Wikipedia:Bots/Noticeboard

Bots noticeboard

Here we coordinate and discuss Wikipedia issues related to bots and other programs interacting with the MediaWiki software. Bot operators are the main users of this noticeboard, but even if you are not one, your comments will be welcome. Just make sure you are aware about our bot policy and know where to post your issue. Do not post here if you came to discuss non-urgent bot issues, bugs and suggestions for improvement. Do that at the bot operator's talk page discuss urgent/major bot issues. Do that according to instructions at WP:BOTISSUE discuss general questions about the MediaWiki software and syntax. We have the village pump's technical section for that request approval for your new bot. Here is where you should do it request new functionality for bots. Share your ideas at the dedicated page

Fully automated edits without BRFA - Request for assistance

I recently came across a large number of automated edits (10,000+) made in October by NmWTfs85lXusaybq without a BRFA (as far as I am aware and can see). I disagree with the edits, for reasons including the ones I posted on their talk page. I don't mean to inflame the situation by posting here, but I feel that I'm somewhat out of my depths regarding knowing what the best thing to do here is, and that input from editor(s) more experienced in this area and/or the Bot Approvals Group would be beneficial, including with regards to the best next steps.

Let me know if there are any queries. All the best, ‍—‍a smart kitten_[meow] 13:11, 3 January 2024 (UTC)

Judging from Special:Contributions/NmWTfs85lXusaybq, these edits are ongoing. Seems like an easy case of WP:BOTBLOCK for running an unapproved bot. Headbomb {t · c · p · b} 20:40, 4 January 2024 (UTC)

Warning left on the initial thread. Primefac (talk) 20:45, 4 January 2024 (UTC)

Warning deleted, it looks like. –Novem Linguae (talk) 01:11, 5 January 2024 (UTC)

Per WP:OWNTALK. NmWTfs85lXusaybq (talk) 01:49, 5 January 2024 (UTC)

@NmWTfs85lXusaybq can you commit to not doing these edits without a BRFA? Galobtter (talk) 01:56, 5 January 2024 (UTC)

Sure. NmWTfs85lXusaybq (talk) 01:59, 5 January 2024 (UTC)

I'm cautious to suggest too much in this discussion, as there are other editors (who are much more experienced regarding bots than me) who may well have better ideas of what the next steps in this situation should be. However, I'd like to make an initial proposal that the edits I initially raised concerns about, defined as:

Edits (including pagemoves) made by NmWTfs85lXusaybq between 17 October 2023 and 24 October 2023 (inclusive) that are tagged with [paws 2.2] (OAuth CID: 4664)

be rolled back. This is due to the concerns I described on their talk page, and due to them being automated edits run without bot approval or consensus for the task. By my estimation, this is between 24,000-26,000 edits that would be reverted. (I would be happy to submit a BRFA to accomplish this on a bot account, using massRollback.js & a slightly modified massMoveRevert.js.)

I'd also ask if NmWTfs85lXusaybq would mind listing the automated tasks/bot runs that they have previously run on their account, so that they can be retrospectively assessed by the community & the Bot Approvals Group. If there are concerns raised with any of the other automated edits, further proposals (such as the above) can be made.

Everything I've just said, however, comes with the caveat that I am not as experienced in bot matters as other editors, so I will likely defer to members of the BAG if they have any other suggestions and/or take issue with any of what I've said.

All the best, ‍—‍a smart kitten_[meow] 16:56, 6 January 2024 (UTC)

I support the idea to roll these edits back (not a vote obvi I just really don't like this), though I'm not a bot-experienced editor, there's really no reason to do this and it makes everything far more confusing. PARAKANYAA (talk) 23:38, 6 January 2024 (UTC)

Mass rolling back did come to mind when I saw what was being done. I would support it too. Headbomb {t · c · p · b} 01:31, 7 January 2024 (UTC)

I support this but I would limit it to the edits from October 20-24; the October 17-19 edits are useful and not worth reverting IMO. * Pppery * _{it has begun...} 03:00, 8 January 2024 (UTC)

