 (UTC)[reply]
I think you have misunderstood the linked discussions, I would suggest reading both threads in full. When Sarefo says I wouldn't be using claude 4 as part of my toolkit for this, they are referring to not using Claude 4, which is not the same as Claude Sonnet 4.1 (though I must note I think they made a mistake referring to the model they're using as Sonnet - I think they meant to say Claude Opus 4.1, but I am using Sonnet here as that's what they've said on their talk), because they feel Claude 4 is worse at this task than Claude Sonnet 4.1. They are giving an example of a model they think would be particularly unsuitable for the task, not saying that they will not continue use LLMs to create articles. Unless Sarefo wants to correct me now (and I am very happy to be corrected!), they have given no indication that they plan to stop generating articles with Claude. On the NSPECIES talk they say that they have around 500 more articles to complete the species in the existing SANSA guide and I'm happy to slow down to around 50 per day. Again, I am happy to be corrected if Sarefo intends to create these pages manually and/or seek WP:MASSCREATE permission, but I do not see any evidence that that is the case. Ethmostigmus 🌿 (talk | contribs) 03:30, 30 September 2025 (UTC)[reply]
hi! trying to clarify as much as I can, if anything unclear let me know.
some context. I'm a biologist (Masters' in genetics/bioinformatics, but big fan of biodiversity, especially spiders). I built the framework for wikipedia's spider section (WikiProject Spiders, doesn't seem very active nowadays) around 2006, using mostly scripting. I've been very active on iNaturalist the past decade or so (my user name is portioid there, if you're interested). I recently realized that Ansie Dippenaar-Schoeman (ADS), SA's most prominent arachnologist, uploaded thousands of the late Peter Webb's high-quality spider photos to iNaturalist under an open license, and also creates quite good guides for all the spider families in the area. So I'm bringing these together to build the SA spider section on wikipedia. these are photos of wild specimens that were then caught and dissected, meaning that we'll have scientifically determined high-quality photos of about 500 species, which is awesome. the spider guides are of rather good quality IMO, and my personal assessment was that these are essentially a collection of pre-written short wikipedia starter pages already that just need to be transferred. I call it Project Spider Webb. if you think it would make sense to create a project page for it, let me know.
to achieve this, I'm using a combination of manual work, python scripting and claude sonnet 4 (now 4.5). at the moment, my sources are WSC (world spider catalog, world authority on spider systematics) and ADS' guides. I have been using claude extensively for coding for the last year or so, so am generally aware of its strengths and weaknesses. the use of claude in this project is mostly to make existing passages flow better, as the original guides are sometimes written in shorthand. I proofread everything it outputs, and tweak it to a varying extent.
I don't fully understand in what way you're both misreading my earlier posts ;), but I'm referring to claude sonnet 4 (now 4.5). I mistyped when I mentioned sonnet 4.1, there's no such thing (I was working with claude opus 4.1 for programming earlier, hence the mixup). 4.5 is very new (it came out today) and allegedly markedly better, but I will have to test this more to have details, especially regarding fabulations. as I mentioned elsewhere, I'm using claude essentially as a glorified text processor as part of my pipeline. it is my impression that this is legitimate under wikipedia's current guidelines.
I am currently in the process of mass uploading the photos to Commons, and creating categories etc. Once that is done, I intend to complete adding SA spider species pages. as mentioned, there are a few hundred left I think.
I was made aware of some mistakes that slipped in, for which I am very thankful. this showed that I did not check the output as closely as I thought I had. at some point, I had gotten a bit careless in the rush of things. I improved parts of my scripts (including the part where claude is involved), and how I proofread the pages. I have been closely reading through hundreds of these since, and found very few mistakes. these mostly happen with claude 4 when it is expecting some input, but I did not give that input.
some people have worried about the short-term frequency of article uploading ("every two minutes", etc.). this may be an artifact of batch uploading previously generated works. I sometimes generate and proofread a batch of articles, then upload them in one go. so the upload rate is not necessarily an indication of creation speed. also, some of the shorter pages are much easier and hence faster to check than more complicated ones. this eg. depends on which spider family I'm currently working on.
my personal suggestion would be to let me finish this project using my current workflow, adding the remaining pages. then I will revisit the earlier additions to make sure they are up to the standard of the later ones. SA spiders are not a highly visited topic (yet!) I assume, so there should be no immediate rush I guess ;) any help is encouraged and appreciated, but not required. I'm happy for anyone joining in with the effort, but am equally happy to bring this to a level everyone can live with by myself.
I am comparing my output with average content on wikipedia, which as you all know is a mixed bag. I think it's better than much content that nobody has contested for years, but not as great as that of some of the great contributors here. I can live with that. it's a first version, and wikipedia is meant to be iterated on. I've experienced in the last week(s) that my content is held to a much higher standard of perfection, which I find only partly justified (because of the amount).
most of the feedback I was given did not actually involve problems resulting from LLM use. again, my suggestion would be to extend to me a certain level of trust that I'm trying to build something of value here. I'm not some random jerk poisoning the waters with hallucinated AI slop.
pain points I'm personally aware of with my past output:
half-sentence additions here and there that while not factually wrong, are not actually explicitly mentioned in the source. this happens much less in newer content, but I don't think it's a drastic problem in earlier content either. after being made aware of two of these, and very thoroughly checking around a hundred articles, I spotted around two of these in my output at the time.
due to the way claude internally processes the content I feed it (WSC + respective ADS guide), it sometimes will give the wrong reference (WSC for ADS content or vice versa). when proofreading, I did not thoroughly check for this. that's not great, but hopefully tolerable. now that I'm aware of it, I'm keeping an eye on it.
regarding fabulations, I can't fully vouch for the earliest articles I created, and hence will revisit them in the coming weeks. there might be a phase where my diligence waned for a bit after initial scrutiny, until I was made aware of some flaws I had overlooked. I don't fully remember when I started using claude. I started doing this manually, and at some point tried claude sonnet 4, and found it to be a valuable tool for part of the creation process.
to summarize. my plan:
finish adding pages using ADS' spider guides content (cc-by)
start adding Peter Webb's photos
revisit earlier pages in this project, to weed out beginner's problems that probably exist in these
take a good rest, because this project is amazing but also exhausting. I have no plans on adding any other mass content to wikipedia for the foreseeable future. I'll be updating WSC data from time to time (using python scripting).
thanks for taking the time, and for keeping this friendly and factual! as I mentioned in the other thread, we all have essentially the same goal, and I'm confident we can find a path to there that brings a smile on all of our faces. Sarefo (talk) 04:36, 30 September 2025 (UTC)[reply]
Expectations around mass creation were not created due to the need to respond to "random jerk"s, they were created due to good faith editors moving too fast and too far and creating cleanup work for others. For example, you are mass-uploading images on Commons calling the author Peter Webb and put in the category "Photos by Peter Webb". However the author of every one of a random sample of your 50 most recent uploads appears based on the source website to be someone called Ansie Dippenaar-Schoeman (eg). This is technically a violation of the licence, and while it can be fixed it's a lot of cleanup. CMD (talk) 05:23, 30 September 2025 (UTC)[reply]
On the subject of accusing people of typing without reading, here is what they said in the second paragraph of their comment:
I recently realized that Ansie Dippenaar-Schoeman (ADS), SA's most prominent arachnologist, uploaded thousands of the late Peter Webb's high-quality spider photos to iNaturalist under an open license, and also creates quite good guides for all the spider families in the area.
Thank you for the clarifications (you has me more than a bit confused with Sonnet 4.1, but I figured it was a safe bet you meant either Opus 4.1 or Sonnet 4.5, and not some secret Claude version I was unaware of), though I must ask again that you clarify the scope of your use of automation in regards to which/how many articles are affected, as these will need substantial clean-up if they resemble your recently created articles.
