User talk:Legobot/2023
This is an archive of past discussions with User:Legobot. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Possible Linter pattern for Legobot
If you are looking for Linter patterns for Legobot to process, please take a look at the "Centered images" note at User:MalnadachBot/Signature submissions. I don't know if it fits within the BRFA, but if it does, the two patterns listed there might be processable by the bot. – Jonesey95 (talk) 05:43, 24 March 2023 (UTC)
- Thanks for the suggestion! I think I got it working, but I'll wait for the current run to finish before turning it on. And I want to get the web interface operational to make it easier to test changes like this... Legoktm (talk) 00:56, 25 March 2023 (UTC)
- I'm not great at reading github pages, but that looks like a fix for gallery tags. The above link is just for regular File: invocations inside square brackets. – Jonesey95 (talk) 04:24, 25 March 2023 (UTC)
- Er, I read the wrong section. I see now what you were actually referencing, now trying it out... Legoktm (talk) 04:48, 25 March 2023 (UTC)
- OK, implemented. Legoktm (talk) 07:06, 25 March 2023 (UTC)
- Thanks. That should be another 5,000 errors that humans won't have to tediously perform. Progress! – Jonesey95 (talk) 12:53, 25 March 2023 (UTC)
- I'm not great at reading github pages, but that looks like a fix for gallery tags. The above link is just for regular File: invocations inside square brackets. – Jonesey95 (talk) 04:24, 25 March 2023 (UTC)
Legobot GIGO minor error
Legobot changed the invalid "font family" to the invalid "span family" instead of to span style="font-family.... See this Invalid font tags section for more GIGO to avoid. – Jonesey95 (talk) 18:21, 29 March 2023 (UTC)
Legobot tt tag opportunity
I'm not sure why Legobot missed this opportunity, but maybe a minor regex change could find more of them. – Jonesey95 (talk) 15:27, 30 March 2023 (UTC)
- It's because it's wrapped in two unbalanced templates (discussion top and bottom); Legobot can't edit those. Legoktm (talk) 02:06, 31 March 2023 (UTC)
Minor Legobot improvements
Two cases where Legobot could be improved.
Not sure why the bot decided to do this oddity and duplicate the color parameter, sticking one inside and one outside the link. It didn't create any errors, but it's a little GIGO-y.
Also, can you avoid doing this. It had font obsoletes, and the placement of them outside the link are a tidy font error (doesn't display link in the user's specified color). After the change to spans, both errors go away, but it still doesn't display as intended unless the color portion is inside the link brackets. If the bot had moved the color inside, it would have displayed as intended and been an excellent edit.
Thank you. Zinnober9 (talk) 04:39, 18 April 2023 (UTC)
RFC maintenance?
I recently posted an RFC on Talk:Elizabeth Holmes...Legobot seems to have just removed the RFC notice, but perhaps I don't exactly understand the process. Worth a double check? Thx, Bdushaw (talk) 07:17, 18 April 2023 (UTC)
- Ah, I had the RFC tag in the wrong place...would that the bot give a notice of that, rather than a quiet anonymous cleanup... Bdushaw (talk) 15:10, 18 April 2023 (UTC)
- @Bdushaw: People starting RfCs are normally expected to have understood WP:RFCST. Putting a contextless
{{rfc}}
tag into any section is not a good move, omitting a statement and timestamp makes it worse, and the fact that you used the lead section (which is for general informational banners) was just plain wrong. In this case Legobot didn't remove it for any of those reasons, but because the next timestamp that it encountered, i.e. 18:28, 11 January 2023 (UTC), was more than thirty days ago. --Redrose64 🌹 (talk) 17:11, 18 April 2023 (UTC)
- @Bdushaw: People starting RfCs are normally expected to have understood WP:RFCST. Putting a contextless
Index pages
I was wondering if there's a way to create a page 2 ? My main index page is getting quite long. Just wondering, no big deal I guess. Thanx. Cheers, - FlightTime (open channel) 19:45, 21 April 2023 (UTC)
Old WP:MFD
@Legoktm any reason why your bot is not archiving closed WP:MFD discussions and not moving 22nd April discussions to old business? Lightoil (talk) 08:40, 30 April 2023 (UTC)
- @Lightoil: not on purpose. Seems fine now? Is there still an issue? Legoktm (talk) 04:32, 2 May 2023 (UTC)
- @Legoktm no the issue seems to be resolved with discussions being archived by your bot. Lightoil (talk) 04:35, 2 May 2023 (UTC)
- This bot appears to have gone down again. * Pppery * it has begun... 17:38, 19 May 2023 (UTC)
- Ack, no, that's not the problem 0 it's adding entries to Wikipedia:Miscellany for deletion/Archived debates/April 2023 but not removing them from the base page. * Pppery * it has begun... 17:40, 19 May 2023 (UTC)
- @Legoktm: * Pppery * it has begun... 21:42, 23 May 2023 (UTC)
- Fixed. Basically the bot stops when it can't find the correct date header, which can leave it in a weird state of some things archived but not everything removed. Presumably it should just add in missing date headers instead of stopping but I'd need to think through the edge cases a bit more before implementing that. Legoktm (talk) 13:59, 24 May 2023 (UTC)
- Oops. Will note that for the future the next time I do manual archiving (which I did in that case to get the Aoril Fools jokes off the page as soon as it was April 2). * Pppery * it has begun... 14:44, 24 May 2023 (UTC)
- Fixed. Basically the bot stops when it can't find the correct date header, which can leave it in a weird state of some things archived but not everything removed. Presumably it should just add in missing date headers instead of stopping but I'd need to think through the edge cases a bit more before implementing that. Legoktm (talk) 13:59, 24 May 2023 (UTC)
- @Legoktm: * Pppery * it has begun... 21:42, 23 May 2023 (UTC)
- Ack, no, that's not the problem 0 it's adding entries to Wikipedia:Miscellany for deletion/Archived debates/April 2023 but not removing them from the base page. * Pppery * it has begun... 17:40, 19 May 2023 (UTC)
Nomination of List of Asian superheroes for deletion
The article will be discussed at Wikipedia:Articles for deletion/List of Asian superheroes until a consensus is reached, and anyone, including you, is welcome to contribute to the discussion. The nomination will explain the policies and guidelines which are of concern. The discussion focuses on high-quality evidence and our policies and guidelines.