To expand on my view of why the Oct 17-19 edits are worth reverting; in my opinion, it’s unhelpful to mass-redirect talk pages of redirects to the talk pages of the target articles. Talk pages of redirects can be useful for discussing the redirects themselves, in addition to (for example) recording previous RfD discussion results. For this reason, I don’t support doing so generally as a blanket measure - without individual consideration as to whether the action is appropriate. (The exception to this is that I support redirecting talk pages in the case of an {{R from move}}, where someone that follows a link to the pre-move article talk page will rightly expect to end up at the current location of the article’s talk page). I also note that WP:TALKCENT says that an editor wishing to implement centralized talk pages should consider first gaining consensus for [the] proposal, which presumably would apply even more so to a mass-centralization such as this; especially when I can’t find a record that existing wider community consensus for such centralization exists. However - to be clear - if consensus is against me on the 17-19 Oct edits being reverted, it is not a problem, and I would still be happy to submit a BRFA to revert the others. All the best, ‍—‍a smart kitten_[meow] 14:18, 9 January 2024 (UTC)

I was quite aware of WP:TALKCENT when I did this task. I will never argue that the talk pages of redirects are useless. However, my point is a talk page can't be useful if it's intentionally blanked, no matter if it's associated with a redirect or an article (that's why I moved or tagged the empty talk pages of articles as well). Any empty talk page of redirects doesn't deserve a {{tpr}} tag per Template:Talk_page_of_redirect/doc#Misuse. Moreover, the task of Oct 17-19 edits is the most thanked one without any objection before this thread. NmWTfs85lXusaybq (talk) 14:55, 9 January 2024 (UTC)

I disagree with you on the principle - I think a talk page of a redirect should point to the talk page of its target unless there's a good reason otherwise, but regardless the October 17-19 edits replaced blank pages with redirects - I'm completely failing to see how re-blanking them could possibly be an improvement. * Pppery * _{it has begun...} 16:01, 9 January 2024 (UTC)

 Note: The abandoned talk pages are filtered by only one major edit and the messy talk pages are filtered by lack of section header from Category:Talk pages with comments before the first section. The ones created from anonymous user or inexperienced user (#edits <= 10) are generally against WP:TPG, like Talk:University of Hail. It's better to blank and redirect it to the talk page of its target, which is exactly what I have done, rather than tagging it with {{tpr}}. NmWTfs85lXusaybq (talk) 09:15, 8 January 2024 (UTC)

Notes

Other bot runs

I looked through NmWTfs85lXusaybq's other edits tagged as using PAWS. I found:

1. Early tests from September 15-October 7, 2023 that aren't actual bot runs
2. September 17: Mass tagging a bunch of talk pages (rightly) as U2 cases. They were later deleted.
3. October 11: Creating a bunch of "foo, the" -> "the foo" redirects. These seem harmless to me, but anyone who disagrees is welcome to R3 them.
4. October 12: Various changes to rcat templates on DAB pages. I'm not an expert in this area, but these seem correct to me.
5. October 14: Tagging redirects for a RfD that was later closed as no consensus
6. October 15: Adding {{R from unnecessary disambiguation}} to a bunch of redirects. Some of these (i.e United States Agricultural Information Network (USAIN) aren't really unnecessary disambiguation redirects) but this is overall harmless (and small enough it could have been done easily with AWB)
7. October 17-19: Centralized empty talk page of redirect. This specific bot run replaced all talk pages of redirects that were blank with redirects to their targets, and seems useful to me despite being included in the request to revert (although I agree it deserved a BRFA)
8. October 20: A bunch of page moves. A smart kitten and others make a reasonable case to revert these, although I'm not convinced it's necessary (there may be few enough to review manually)
9. October 22 #1: Some more page moves basically identical to the October 20 moves
10. October 22 #2: For some inexplicable reason Talk:2021 Polish census was created via PAWS
11. October 22-23: ‎ Centralized abandoned talk page of redirect from anonymous user. This is the meat of the complaint and I concur these edits (which go beyond the above link) need to be mass reverted.
12. October 24: ‎ Centralized messy talk page of redirect from inexperienced user. These are a (much smaller) extension of the previous run, and should also be reverted.
13. October 26: Creation of a few redirects to clean up after a page move. Innocuous.
14. October 27 #1: A re-run of the October 17-19 bot run except applying to pages outside of mainspace.
15. October 27 #2: Replace blank talk pages of redirects with soft redirects to Commons. Seems useful.
16. October 29-30: "‎unlink language label for transliteration template in disambiguation pages". Seems useful.
17. November 5-9: Adding {{Talk page of redirect}} to talk pages of redirects. Seems useful. This run repeats sporadically through the coming months (as recently as January 4), only affecting a few pages each time
18. November 11: A re-run of "‎unlink language label for transliteration template in disambiguation pages"
19. November 13: Add reference lists to a bunch of templates. Seems useful
20. November 24: Mass PRODing of disambiguation pages
21. November 26: Mass addition of {{One other topic}} to disambiguation pages
22. December 1: Mass redirection of disambiguation pages (and later set indices) with only one entry. Most of these are useful but this is the bot run that brought us oddities like Walker Elementary School (later deleted at RfD). I'm not sure what to do here, but this could use attention. Many of these edits have been deleted because they redirected a disambiguation page at X containing only X (Y) to X (Y) and then X (Y) was moved to X.
23. December 3: Moving of template documentation subpages. Innocuous
24. December 19-20: Mass changes to WikiProject banners. These seem to consist of removing {{WikiProject Disambiguation}} from talk pages of set indices, adding {{WikiProject Anthroponymy}} to talk pages of name pages, and adding {{WikiProject Lists}} to lists
25. December 21-22: Reclassifying set-indices and lists as list-class rather than disambiguation class in WikiProject banners, and removing the class parameter entirely for articles
26. January 3: Moving several dozen articles per a RM discussion