I don't necessarily have a problem with you using automation in this process but, while I can tell you are acting in good faith, creating such a large volume of articles with a LLM without disclosing this and without going through the normal page review process seems like a serious lapse of judgement. Mass-created articles and articles generated by LLMs are two categories of article that require extra scrutiny, not less than is expected for the average article by an extended confirmed user, and you've combined these two methods at higher risk of error into a single project while entirely bypassing the safety net of new page patrol. You've started a large-scale automated article creation process without disclosing the level of automation or putting any checks and balances in place besides your own personal review, which is not enough when producing such large amounts of content. It seems like the only reason the errors in these articles have come to light is because @Novedevo noticed Claude saying "here's the article you requested" and flagged it on your talk page, which led other users to investigate the rest of your articles for further errors. I am certainly seeing issues that others have not flagged that need addressing and seem to be common across the board with your recent articles.
I appreciate that you've paused creating new articles since becoming involved in this discussion and would kindly request you continuing to keep this on hold for the moment until this is resolved. This is a project with a lofty goal but I am really deeply disappointed by the failure to disclose the use of automation so other editors can ensure the quality of these articles, and I maintain that you have not held to the high standards expected from an experienced autopatrolled user. Ethmostigmus 🌿 (talk | contribs) 06:05, 30 September 2025 (UTC)[reply]
this wasn't really a problem when I auto-created lots of articles in 2006 IIRC. I've been working on iNaturalist for the last years, and I was generally met with a wall of silence on its forum when trying to talk about projects, so that was the mindset I was in. build this, make sure it's of good quality. as I mentioned earlier, should I create a project page for this, so we can discuss improvements in a central place? or any other suggestions to centralize this? Sarefo (talk) 21:15, 30 September 2025 (UTC)[reply]
A lot of things have changed in Wikipedia since 2006. I used a lot of really poor sources then, and created some very iffy articles. We have tightened up our standards for sources, articles, and the pace of creating articles since then. My personal advice is to slow down the rate of creating new articles and spend more time expanding each article before taking it to main space. Even better, spend part of your time expanding existing stubs. I spend as much time on major expansions of existing articles as I do on creating articles. Donald Albury22:09, 30 September 2025 (UTC)[reply]
We can assume good faith on the part of the relevant editor here. I don't think the details of how LLMs were used in this case are particularly relevant. The crux of the issue is that the autopatrolled right is completely unnecessary for article creation. It has no impact whatsoever on this editor's capacity to create more articles, and one hopes he will continue to do so. The purpose of autopatrolled is to remove works created by trusted editors from the NPP log, so as to reduce the overall burden on new page patrollers. In a case where an editor has used, and plans to continue to use LLMs for article creation, and where numerous errors have been identified in his past creations, it does not make any sense to treat that editor as trusted, and allow him to continue to retain the autopatrolled right. For the encyclopaedia, it is imperative that a second pair of eyes be drawn to any content resulting from LLM-based editing before it is indexed. In fact, this benefits the editor, because the patrollers may pick up on phenomena he did not detect. Given this, I must support the stripping of this editor's autopatrolled right, with no prejudice against his past and continued contributions to the encyclopaedia. Yours, &c.RGloucester — ☎06:06, 30 September 2025 (UTC)[reply]
[triple edit conflict] Sarefo, as far as mass addition — just slow down and add some of these articles more gradually, like one or two per day, and people won't object to the quantity. (Imagine that you were doing all the work manually, and upload them at this rate.) Obviously once they're written, it won't take much work at all; just log in for a couple minutes and copy/paste from your offline saved information. Or better yet, mass-create them in userspace (User:Sarefo/Spider species 1, User:Sarefo/Spider species 2, etc.), wait a while, and move them into mainspace at your convenience. As far as files — as Chipmunkdavis notes, the linked image does need to be attributed to "Ansie Dippenaar-Schoeman", even if it were a Webb creation, because the source specifies Dippenaar-Schoeman as the required attribution. Be sure that you're complying with the source when attributing copyrighted-but-freely-licenced images. If you have good reason to say that they're Webb creations, you can note this in the file description page, and you can still use the Webb category.
As far as the bulk of the content — since you're reviewing and modifying the AI-generated text, it's not a fundamental problem, but if your output contains errors, yes this will be a problem. I don't want to deter you from this work, but supplying erroneous information is disruptive, so you need to be careful not to disrupt things; you've been around here long enough that you know that people get sanctions of various sorts if they keep disrupting things. And this applies both to the text of the article ("Spider A lives in America" when it really lives in Europe) and to the citations (citing a statement to a source that doesn't say so), so yes be careful to keep a very close eye on it.
South Africa, I believe Sarefo is in contact with some of the researchers and managed to secure image release in a compatible licence. CMD (talk) 06:36, 30 September 2025 (UTC)[reply]
the observations on iNaturalist also say these are observations and photos by Peter Webb, so I don't think there will a problem here?
any work that is actually done will contain errors. the point is to minimize errors as much as possible. I'm doing this. I think the output is of pretty good quality. I don't think holding me to a 100% perfection standard is fair. I'm also happy to run a bot over this set of articles, eg. removing the status information (IUCN etc.) which I scripted in, but ethmo told me recently under wrong assumptions. I really don't think I'll be putting any unnecessary burden on others.
South African spiders :) Peter Webb was a colleague of Ansie Dippenaar-Schoeman, SA's most prominent spider scientist. he died recently, so she spoke with the family, and they granted open rights to make these fantastic photos accessible. she uploaded most of them to iNaturalist, that's how I noticed it. Sarefo (talk) 21:25, 30 September 2025 (UTC)[reply]
This is just about my exact sentiment, but you put it into words much better than I did! I don't want Sarefo to stop creating articles, but this is not an appropriate use of the autopatrolled right. Ethmostigmus 🌿 (talk | contribs) 06:15, 30 September 2025 (UTC)[reply]
It is worth noting that, since last year, autopatrollers can unpatrol their own creations. I wrote the current rough guidance on this—Use cases for this include where most of the page's content was written by someone else who is not autopatrolled, where the user has a conflict of interest, or where the user is unfamiliar with the topic area or otherwise thinks the page would benefit from an outside review—and I think it would make sense to add "where the page's creation involved generative AI" to that list. @Sarefo: Would you be willing to self-unpatrol any AI-created articles you publish in the future? There should be a button in the tools menu saying "Add to the New Pages Feed". -- Tamzin[cetacean needed] (they|xe|🤷) 09:18, 30 September 2025 (UTC), ed. 10:14, 30 September 2025 (UTC)[reply]
The link is called "Mark this page as unreviewed" but as Tamzin says, this does exist as a feature since last year. I've attached screenshots of how it should look for folks with autopatrolled. This is how the link looks like in Vector22which then opens up this dialogSohom (talk) 12:18, 30 September 2025 (UTC)[reply]
With all respect, offering that option does not seem appropriate to me. I want these articles to be generated, and I'm sure Sarefo has the best intentions (and good material). However, what the autopatrolled right comes down to in the final analysis is "does the community think this user has sufficient judgement to not require oversight". I'd suggest that the flaws discussed so far, and the discussion had here, show that this judgement isn't quite there. Jank autopatrolled. It's not a collectible hat, and this only places them back on the same level as the vast majority of valued and productive users. --Elmidae (talk · contribs) 13:31, 30 September 2025 (UTC)[reply]
Speaking as a patroller, I'm not entirely sure whether it would do anyone any good for NPP to be flooded with articles about spiders: NPP is backlogged enough as it is. It therefore doesn't strike me as a very good idea either to revoke autopatrolled or to ask Sarefo to tag the articles as unpatrolled. In my view, if editors are going to create huge numbers of semi-automated articles, this should be on the understanding that they will check them rigorously for themselves and not publish them in mainspace until after they are properly checked: I mean, I'm not sure if any editor should be mass-creating articles at all unless they are autopatrolled and check their articles in the thorough manner that that user right presupposes.
