Wikipedia:Edit filter/Requested

{{User:ClueBot III/ArchiveThis }} {{archives|search=yes}}
 * age=720
 * minkeepthreads=4
 * archiveprefix=Wikipedia:Edit filter/Requested/Archive_
 * format=%%i
 * maxarchsize=900000
 * numberstart=21
 * index=no
 * archivenow= {{archive now}},{{User:ClueBot III/ArchiveNow}},{{resolved,{{stale,{{EFR|denied,{{EFR|impossible
 * header= {{talkarchive}}
 * headerlevel=2

User:Drmies wants a filter
Nyttend (talk) 12:54, 24 June 2024 (UTC)
 * Task: Drmies left a note at WP:AN linking to Special:Contributions/Learoy4, all of whose edits had the same summary, all of which have been revdelled as Grossly insulting, degrading, or offensive material. Example, if you're an admin and able to see the revdelled content
 * Reason: Based on the AN request, I suspect that this summary is being used by other accounts or IPs. Drmies blocked Learoy4 for vandalism, so we won't see further problems from this account.
 * Diffs: Every edit by Learoy4 has the same summary, so preventing further edits by other accounts or IPs should be trivial. If the wording is changed a little, well, you're the filter maintainers and I'm not; maybe you could find a way to make things better.


 * @Nyttend, @Drmies: see the comments for Special:AbuseFilter/1314. —Ingenuity (t • c) 14:57, 24 June 2024 (UTC)
 * , you can check 's log--I blocked a few but they blocked more. Thanks. Drmies (talk) 15:00, 24 June 2024 (UTC)
 * Yup. Dozens of them. It's some known LTA case, but I don't care which. I'll keep playing whack-a-vandal while I can. Keep an eye on 1314. —Smalljim  16:19, 24 June 2024 (UTC)
 * Maybe, just maybe, there's a tiny chance if we can set to disallow or add the summary regex to 52 and disable 1314, but further discussions should not happen here. Codename Noreste  🤔  Talk  04:01, 25 June 2024 (UTC)
 * Just curious, is there any page where discussion can happen securely (something requiring admin or filter-editor rights just to view) without relying on the email address provided in the edit notice? I've looked at filter 1314's notes, and I can see people saying "To explain why..." and "Is this so-and-so" (as Ingenuity recommends), but nowhere that's being used for discussion.  Nyttend (talk) 10:03, 25 June 2024 (UTC)
 * Nope. I've proposed a private wiki for facilitating this kind of discussion before but it did not get much traction. The current canonical venue is always the mailing list. 0x Deadbeef →∞ (talk to me) 10:53, 25 June 2024 (UTC)

I've tested possible code for this filter on Test Wiki (see here), and it seems to work well.   [[User:CanonNi ]]  (talk • contribs) 13:12, 25 June 2024 (UTC)
 * Task: Prevent the removal of past AfC decline and rejections.
 * Reason: They're not supposed to be removed by non-reviewers. (There's a invisible comment that says  beside the templates)
 * Diffs: A lot.