* Pppery * _{it has begun...} 02:56, 8 January 2024 (UTC)

So in summary, most edits look good, but you're proposing 11 and 12 be mass reverted? –Novem Linguae (talk) 10:00, 10 January 2024 (UTC)

I'm proposing 8,9, 11, and 12 be reverted. * Pppery * _{it has begun...} 14:32, 10 January 2024 (UTC)

The edits summarized by cleanup before move in 8 and 9 are actually different except that they are all intended to deal with the talk pages of redirect when that of the target doesn't exist. The case Special:Diff/1181304837 proposed by A smart kitten is in 9, not 8. The cases in 8 are generally uncontroversial, as the moved talk page is the only extant page among the talk pages of the target and all sources while most of them contain nothing but {{WPDAB}} and {{Tpr}}. Therefore, a similar argument may still apply: The revert of edits in 8 couldn't be an improvement unless there's a good reason to move the centralized talk page back without leaving a redirect. NmWTfs85lXusaybq (talk) 15:47, 10 January 2024 (UTC)

I was assessing a bunch of unassessed articles earlier and I found a bunch of cases where this editor automatically added WikiProject List to articles that weren't lists. It also was out of the template shell. Unsure how many of those there were but I noticed quite a few PARAKANYAA (talk) 16:49, 10 January 2024 (UTC)

The articles were only tagged with {{WikiProject List}} when they had been categorized as SIA while not tagged with any banner, or at least categorized in one list category, like "Lists of ...". NmWTfs85lXusaybq (talk) 23:50, 10 January 2024 (UTC)

Ah. Well that's someone's problem, but probably not yours then. PARAKANYAA (talk) 17:15, 11 January 2024 (UTC)

Unarchived, at WP:ANI

This was unarchived, but it is also being discussed at Wikipedia:Administrators' noticeboard/IncidentArchive1149#Disruptive editing of disambiguation pages. So as not to split the discussion, it is probably best to discuss it there for the time being. Rgrds. --BX (talk) 18:02, 15 February 2024 (UTC) Link updated post-archive. Primefac (talk) 17:23, 23 February 2024 (UTC)

These are not really that related. ANI is complaining about 22 in my list above. This discussion was originally complaining about 7-12. * Pppery * _{it has begun...} 18:09, 15 February 2024 (UTC)

Noting that the ANI thread has now been archived. Best, ‍—‍a smart kitten_[meow] 17:21, 23 February 2024 (UTC)