So I wonder whether it might be better to ask Sarefo to create these articles in draftspace and move them to mainspace only after the whole process of creating and checking has been completed. Dionysodorus (talk) 16:09, 30 September 2025 (UTC)[reply]
I'm also putting a lot of personal effort into it. My fingers hurt. I'm seeing spiders the moment I close my eyes. I'm essentially reduced to working on this and sleeping for the last weeks.[5]
As there is no time limit on getting every spider in South Africa added to the encyclopedia, I think it would be better all around for them to slow down. A problem with these kinds of mass creations (LLM, scripted, or human written) is the break in feedback. One missed error becomes 50 missed errors when you are publishing 50 articles at once, and one is much more likely to miss errors when working flat-out every day. REAL_MOUSE_IRLtalk18:14, 30 September 2025 (UTC)[reply]
I'm also fine with this. As I keep mentioning, while I highly appreciate any input on how to improve this dataset, I'm also very happy to work on it by myself to bring it up to a level everybody is happy with. I don't want to put a burden on anyone.
I also think that the recent pages are totally good enough to stay in mainspace while I improve on them.
You'll also have to consider that I have a life outside of this project (soon, hopefully). I still think if you let me do this my way, while I incorporate input I get to improve, is the best way to go for everyone. Everyone does things differently. In the end, it's the result that counts. Sarefo (talk) 21:32, 30 September 2025 (UTC)[reply]
If I am reading correctly, they aren't hooking the model directly up to the editing box, they are using it as part of a workflow which they manually review prior to posting any articles. jp×g🗯️01:35, 1 October 2025 (UTC)[reply]
that is correct. it is also correct that I improved my workflow in the process of doing this, also due to feedback from here. that's why I want to revisit older articles once I'm finished to clean up what I might have missed. again, as it's not a highly frequented area anyway, I think the easiest thing would be to just let me wrap this up in mainspace and clean up whatever might be necessary as the next step. I don't think there is anything dramatic there. here and there a sentence that while not wrong is not written as such in the source. very rarely some fabrications I might have missed. it took me a few days to adapt from reviewing the model's coding output to using it for text processing. Sarefo (talk) 04:55, 1 October 2025 (UTC)[reply]
Sarefo came to my attention while following up on reports at CopyPatrol. Substantial portions of the text of articles created matched texts at the websites discussed above, which triggered a red flag. The text was cited, and IIRC, Sarefo felt that was sufficient, not realizing that an attribution template was required. Sarefo accepted this observation, immediately modified their workflow, and subsequent articles created included the proper attribution. I backfilled the attribution on approximately 250 articles. While understanding that my focus was on the copyright issues, I did peruse some of the content, and what I looked at seemed fine. It's my understanding that Sarefo is not using an LLM create content ab initio, but mainly for organization of material already created. I'm aware that the community has serious concerns about the use of LLM's, but this strikes me as a perfectly acceptable use case.S Philbrick(Talk)13:18, 3 October 2025 (UTC)[reply]
There is no mechanical distinction between organizing content and creating content ab initio, all an llm does is absorb the prompts it is given into its mathematical formulas and produces an output that it converts into words. You get a sliding scale of 'ab initio' content depending on how controlled the prompts are, but the underlying mechanism is the same and reorganization can (and in this case has) produce 'hallucinations'. CMD (talk) 04:26, 4 October 2025 (UTC)[reply]
With all due respect, I understand how Sarefo is creating their articles. This is a discussion about whether or not this is an appropriate use of the autopatrolled user right, not about what is and is not acceptable LLM use. Sarefo is not violating any policies by using Claude to generate articles and I am not trying to stop them from doing so, only raising my concerns about oversight because I believe it will improve the quality of their articles. Not to single you out in replying to your comment, Sphilbrick, (yours is just the latest and makes the most sense to reply to) but it feels like about half of the people who've commented on this discussion so far have done so not to address my actual concern but to instead have a completely separate discussion about the ethics of LLM use, as if talking past me. Absolutely no offence intended, but I would desperately like to steer this discussion back to the specific topic of whether or not large numbers of articles mass-created by LLMs should be autopatrolled. Ethmostigmus 🌿 (talk | contribs) 10:31, 5 October 2025 (UTC)[reply]
I haven't historically been active in either granting or removing autopatrolled rights, so I decided to re-review Wikipedia:Autopatrolled. Much of the content was straightforward but I was somewhat surprised to see so little content in guidelines for revocation. It's short so I will repeat the content in full:
If an autopatrolled user hasn't edited for three or more years, their right may be revoked.
I don't view this as exhaustive — if an editor is granted the right and generates some new articles that, in retrospect, ought to have been subject to review, it may be appropriate to remove the right.
If the discussion is not about the use of LLM, then I would see the decision turning on whether this editor has recently created multiple examples of articles that would've been flagged by new page patrollers as problematic. I am aware that there may have been some examples early in the process but how many examples can you provide from, say, the last month? S Philbrick(Talk)14:58, 5 October 2025 (UTC)[reply]
Apologies for the late reply, real life has been very busy lately. Yes, I also noticed that the guidelines for revocation were severely lacking... This is something I would really like to see expanded. I am autopatrolled, and it's a privilege that I work hard to show that I deserve and will use with responsibility - if I were to stop being so diligent, it's only right that I should not be afforded such a privilege. We need to develop a guideline that explains what behaviour is and is not acceptable for autopatrolled users. I note that the admins responding to requests at Wikipedia:Requests for permissions/Autopatrolled are quite thorough in vetting users requesting the right, and I think that kind of thorough scrutiny should continue throughout a user's editing career.
RE: examples of issues - I must quickly point out that when you say there may have been some examples early in the process but how many examples can you provide from, say, the last month?, the earliest issue I'm aware of being flagged was by @Novodevo on Sarefo's talk page on 21 September, so well within the last month (not sure when Sarefo began creating these articles). A non-exhaustive list of issues I and others have noticed:
Including Claude's response to the prompt ("Here's a short distribution section for the genus *Amaurobioides* article:") at Amaurobioides (flagged in user talk discussion by Novodevo)
The article for Apochinomma deceptum (the valid name of the species) was referred to by the incorrectly formed basionym Apochinomma decepta within the article body despite being located at the correct article name - this is an error stemming from prompting Claude to draw primarily from a source which used the incorrect name (flagged in user talk discussion by @Jlwoodwa)
Also on the article for Apochinomma deceptum, Claude referred to two ant genera as species, and refers to a genus "Streblognatha", which does not exist (I believe it to be a misspelling of Streblognathus) - these may be a result of Claude closely paraphrasing the source text, however, I cannot check this as it absolutely refuses to load in my browser and consistently fails to download - if someone can confirm/disprove this I would appreciate it. The article also seemed to claim that the species was present in three particular protected areas (Ndumo Game Reserve, Tembe Elephant Park, and Kruger National Park) which are not actually stated in the source text and appear to have been mere guesses by Claude (incorrect use of the word species and hallucinated park occurrences flagged in user talk discussion by @Chipmunkdavis and Sarefo respectively, possible misspelling noticed by myself just now)
Almost all articles incorrectly use the |status= parameter in the Speciesbox - this is something I have manually fixed on around 40 pages and notified Sarefo about. This comes in two forms or error: articles that incorrectly assign a IUCN conservation status in the Speciesbox when the species in question has not actually been assessed for the Red List, and articles that incorrectly assign a Red List of South African Plants status to spider species. This is an issue arising from the authors of the text Claude is primarily drawing from assigning statuses based on similar criteria to the Red List, but which are not the same as an actual assessment - South Africa has its own special system for this which is similar to but distinct from the IUCN Red List (flagged in user talk discussion by Chipmunkdavis - this was something that immediately stood out to me as a frequent user of the |status= param who knows about SANBI's weird independent system)
Some articles make reasonable but unsourced claims about the etymology of specific epithets, eg. Olios lacticolor claiming that The species name lacticolor derives from Latin, meaning "milk-colored" but providing no source - this seems to just be the LLM just making an assumption rather than drawing from a source. In the example of O. lacticolor, I checked all three sources in the article to make sure it wasn't just a failure to correctly attribute the information, but no, none of them give an etymology (noticed by myself while cleaning up some of these articles, I have flagged the ones I saw as needing citation)
This is relatively minor compared to factual inaccuracies, but many articles have general issues with formatting, prose quality, and not following WP:MOS. I think many articles would benefit from a cleanup template like Template:Tone or Template:Copyedit. A near universal example of this is that common nouns are capitalised - this seems to stem directly from the primary source the LLM is drawing from also using this sort of weird capitalisation (eg. Savannah and Thicket biomes in Olios lacticolor). Another very common example is that sentences discussing taxonomy attempt to attribute the description of a species to an author, but only provide a last name (eg. Olios zulu was described by Simon in 1880 in Olios zulu - I know which Simon this is referring to because I'm a nerd - it's the French scientist Eugène Simon, who had a huge impact on arachnology - but to an average reader this is obviously not helpful). The author is usually linked in the |authority= param in the Speciesbox, but how many readers will know to look for that?