 * Looks good, except you forgot exempting new page reviewers in the test wiki code, so maybe make it something like ? –  Pharyngeal Implosive7   (talk)  01:36, 26 June 2024 (UTC)
 * Ultimately it doesn't particularly have much effect, since I can't really think of any patroller who isn't extendedconfirmed already. The only ones who would be are bots, who already operate with a bot flag. EggRoll97 (talk) 03:02, 26 June 2024 (UTC)
 * True. I didn't think of that, but one might keep it there just to be safe? –  Pharyngeal Implosive7  (talk)  01:52, 27 June 2024 (UTC)
 * This probably needs wider discussion. I'd support it, but I suspect the anti-draftspace people would object. At a minimum, should probably make a post at WT:AFC. – Novem Linguae (talk) 02:01, 26 June 2024 (UTC)
 * Good idea. I've posted a please see there.   [[User:CanonNi ]]  (talk • contribs) 03:43, 26 June 2024 (UTC)
 * meh. There are two situations where the AFC submission tags are being removed. In the first case, the draft-writer is attempting to hide past declines and/or unaware that they shouldn't replace declines with a new submit tag. In the second case, someone (and it could even be the draft creator) is moving the draft to the article space, which meets the before article has been created clause of the hidden comment. Can the filter tell the difference between these two cases? If not, then I do not think it will be a helpful filter (unless it is log-only). Primefac (talk) 13:19, 26 June 2024 (UTC)
 * I think it can. The  line is used to not catch drafts that were turned into redirects (likely from a page move).   [[User:CanonNi ]]  (talk • contribs) 13:26, 26 June 2024 (UTC)
 * I suppose my concern is if someone wants to clean up the draft before they move it to the article space, it will flag it as a violation, no? Primefac (talk) 14:57, 26 June 2024 (UTC)
 * Hmm... that's a good point. Maybe the template can say something like "Only remove this template if the draft has been moved into mainspace."?   [[User:CanonNi ]]  (talk • contribs) 02:14, 27 June 2024 (UTC)
 * Setting the filter to warn rather than disallow as you propose sounds like a good compromise. – Novem Linguae (talk) 03:14, 27 June 2024 (UTC)
 * I can see that a draft is OVERWRITTEN by a different draft. That could cause an issue here. There is no collision detection at Article Wizard, so if you select an existing draft article name, and create a new draft, that will delete any rejection notices with a fresh draft. I've seen different users create new drafts overwriting one another. -- 64.229.90.32 (talk) 07:08, 1 July 2024 (UTC)
 * I expect that trapping conversion to redirect would help with if someone merges a nonnotable-rejection into a broader topic draft that could be notable. the Merge-and-Redirect activity would capture the edit history as a redirect's contribution history. ? -- 64.229.90.32 (talk) 07:08, 1 July 2024 (UTC)
 * Support Apart from the correct housekeeping removal on acceptance, ideally but not always done by the AFCH script, I see only two reasons an editor, not necessarily the creating editor, will remove the material:
 * With goodwill, thinking this is correct despite the hidden comment
 * To conceal prior review history.
 * I see this proposal as a benefit provided the exception cases are sorted out. I have no objection to offering a warning, though would prefer outright prohibition. 🇺🇦 Fiddle Timtrent  Faddle Talk to me 🇺🇦 07:58, 1 July 2024 (UTC)

COI filter
Rusty talk contribs 23:21, 30 June 2024 (UTC)
 * Task: Prevent edits common COI edit summaries
 * Reason: Reduce the workload of patrollers, help out new users who may be unfamiliar with Wikipedia's policies.
 * Diffs: Don't have any on hand right now, but generally use phrases like "I am/We are ______ and am/are updating the article...", etc.
 * That wouldn't be in keeping with policy. COI edits are discouraged, but not outright forbidden. We certainly should not be preventing COI editors from removing obvious BLP violations, vandalism, etc. Spicy (talk) 13:31, 1 July 2024 (UTC)
 * Would you also object to a warn-only filter? This would certainly be in line with "discouraged, but not outright forbidden". Animal lover &#124;666&#124; 12:33, 7 July 2024 (UTC)

Warn about a Wikipedia mirror
Ed-Tech Press, also known as "Scientific E-Resources, is a Wikipedia mirror. They print copies of books that are just Wikipedia articles.  Per WP:CIRCULAR, we should never cite them in articles.  Unfortunately, these books are listed in Google Books, and there's no obvious warning on them.  I've inadvertently cited them twice recently.  While I really appreciate reversions like this one, it seems like this is an area where an ounce of prevention is worth a pound of cure.  Could we please have an abuse filter set up for this string:

which should catch most cite book uses? If it would be great if it could produce a warning message like "Ed-Tech Press and Scientific E-Resources are Wikipedia mirrors. They are not reliable sources and should not be cited in articles per WP:CIRCULAR." I think that the 'warn' setting should be sufficient. WhatamIdoing (talk) 01:42, 7 July 2024 (UTC)
 * Thank you making this request - this publisher is just the worst. There is deliberately no attempt to identify the nature of the copied materials; it's just a straight up scam. There are three things I usually search for: "Ed-Tech Press", "Scientific e-Resources" (which is typically displayed when a google books link is resolved in a template), and the URL of "edtechpress.co.uk". I do agree with the warning being sufficient as I don't recall this ever being used on-wiki by a bad-faith actor. Sam Kuru (talk) 02:35, 7 July 2024 (UTC)
 * Yeah. Possible filter code for catching this could be:

page_namespace == 0 & !contains_any(user_groups "bot", "sysop", "extendedconfirmed") & (  mirrors := "(?:\|publisher\s*\=\s*(?:(?:[Ss]cientific [Ee]\s?-\s?[Rr]esources)|(?:Ed\s?-\s?[Tt]ech [Pp]ress)))|(?:\|url\s*\=\s*edtechpress\.co\.uk)"   added_lines irlike mirrors &   !(removed_lines irlike mirrors) )
 * I would create a log-only filter at first, and if it does well, ramp it up to warn. –  Pharyngeal Implosive7  (talk)  22:05, 13 July 2024 (UTC)
 * Thanks for this. I understand that starting as a long-only filter is common, and I've no objection. WhatamIdoing (talk) 00:20, 14 July 2024 (UTC)