Proposal closure

I've thought about requesting formal closure of the proposal above, given that discussion seems to have died out. However, I'm conscious that there aren't that many comments in response to the proposal, which may make it difficult to infer a strong consensus (especially for >24,000 edits). I'd therefore appreciate it if more experienced editors than myself could comment on what might be the best next step here; including whether it's worth notifying any other venues (e.g. WP:AN) of this mass-rollback proposal, in order to hear more editors' opinions on the matter. All the best, ‍—‍a smart kitten_[meow] 17:21, 23 February 2024 (UTC)

Noting here for reference my post at WP:BOTREQ#Bot to mass-undo edits & pagemoves. All the best, ‍—‍a smart kitten_[meow] 18:45, 6 March 2024 (UTC)

I've asked for more input at VPP. Snowmanonahoe (talk · contribs · typos) 01:41, 2 April 2024 (UTC)

Well, that didn't work. Snowmanonahoe (talk · contribs · typos) 14:44, 15 April 2024 (UTC)

Blocked user-operated bots

I have an idea that may be unpopular, but I suggest that bots operated by blocked users should not just be blocked, but should instead be "adopted" to a new owner who will then refurbish it, rename it, and make it do new tasks. I see no reason to block innocent bots that have not done anything wrong because their owners have been blocked, instead it just slows down the project's productivity. 2003 LN6 19:30, 8 April 2024 (UTC)

We have plenty of bots, inactive or otherwise, who have had their tasks taken over by someone else. If the code is available there is no reason to take over the bot; just copy over the code. Primefac (talk) 19:40, 8 April 2024 (UTC)

It takes a lot of work to adopt (on Toolforge) or fork a bot. The idea of blocking first due to the bot operator losing trust, then other folks taking their time and adopting or forking the bot if the bot is important and if it is hosted in our ecosystem and/or has open source code, is already practiced. The idea of usurping an existing Wikipedia bot account of a blocked user seems unnecessary since a new account can be created to run the same program code as the original bot. –Novem Linguae (talk) 19:40, 8 April 2024 (UTC)

That is not a good idea because bots are just an extension of their operator, so the account should certainly not be just handed over. Of course the task can be replaced by anyone that wants to under a different bot account if it is still needed. — xaosflux ^Talk 10:29, 1 May 2024 (UTC)

Meatbot sisterlink updates

Well this may be pretty poor timing judging by the threads above on this page, but… :)

I'm about to run PWB's replace.py on my main account (from PAWS) over the ~267 articles in Category:Child Ballads to update a bunch of sisterlinks to Wikisource after a page reorganisation there. The changes will be variations on this edit; I'll be manually checking and approving each edit; and the rate will be as high as I can possibly make while still manually checking (i.e. not very, on average).

So far as I know this should require no particular approval (much less BRFA), but… 1) I haven't been paying attention to enWP for a couple of years (and I note WP:BOTPOL has been rather actively edited lately), so I could be out of touch. 2) It's likely I'll need to do similar cleanup on enWP, after wielding the mop over on enWS, in the future so I figure better safe than sorry.

So… thumbs up? No don't do it? RFC? BRFA? Xover (talk) 17:50, 18 April 2024 (UTC)

267 edits is a drop in the bucket, we wouldn't even blink if someone did that on AWB. Primefac (talk) 18:15, 18 April 2024 (UTC)

That's what I figured. Thanks. Xover (talk) 18:38, 18 April 2024 (UTC)

More specifically, the manual checking of each edit makes it fall under WP:SEMIAUTOMATED, and planning to be not very fast helps avoid trouble too. Use a good edit summary and it should be fine. Anomie⚔ 21:32, 18 April 2024 (UTC)

Inactive bots (May 2024)

Bot account	Operator(s)	Last edit (UTC)	Last operator activity (UTC)
Thehelpfulbot	Thehelpfulone	26 Dec 2012	01 Mar 2022
TheMagikBOT	TheMagikCow	21 Dec 2018	13 Apr 2022
PowerBOT	力	01 Aug 2019	30 Apr 2022

Per User:MajavahBot/Bot status report. * Pppery * _{it has begun...} 02:00, 1 May 2024 (UTC)

Notifications have been left. Primefac (talk) 08:47, 1 May 2024 (UTC)