These are just what I'm aware of thanks to the work of others and my own viewing of around 50 of the LLM-generated articles, but I think they are enough to illustrate that these articles suffer from a pattern of poor quality control owing to the inherent flaws of LLMs, Sarefo only being a single person not able to catch every possible issue, and the lack of any further oversight from the review process to catch the issues slipping through. FOARP has also raised the issue of WP:MASSCREATE below which I think is highly relevant here. Ethmostigmus 🌿 (talk | contribs) 02:50, 9 October 2025 (UTC)[reply]
If we are discussing Apochinomma deceptum, a significant example was the sentence "Adult females have been collected during February, May, and November, while males were collected in April, July, and October" was apparently completely made up from nothing (ie. not even a mangling of the source). This was only found after I spent substantial time checking through the article. CMD (talk) 03:22, 9 October 2025 (UTC)[reply]
Revoke autopatrolled until the Sarefo can show again that they should be entitled to have it - Using LLM-generated text without sufficient review is a perfectly good reason to do this. Even if done in good faith, this is a WP:CIR issue. I take the point about NPP being flooded, but in that case the author needs to slow down and do more checking, and NPP reviewers can flag that up to them if they are finding a lot of spider articles jamming the feed. Sarefo also needs to go through WP:MASSCREATE for mass-creation of articles using automated or semi-automated tools (which is what using an LLM is). FOARP (talk) 09:34, 7 October 2025 (UTC)[reply]
Hello. I respectfully ask the community to reconsider my ban.
I understand now why I was banned: I created edit wars because I did not follow the rules to start discussions on the talk pages and wait for other opinions. Some of my words also sounded uncivil in English. I am sorry for this, and I have learned from it. I will not repeat that behavior.
For the future, I will not edit existing articles about Freemasonry anymore. I know that this was the wrong way and it only created conflicts.
But I would like to be allowed to finish my own article draft about the contradictions of Freemasonry (I will rename it to that – because I also renamed it in my German article). This was the reason why I came to en-WP in the first place. I started that draft already before my ban (see: User:Wikiprediger/Freemasonry criticism). It is based on facts and reliable sources, and I want to translate it step by step.
See also the actual German version of this article: https://de.wikipedia.org/wiki/Benutzer:Wikiprediger/Widerspr%C3%BCche_in_der_Freimaurerei.
This is the Google-Translator Version (english): https://de-m-wikipedia-org.translate.goog/wiki/Benutzer:Wikiprediger/Widerspr%C3%BCche_in_der_Freimaurerei?_x_tr_sl=de&_x_tr_tl=en&_x_tr_hl=de&_x_tr_pto=wapp
I want to remark that I am on de-WP since January 2014 and had never any conflicts until my edits on Freemasonry in 2024.
I understand that topics like Freemasonry are sensitive, and that is why I started a separate draft long ago – after I recognized the intensity of the conflicts – but the idea unfortunately came too late. My aim is not to attack, but to write a fact-based article that highlights contradictions and critical perspectives, supported by reliable sources.
I promise to be very patient: first completing the German version, then waiting for review and discussion there. Only after that I will create an English version. I will follow all rules, respond to discussions, and start a review before publishing. My only goal is to contribute to Wikipedia and i think such an article is missing here.
Please unblock me with the condition that I stay away from existing Freemasonry articles, but that I may continue to work on my own draft. I believe this would be a fair compromise. If I make mistakes, they can always be corrected or discussed.
Thank you very much. Wikiprediger (talk) 14:16, 30 September 2025 (UTC)[reply]
I'll note that while I'm glad Wikiprediger is willing to take a TBAN, I did suggest that they not ask for the exemption for their draft. I would also support a TBAN from freemasonry, broadly construed, without the exception for the draft. voorts (talk/contributions) 22:55, 30 September 2025 (UTC)[reply]
I still don't think this should have been a CBAN, but now I don't think Wikiprediger should be unblocked because it's clear that they only want to use Wikipedia writ large to POV-push their views on freemasonry. voorts (talk/contributions) 23:28, 1 October 2025 (UTC)[reply]
Looking at Wikiprediger’s edits at de.WP, I am concerned that he is a bit of an SPA, focused on Freemasonry (from a negative POV). This was the topic area that resulted in his being blocked in the first place. Suggest a topic ban (broadly construed) if he is to return. This would include not working on his desired draft. Blueboar (talk) 21:55, 30 September 2025 (UTC)[reply]
@Blueboar: I understand the concern, but in Wikipedia we should not be judged by personal opinion, only by how we use good sources.
There should not be a problem with my perspective on Freemasonry, as long as I use reliable sources and remain neutral.
There will not be an Edit-War again. We can discuss it on the talk page and ask for third opinions or reviews.
@Wikiprediger: it's not censorship. A TBAN (which I thought you agreed with based on your unblock request) is used to avoid disruptive editing. You were banned by the community for being disruptive in that topic area, and you have indicated that you would like to continue editing in it. If you can edit productively for a significant amount of time (at least 6 months, preferably longer), you can come to the community and ask to have the restriction lifted. voorts (talk/contributions) 01:58, 1 October 2025 (UTC)[reply]
I think I would support this unless other issues that have not been mentioned come up but I don't see any way to enforce a topic ban EXCEPT for in the editor's sandbox. I think if this is the plan, it would be better for the editor to work on the English language verion of this article on the German Wikipedia and after appealing the topic ban here after a few months, then moving that article to this Wikipedia if that appeal is granted and the topic ban is lifted. However, this proposal might not get wide approval among other editors since the focus of this editor's work is still the subject that originally got them into this mess in the first place. LizRead!Talk!21:45, 30 September 2025 (UTC)[reply]
@Liz: Thank you for your comment. I can accept working in my sandbox if that is the best way.
It will take some time anyway. First the German article surely will be discussed and maybe improved, then I can translate it step by step in the sandbox.
It would be best to do this in en:WP, so I can ask for review here. Also we can use the talk-page for discussions. I am used to finish articles as draft. When we have reached a consensus, the draft can then be moved into the article namespace. Wikiprediger (talk) 01:46, 1 October 2025 (UTC)[reply]
Comment from Blocking Admin I implemented the consensus of the discussion and have not been able to follow the recent discussions on @Wikiprediger's Talk and will not have the on wiki time to do so. I support whatever consensus the community comes to on the unblock and potential topic bans. StarMississippi02:48, 1 October 2025 (UTC)[reply]
Concerned. This doesn't feel right. Typically, editors who have experienced problems in a specific topic area but who want to contribute to the overall project will be happy to accept a topic ban from that specific area in order to be unblocked. Here we have someone who has never shown any interest in editing other subject areas, but who promises not to edit existing articles in their favoured topic area, and only wants to write their own article about the subject. I am concerned that this is, in a sense, setting them up for failure - this is a collaborative project, nobody gets to write their own article about any subject, and when they discover that the 'I won't edit existing articles on the subject' firewall is not a two-way street I fear that we will be in for a lot more drama down the line. I would be happy to support an unblock with a topic ban on the subject of freemasonry, broadly construed, which could be appealed after six months of constructive editing in other subject areas while they learn the ropes. Six months of editing in the walled garden of their sandbox will not help them develop as a collaborative editor. GirthSummit (blether)03:34, 1 October 2025 (UTC)[reply]
In light of their recent comments (referenced below by jpgordon and others), I would no longer support an unblock with TBan in place. Oppose any unblock at this time. GirthSummit (blether)00:28, 2 October 2025 (UTC)[reply]
Comment I took a peek at the de-WP draft.[6] This is IMO the kind of article that will often contain a lot of WP:SYNTH and WP:OR. For example, we have under the heading "Ethical understanding of well-known members of Freemasonry" (I'm using google translate, I don't speak german) the sentence "Harry S. Truman ordered the atomic bombing of Hiroshima and Nagasaki" with two cites. That is uncontroversial enough, but the cites makes no mention that this has any connection to Truman's FM, and is contradictory to it. An en-WP article about "Contradictions in Freemasonry", is supposed to be a summary of WP:RS that discuss that subject, not a list of selected examples the WP-editor finds "contradictory". Gråbergs Gråa Sång (talk) 05:44, 1 October 2025 (UTC)[reply]
Oppose unban I speak fluent German, and had a thorough look at the draft over there. The title of the draft itself has already an npov-tilt, and we already are, at least in part, in OR, but more so in WP:SYNTH territory. A TBAN from freemasonry, broadly construed, including the draft would work imo. Lectonar (talk) 09:46, 1 October 2025 (UTC)[reply]
Just want to add that even if an article was accepted in DE-Wikipedia, it would not mean that it would be automatically accepted here. The rules for sourcing and notability are not identical. Lectonar (talk) 09:48, 1 October 2025 (UTC)[reply]
Oppose any unban not accompanied by a complete TBAN for at least a year. And after that, only lift the TBAN if Wikiprediger has done significant useful work on other topics. It seems entirely evident (see e.g. the comment about 'censorship' above) that Wikiprediger wishes to continue with the same POV-pushing that led to sanctions in the first place. We don't need that. And nor do we need synth-ridden POV forks padded out with random stuff that seems to have little to do with the supposed topic. Even if such nonsense is acceptable on de.wikipedia (which would surprise me), it won't be here. AndyTheGrump (talk) 13:36, 1 October 2025 (UTC)[reply]
Support unblock only with a broadly-construed topic ban to freemasonry definitely including the draft I think it would be unwise to allow a person who was cbanned for edit-warring over freemasonry to make what appears to be a WP:COATRACK article. Simonm223 (talk) 14:26, 1 October 2025 (UTC)[reply]
Support unblock, but only with a topic ban on freemasonry, with no exception for any current drafts. If we trust this editor to write about freemasonry, there's no reason to prevent them from writing about freemasonry. And if we don't trust this editor to write about freemasonry (I'm in this camp), then there's no reason to allow any edits about freemasonry. I am explicitly against any sort of St. Augustine "Lord, give me a topic ban on freemasonry, but not yet!" nonsense. There's nothing preventing this editor from saving their draft on their own device and working on it offline, but allowing them to work on this topic on Wikipedia is a terrible idea.CoffeeCrumbs (talk) 18:55, 1 October 2025 (UTC)[reply]
Oppose unban after seeing the latest comment on their talk page that is referenced below. The veiled threats in order to exert leverage over the community in order to get to edit this article make me seriously question if the problems collaborating of others would be limited to the freemasonry topic. CoffeeCrumbs (talk) 23:16, 1 October 2025 (UTC)[reply]
Oppose unban I was planning on staying out of this discussion, since I started the one that got him cbanned in the first place, but after reading his last comment, where he seems to claim we'll damage public trust in Wikipedia by not letting him edit on the topic, I don't think an unblock is appropriate at this time. --SarekOfVulcan (talk)19:24, 1 October 2025 (UTC)[reply]
Oppose unban. Just wow: I cannot accept being told I am not allowed to work on this subject...If it is blocked here, I will always note on my user pages that the ban in en:WP feels like censorship. I will also try to find other solutions to be allowed to speak... A ban on me may not prevent the subject from appearing sooner or later — it can only prevent that I take part in the discussion. I don't see how this person's attitude aligns with Wikipedia's policies and polity. --jpgordon𝄢𝄆𝄐𝄇00:19, 2 October 2025 (UTC)[reply]
(edit conflict)Oppose unban in its entiretyThis says it all: I cannot accept being told I am not allowed to work on this subject. This is the reason I came to en:WP, and I want to continue to translate this article into other language versions (French, Italian,...). If it is blocked here, I will always note on my user pages that the ban in en:WP feels like censorship., If en:WP is not prepared to allow critical discussion of certain topics, it may have a clear effect on how the project is perceived from outside. In the end, it may be the en:WP community that is questioned for this decision., andI will also publish the article in other languages step by step, so that all major Wikipedias will eventually cover this topic. If readers are then told that such an article cannot exist on the English Wikipedia because of a ban related to conflicts about Freemasonry more than a year ago, it will naturally raise questions and may reflect poorly on Wikipedia. makes it very clear that this is a user who is not here to help build an encyclopedia. They have no interest in editing topics other than their pet hobbyhorse, and speaking frankly if they were to be unbanned with a topic ban I can't help but have a feeling they would try to build a good reputation, get the topic ban lifted, and immediately push that POV. - The BushrangerOne ping only00:24, 2 October 2025 (UTC)[reply]
Oppose any unban. I just don't see any indication that this person actually wants to edit constructively on other topics. At best, I think we'd get the situation Bushranger described, but I don't even think they'd get that far given their comments. ♠PMC♠ (talk)00:48, 2 October 2025 (UTC)[reply]
The discussion above is closed. Please do not modify it. No further edits should be made to this discussion.
Number 6 might be shoehorned in under "character handling issues", and Commons:File naming suggests not abusing Unicode, and that "strange punctuation can be replaced with standard quotes and commas".
I'm happy for the curly d' to be replaced. I didn't upload it as 'strange punctuation'. It was just my file name, so there's a weakness in the Commons processing of uploads of the apostrophe. Charlesjsharp (talk) 20:12, 1 October 2025 (UTC)[reply]
Moving them all to draftspace would be an admin action and I want to make sure I'm doing the right thing. I think it would be the right thing, but then I've seen other admins get in to trouble to going in without discussing at all first. FOARP (talk) 18:50, 3 October 2025 (UTC)[reply]
having failed to get the desired result at afd, it turned into a bizarre proposal about the hypothetical creation of already created articles at vpp, and here it is coming to fruition as a strange end-run around the deletion process. If you want to delete or draftify, at minimum you need consensus... on a question about deleting or draftifying. — Rhododendritestalk \\ 19:27, 3 October 2025 (UTC)[reply]
I understood it that way. You said explicitly "This RFC isn't about deletion" (which would then follow it wasn't about draftificaiton) in the RfC. Please see User talk:JoelleJay#Our mutual friend.
That being said, I have the capability to draftify all the List of Olympians articles myself and am willing to do that. Some care needs to be taken because I understand there are redirects with history (i.e. BLARed articles) that redirect to the list, and just draftifying the list pages would delete them. Ideally the by-sport list recommended would be created instead, and the redirects would be retargeted to rows in that list (or an arbitrary row e.g. for multi-sport athletes). --Habst (talk) 05:26, 5 October 2025 (UTC)[reply]
Thanks for your comment, from which I gather you're OK with drafting these articles. I've draftified all the articles. I think, given the result at VPP, you should make a trial version first in draft rather than just mainspace the lot. FOARP (talk) 15:32, 5 October 2025 (UTC)[reply]
Are the redirects specific to an alphabetical listing of all Olympians? If so they are unlikely to be needed. Redirects are anyway cheap, and not a reason to block draftification whilst the issue identified in the VPP discussion are addressed, if doing so is otherwise acceptable. FOARP (talk) 15:49, 5 October 2025 (UTC)[reply]
Per my comment, I'm talked about BLARed articles with page history, which are not WP:CHEAP. They are not specific to an alphabetical listing of all Olympians. --Habst (talk) 15:53, 5 October 2025 (UTC)[reply]
There's no rush to have these moved to draftspace. We can take our time to ensure proper care is given. FWIW, that proposal was so confusing ("is there support for creation of lists that already exist?") that I didn't really participate (aside from pointing out a few falsehoods), and probably others thought similar. BeanieFan11 (talk) 15:58, 5 October 2025 (UTC)[reply]
I'm OK with taking time, so long as it does not simply become obstruction-through-technicalities. It would be good for you to explain what exactly you want to take time for, now that the VPP discussion already closed some days ago. FOARP (talk) 16:07, 5 October 2025 (UTC)[reply]
According to Habst, there are Olympians that redirect to some of those lists. We need to figure out what happens to them before moving the lists. BeanieFan11 (talk) 16:17, 5 October 2025 (UTC)[reply]
The most obvious large-scale, reasonable sorting would be a "List of Olympics from <country>", which still likely would need to be sub-listed to some extent. I'm surprised I'm not quickly finding such lists to start with. Masem (t) 16:29, 5 October 2025 (UTC)[reply]
I'm actually at the Wikipedia NYC hackathon right now and met some users who commented in the RfC IRL -- it was kind of surreal to have these types of conversations in person and has given me a new perspective. Please give us some time while the hackathon is ongoing, there is a plan for the future. But yes, there are Olympians that redirect to the lists. --Habst (talk) 16:20, 5 October 2025 (UTC)[reply]
(edit conflict) Habst, please undo your moves to mainspace. Speaking as someone genuinely undecided on a list of all Olympians, there is no way to read the VP discussion as anything other than a consensus against their existence in mainspace at this time, and so moving them back to mainspace is effectively ignoring community consensus, regardless of what your intentions might have been. Broken redirects are a triviality. We have tools to deal with them. Vanamonde93 (talk) 16:24, 5 October 2025 (UTC)[reply]
We can wait to figure out an appropriate solution for the redirects. Rushing and deleting redirects with important history is not appropriate. BeanieFan11 (talk) 16:27, 5 October 2025 (UTC)[reply]
You realize you could draftify the redirects? Or ask for a refund, if they were deleted in the meantime? There's a ton of options here that don't involve ignoring an active consensus. Vanamonde93 (talk) 16:37, 5 October 2025 (UTC)[reply]
OK, I will do that after the Wikimedia hackathon is over today. I did not move them all back to mainspace; only a few were moved at the time that FOARP said "I've draftified all the articles". Per an in person discussion I had, some care is needed because the existing tools to deal with broken redirects would delete them via bot or move them to draftspace, which would remove the Wikidata link. Doing a refund would not restore the Wikidata link. They are quite important. --Habst (talk) 16:39, 5 October 2025 (UTC)[reply]
@FOARP, please give us until the end of the day here. We are at a Wikimedia hackathon and the contents of the lists are being used as part of a project about to be presented. Also, there are redirects that need to be retargeted. That's completely reasonable. --Habst (talk) 17:57, 5 October 2025 (UTC)[reply]
As an admin, you should absolutely not be edit-warring and fully protecting the pages to get your way. What is unreasonable about asking for one day? Both of you need to stop it. BeanieFan11 (talk) 17:59, 5 October 2025 (UTC)[reply]
Guys, this had already been explained to you, and not just by me. This is not me "getting my way". This is the consensus that these articles don't belong in mainspace being carried out. Vanamonde93 already explained to you what to do about the redirects. FOARP (talk) 18:09, 5 October 2025 (UTC)[reply]
What should we do about the redirects? Also, this is being used in part for an in person project at the Wikipedia hackathon right now. Please give us until the end of the day. --Habst (talk) 18:11, 5 October 2025 (UTC)[reply]
"You realize you could draftify the redirects? Or ask for a refund, if they were deleted in the meantime? There's a ton of options here that don't involve ignoring an active consensus."FOARP (talk) 18:15, 5 October 2025 (UTC)[reply]
(ec) Once again, you, as the main editor in this dispute, are not in a position that would make it appropriate to fully-protect the pages while refusing a polite request for a mere one day to look into the redirects further. What is unreasonable about the request? This is disappointing to see from an administrator. BeanieFan11 (talk) 18:16, 5 October 2025 (UTC)[reply]
I responded to that -- draftifying would remove the Wikidata links which a refund would not restore, and the drafts would need to be decategorized. There is a plan to do this properly, I am busy at the Wikipedia hackathon right now. --Habst (talk) 18:18, 5 October 2025 (UTC)[reply]
I am not so fussed about the timeline as to suggest a sanction for a few additional hours. But I am assuming that what you have is a real plan to implement the VPP consensus; that is, a plan that ends with these pages in draftspace, soon. If this is merely a delaying tactic, I do think we will be looking at sanctions here. That said, FOARP, that's not a great protection. Vanamonde93 (talk) 00:25, 6 October 2025 (UTC)[reply]
Thank you for assuming good faith. I have completed the archival of redirects and draftified the list articles per Wikipedia talk:WikiProject Olympics#Keeping the articles. My hackathon project won the "Best Game That Contributes Back to a Wiki" award and is relevant to these discussions we've been having; I look forward to improving it more and rolling it out to help bridge the gap between both sides on these issues. --Habst (talk) 01:20, 6 October 2025 (UTC)[reply]
I closed the AfD in question, and in my closure recommended a community-wide RfC about the intersection of NOTDB and NOTDIR with lists of sportspeople. I also explicitly noted that there was no consensus on the existence of these lists, and that further discussion would be needed as to their inclusion or exclusion from Wikipedia. The RfC that was launched was much narrower, and I might perhaps have worded it differently, but I don't think it is reasonable to argue that it was out of process: it is the logical continuation of the dispute that did not conclude at AfD, and there is precedent for large-scale draftification or deletion following discussions at VPP. Vanamonde93 (talk) 19:44, 3 October 2025 (UTC)[reply]
This is my thinking on the matter, and also having discussed this with an experienced former admin (FRAM) their view as well. FOARP (talk) 18:41, 4 October 2025 (UTC)[reply]
A recent Phabricator ticket (T405999) added the abusefilter-modify-restricted right to English Wikipedia edit filter managers to allow non-administrator edit filter managers to modify filters using the "Revoke the user's autoconfirmed status" action. The change's deployment and the extent of prior community discussion and support have been raised as issues on the ticket. It may also be a good opportunity to review how configuration changes are evaluated and deployed following community discussions. Daniel Quinlan (talk) 09:12, 4 October 2025 (UTC)[reply]
@Daniel Quinlan can you clarify what the intent of this section is - is it purely informational? Are we discussing (here? on the phab ticket?) whether we want this change on en.wp? Something else? Thryduulf (talk) 10:27, 4 October 2025 (UTC)[reply]
@Thryduulf: Original proposer here. The question being raised is whether the change has the support of the broader community. I maintain that managing edit filters is within the remit of EFMs, and the abusefilter-modify-restricted right only allows the additional ability of editing filters with "Revoke the user's autoconfirmed status" used as an action (this allows for socks who game autoconfirmed before being disruptive to have their autoconfirmed revoked automatically). It is worth noting that in the original RfC covering assignment of the EFM right to non-admins, the original text stated (and the current text at WP:EF still states): The assignment of the edit filter manager user right to non-admins is highly restricted. It should only be requested by and given to highly trusted users; when there is a clear, demonstrated need for it. Demonstrated ability that one can and will use it safely is absolutely critical. It seems unnecessary to separate filter editing abilities away from non-admin EFMs, when those same non-admin EFMs are trustworthy and capable of editing those filters. If we proposed that disallow filters not be editable by non-admin EFMs, the proposer would be laughed out of the room. I fail to see how it is any different for revoke autoconfirmed.
I note also that non-admins with the edit filter manager group can already restore autoconfirmed when it is revoked by a filter that is set to revoke autoconfirmed (see Special:AbuseFilter/tools with an account with EFM added). Edit filter managers can also add all actions, including disallowing edits, even before the change. The only difference is that they are now able to add revoke autoconfirmed. The question being asked here is whether the community trusts non-admin edit filter managers to, well, manage edit filters. I personally didn't think it would be that controversial of a change, though I welcome Daniel bringing it here since apparently I underestimated that. EggRoll97(talk) 16:29, 4 October 2025 (UTC)[reply]
So for clarity if the editor is already autoconfirmed the filter action will do nothing, it only will stop editors receiving autoconfirmed when they don't have it? Is it a permanent change to the editor so other edits not caught by such a filter will also not grant autoconfirmed or does it only mean those edits don't count? Nil Einne (talk) 18:43, 4 October 2025 (UTC)[reply]
I think this is an issue of interest to administrators. The post is both informational and for discussion. It's not a call for a specific action, at least not yet. I'm most concerned that the community wasn't adequately notified and that significant configuration changes can be implemented with minimal input. At the very least, the community needed to be made aware of the change. I don't think Phabricator is the best place for discussion so here is fine. Daniel Quinlan (talk) 16:34, 4 October 2025 (UTC)[reply]
How many people does this actually apply to? It seems we only have a handful of non-admin EFMs, and even fewer who aren't former admins. I am fine with the proposed change, although it would seem better to me to just promote the relevant people to sysop. (It should be easier than a software configuration change, but I know it isn't). —Kusma (talk) 16:40, 4 October 2025 (UTC)[reply]
It applies to 16 people and one bot. Of the people, I would estimate only about a third are actively involved with edit filters, and six have fewer than 20 edits in 2025. Daniel Quinlan (talk) 16:54, 4 October 2025 (UTC)[reply]
My admin coloring script (because I like to have a colorful display and all) shows me 16 non-admin EFMs, of which 8 are former admins (and one is a former steward). There is another EFM account that is a bot, run by one of the current EFMs, which does some minor updates to one specific filter. That makes 8 non-admins who have passed a request and community consensus process to be assigned EFM, with the rest passing RfA. I'd say promote the relevant people to sysop is somewhat of a harder sell. Some pass, though it's worth noting I tried that path and didn't succeed (see Wikipedia:Requests_for_adminship/EggRoll97_2), for good reasons. EggRoll97(talk) 16:56, 4 October 2025 (UTC)[reply]
If this is a call for opinions, my opinion is that edit filter managers can already cause all sorts of problems vastly more efficiently (and often with less oversight) than any admin. This adds fairly minimal additional vectors for abuse or poor judgement to very trusted editors, and makes their lives easier. Rusalkii (talk) 21:04, 4 October 2025 (UTC)[reply]
That's not an unreasonable take about the specific change. As I mentioned, I'm most concerned that the community wasn't adequately notified and that significant configuration changes may be implemented without sufficient notification and input. Regarding less oversight, I think there's far more oversight for filters than most administrator actions (blocks, deletions, page protections, etc.) because users are able to report false positives to WP:EFFPR every single time an edit results in a significant action (e.g., warning or disallow), WP:EFFPR is monitored by a broad set of people, and administrators, edit filter managers, and edit filter helpers all regularly review and iteratively improve filters. To put it another way, there are only hundreds of active filters, but there are many orders of magnitude more administrator actions that go without substantial review. It's much easier for some types of problematic administrator actions to go unnoticed for a long time (e.g., dubious deletions). Daniel Quinlan (talk) 21:25, 4 October 2025 (UTC)[reply]
That's fair. I meant there's generally more eyes on deletions than edit filter changes, but if reporting false positives is common reasonably common then it seems like problematic changes might be more likely to be caught, I know there's a mechanism for it but don't have. a great sense of how it's used in practice. Rusalkii (talk) 00:49, 5 October 2025 (UTC)[reply]
Can someone figure out what is going on with these tropical storm drafts?
The following discussion is closed. Please do not modify it. No further edits should be made to this discussion.
Apologies in advance that this will be a bit scatter-brained. Writing quickly on mobile before I go offline for a few hours.
Yesterday I warned three editors for an edit war over Draft:Matmo, which for whatever reason is now at Draft:Tropical Storm Matmo 2025, and at the parallel draft Draft:Tropical Storm Matmo (2025), which after another strange move eventually found its way to be the article Typhoon Matmo (2025). I really can't tell what's going on here—why there have been so many parallel drafts, and why everyone is edit-warring over them rather than merging them—but it was clear that the BLAR-warring was against policy (citing a misunderstanding of the sock policy with respect to proxies, plus an extremely bizarre appeal to MOS:DATE), so I issued my warnings and left it at that. Liz also gave one user a warning for bad page moves.
To those users' credit, they've stopped edit-warring, but now I see an edit war continues about whether Draft:Tropical Storm Matmo 2025 should be a redirect. On the one hand, I don't see why it should exist parallel to the article, but on the other hand, the two users to blank it have included personal attacks in their summaries, one making unsubstantiated allegations of sockpuppetry and the other (who also incorrectly speedy-tagged the redirect) saying "rvv". It's also not clear if all the merging that's happened has followed WP:CWW.
I'm not pinging the involved users just yet because I'm not trying to start a thread to collectively discuss their content, and more hoping another admin can step in and figure out what combination of warnings, protections, and maybe blocks are necessary here. To be clear, for all I know it's correct that the IP is a sock, but so far in this saga they've been the only one to communicate clearly. Everything else here feels like some alternate reality version of Wikipedia.
Wouldn't life be so much easier if we got rid of draft space and developed articles in main space, like we used to? This is supposed to be a wiki, after all. Phil Bridger (talk) 13:10, 5 October 2025 (UTC)[reply]
@Tamzin:, I'm an admin who knows the topic area pretty well. So it looks like Draft:Tropical Storm Matmo 2025 was created first by the IP on October 1, and a different draft was made a day later. The latter one has the edit history, so I don't think the draft needs to be preserved for edit history reasons. However, I now see there have been lots of edit wars over these drafts, so I don't know if they should remain for that reason. If not, it can be deleted as an unnecessary fork. ♫ Hurricanehink (talk) 17:34, 5 October 2025 (UTC)[reply]
@Hurricanehink: Not too sure if I should chip in here. Just one quick point: The "latter", as you called it, was the fork instead. And the one created on 1 October got considerably longer edit history (i.e. with more revisions) and has been kept more up-to-date than the fork.[7]203.145.95.215 (talk) 17:54, 5 October 2025 (UTC)[reply]
To repeat myself: They don't seem to be interested in updating their article after occupying the main article namespace forcibly. They don't even mention in their article under its meteorological history section the landfall of the storm on the Leizhou peninsula earlier on 5 October, after almost twelve hours since it happened (when the storm has already re-emerged on the other side of the peninsula). 203.145.95.215 (talk) 17:56, 5 October 2025 (UTC)[reply]
The discussion above is closed. Please do not modify it. No further edits should be made to this discussion.
Deletion requested due to unauthorized publication by the university
The following discussion is closed. Please do not modify it. No further edits should be made to this discussion.
This page about Ahlul Bayt International University was created without authorization from the university and contains potentially inaccurate or promotional information. Deletion requested by the official representative of the university.
Wikipedia articles do NOT require authorisation, however that article is on The Persian Wikipedia which is an entirely seperate project you will need to gom there to request deletion. Theroadislong (talk) 13:34, 5 October 2025 (UTC)[reply]
@WikiBuilder2025: that article is in the Farsi Wikipedia, which has nothing to do with us here at the English Wikipedia; each language version is a completely separate project. If you wish to discuss that article, you need to contact the Farsi Wikipedia's editors and/or administrators. -- DoubleGrazing (talk) 13:34, 5 October 2025 (UTC)[reply]
Forwarding appeal on behalf of @Jack6754 for unblock request @ User talk: Jack6754. Their reformed request as suggested by @Yamla (the declining admin) and myself:
"Hello, I am requesting an unbanning from editing Wikipedia, as well as a lifting of the blocks imposed upon my account. I have reflected on my actions over the past few years and am apologetic about my previously disruptive and disrespectful behaviour. My previous account is subject to a community ban and indefinite block due to my past actions as an editor, and I regret being involved in those actions.
I accept that I have edit warred in the past and I am sorry for doing so, and will take pains to refrain from doing so in future. The article 'Black War' which I edit warred on is quite an emotional subject for me and I apologise for allowing that to cloud my judgement, an edit war was not an appropriate means to resolve the dispute of the way that dark historical topic ought to have been described by the opening lede of an article.
I am also sorry for my unproductive contributions to AfD years ago. I've matured in my approach to the site since then, and have gotten a little older. I'm confident that I am better equipped to participate as a productive article creator without getting bogged down in these kinds of procedural arguments that exist in the background.
I would like to resume contributing to the articles on this website. I love creating articles and would like to get back to being a productive editor. It has been over six months since I previously edited the site in any way, and I hope that my willingness to comply with my previous ban demonstrates that I have a newfound respect for the rules and have the ability to be a positive member of the community in the future.
I'm requesting that my account be unbanned 'standard offer' as I have complied with all these restrictions for six months. My intention upon being unbanned is to resume contributing; especially in relation to articles about Australian law, history, music, culture, and architecture.
I would be happy to have a 0RR restriction be imposed on my account as a way to show a willingness to comply with edit warring rules given that I have breached these rules in the past; especially if this restriction would give greater faith in my ability to contribute as an editor. Thank you for considering this request."
Support with the WP:0RR restriction. I think that's enough to minimise disruption and to give them the opportunity to demonstrate they can be constructive. --Yamla (talk) 21:21, 5 October 2025 (UTC)[reply]
Support with the WP:0RR restriction as a WP:LASTCHANCE (and I really mean last chance, next time is an indefinite block). They are currently topic banned from deletion discussions; being unblocked does not lift that topic ban and you are still not permitted to discuss AfDs on your talk page, even pointing out relevant policies and guidelines (which did happen previously). Before being sitewide blocked, Jack had indefinite pblocks from WP: and WT: space; those should be reinstated upon the lifting of the siteblock.
This is a laundry list of restrictions, and having reviewed Jack's previously unblock attempt I was left supremely unimpressed with the wikilawyering surrounding the standard offer and their topic ban they engaged in: see generally User talk:Jack4576#Unblock Request. (To Jack's credit, they were not the one engaging in most of the wikilawyering nor did they join the argument that they were permitted to criticize AFD nominations on their talk page.) That being said, I am persuaded that Jack now understands that they were disruptive, and that they can move forward carefully. I sincerely hope that Jack can contribute with fantastic edits and, in due time, appeal each sanction. HouseBlaster (talk • he/they)22:47, 5 October 2025 (UTC)[reply]
Support - I was involved in the page on at least one of the affected pages, but Jack is a content creator with some good work to show for it. I'd be inclined to accept under the Standard Offer. This support applies to lifting the block. If there are topic bans I am unaware of, I have no view on those at this time, but would remain supportive of a lift of the indef. Sirfurboy🏄 (talk) 06:24, 6 October 2025 (UTC)[reply]
Support Thanks for reflecting on your actions and taking the time to write this out, Jack. I see you can recognise where you may have gone wrong. I'm convinced you'll contribute constructively from now onwards. jolielover♥talk16:50, 6 October 2025 (UTC)[reply]
Support Thank you for apologising and recognising your fault I support your unblock request as I beleive you have recognised your fault and are willing to contribute productively now. GothicGolem29 (talk) 23:41, 6 October 2025 (UTC)[reply]
The following discussion is closed. Please do not modify it. No further edits should be made to this discussion.
I have been erroneously labeled a sockpuppet of WP:WIKINGER on the said page.
This stems from the aforementioned LTA creating sockpuppets with names similar to mine as trolling. I cannot edit the page due to insufficient permissions, this is important as I do not want to be labeled as a LTA. As you can see in my edit history, I am a genuine editor with significant activity (over 7000 edits) on Polish Wikipedia and engaged in reporting LTA sockpuppets. Jeż0216 (talk) 17:37, 6 October 2025 (UTC)[reply]
A block for violations of the community ECR would be imposed and logged under WP:GS/KURD. However, if those ECR violations also violated CT conduct expectations (WP:CT#Editing a contentious topic) and you don't believe they are unaware of the CT designation (WP:CT#Awareness of contentious topics), it can instead be imposed as CT enforcement and logged at AEL. (Logging a sanction as both GS and CT would be potentially confusing (e.g., for appeals).) ~ Jenson (SilverLocust💬) 21:29, 7 October 2025 (UTC)[reply]
I am categorically opposed to TBANs from the entire BLP topic area; they're always either overkill (as in a narrower sanction would suffice) or underkill (as in the user should be sitebanned). For that reason, my baseline here is that I'd support a step-down restriction to "biographical material about living persons that Stuartyeates has any degree of personal connection to", which would cover the scenario that led to this TBAN's imposition. But that's mostly a procedural view, matching what I would say if the AN/I thread were to happen today. On the merits, I lean opposed to any lifting beyond that, because this appeal does not address the edits that led to the ban. -- Tamzin[cetacean needed] (they|xe|🤷) 08:14, 8 October 2025 (UTC)[reply]
While I agree with Tamzin in spirit, I think it would be difficult to manage narrowing the scope to people with whom appellant has a conflict of interest. How would we know? Having said that, perhaps @Stuartyeates: can more fully address the edits that led to the topic ban.---- Deepfriedokra (talk) 08:32, 8 October 2025 (UTC)[reply]
In terms of the edits that lead to the TBAN: the last time I checked (immediately prior to my previous appeal), none of the articles I created had been deleted, the content of the edits were still on en.wiki supported by the same sources, although some had been moved to different pages, particularly a consolidation to Listener letter on science from BLPs, including newly minted ones. Creating a page on the issue was probably what I should have done to start off with, rather than repeating similar material on several BLPs. The social media posts which most of the original discussion centred on are long gone, since I'm no longer on twitter. I'm more careful about the perception of off wiki coordination, for example the draft I helped with at Draft:Blacksky (which I helped with based on social media mentions); I won't be making any decisions about promoting that to mainspace. I'm also being more selective about the sources I use in BLPs (more aggressively avoiding potentially ethnically partisan sources) and double-sourcing some claims. Stuartyeates (talk) 09:13, 8 October 2025 (UTC)[reply]
We would know if Stuartyeates decided to disclose them. Or we can look at the persons they got banned over, note their similar qualities, and decrease the topic ban accordingly 178.246.132.152 (talk) 11:22, 8 October 2025 (UTC)[reply]
If we were to decide that Tamzin's proposed narrowing is unenforceable but that narrowing the TBAN would be desirable in principle, would living New Zealand academics be a workable scope? It's easier for editors other than Stuart himself to identify the boundaries of, and seems to cover all of the articles in the precipitating incident. If that's too narrow, then we could have either living New Zealanders or living academics. Caeciliusinhorto-public (talk) 15:32, 8 October 2025 (UTC)[reply]
After a motion, arbitration enforcement page protections no longer need to be logged in the AELOG. A bot now automatically posts protections at WP:AELOG/P. To facilitate this bot, protection summaries must include a link to the relevant CT page (e.g. [[WP:CT/BLP]]), and you will receive talk page reminders if you forget to specify the contentious topic but otherwise indicate it is an AE action.
Sections inactive for over seven days are archived by Lowercase sigmabot III.(archives, search)
.) I'll see if I can sort this out myself, but if any admin knows this topic area well, thoughts/intervention still welcome. -- Tamzin[cetacean needed] (they|xe|🤷) 15:44, 5 October 2025 (UTC)
Administrator changes
_male_Montagne_d%E2%80%99Ambre_2.jpg){kind=link}
_male_Montagne_d%E2%80%99Ambre_2.jpg){kind=link}
_male_Montagne_d%E2%80%99Ambre_2.jpg){kind=link}
_male_Montagne_d%E2%80%99Ambre_2.jpg){kind=link}
CheckUser changes
Arbitration