Users may edit the article during the discussion, including to improve the article to address concerns raised in the discussion. However, do not remove the article-for-deletion notice from the top of the article until the discussion has finished.
Dronebogus (talk) 09:31, 17 June 2023 (UTC)
Legobot does not support diacritics
The entry for Aleksandar Vučić displays č and ć as question marks. Any possibility to make it display characters properly when transcluding RfCs? Szmenderowiecki (talk) 22:02, 19 June 2023 (UTC)
- @Szmenderowiecki: This is a known problem, see the archives of this page and of User talk:Legoktm. It is why the redirect Talk:Aleksandar Vu?i? was created, and also why I pushed for "keep" at Wikipedia:Redirects for discussion/Log/2023 June 8#Redirects for Legobot no longer linked to. Legoktm has been saying for some years now that they want somebody else to take over all of Legobot's current tasks, and isn't fixing any bugs. In short: if you know how to fix Legobot, please do so. --Redrose64 🌹 (talk) 17:59, 20 June 2023 (UTC)
- I started poking around a bit. Relevant bot task is here. Source code is here. The regexes are lacking a /u flag, so that might be one idea to try. Might also be worth checking the $rfadb settings in SQL. But honestly this code is complex enough that setting up a test case on testwiki or localhost and then step debugging it might be the way to go. –Novem Linguae (talk) 19:32, 20 June 2023 (UTC)
- Sorry, I can't really code, though I do hope to learn at least one language. For now it's basic HTML for me and that's it. Szmenderowiecki (talk) 19:52, 20 June 2023 (UTC)
Proposal: add anchor with Rfc id when removing Rfc header
When Legobot removes an Rfc header, let's add {{anchor|Rfcid}}
at the top of the expired or withdrawn Rfc to facilitate navigation to the Rfc discussion from links left elsewhere while it was still live, notably Talk page notifications left by Yapperbot.
When following a link on my Talk page from Yapperbot's randomly assigned Rfc's, I not infrequently end up at the top of some Talk page, and I'm left staring at a hatnote stating:
- No search results found for section "rfc <hexcode>" in archives. It may have been removed or renamed, or you may have followed a malformed link.
This may be because I waited too long and it's been closed, or because it's been withdrawn. Whatever the reason, that doesn't mean I'm uninterested in seeing how it came out, and sometimes Rfc's are converted into discussions which I want to take part in, and not being able to navigate directly to the discussion from my Talk page is an annoyance. This is usually just a minor time-waster, as I hunt for a ToC item with "Rfc" in the title, somewhere near the bottom of the page. But sometimes they don't have that word, and sometimes there's more than one Rfc, and I end up at the wrong one and things can go awry, which is what happened to me today (see discussion).
Considering any possible negative consequences of implementing this proposal, it might appear to a newer user unfamiliar with the concept of an Rfc header, as if a withdrawn Rfc were still live, and they might continue to vote or spend time on it needlessly. To mitigate this, Legobot should add a hatnote at the top of any Rfc it withdraws, which will be just as useful to veteran users following Yapperbot links imho, as it will be to newbies.
Sometimes a user does this manually, when removal of the Rfc header happens by one of the editors in the discussion. For example, this diff shows removal of a header and addition of a hatnote by Mandruss, which is very helpful, and also the first thing visitors to the Rfc will see. (As a follow-up, o/t to Legobot but helpful to users: to make it even easier and to support helpful editors like this, it would be quite easy to create a template, where one might code, {{rfc demote|rfcid=CC0FC1C}}
which would add the hatnote, and also embed an anchor with the given rfcid at the same time, so that in-links would continue to work. That could be done independently if there's an appetite for it, regardless whether the Legobot enhancement proposal is accepted or not.) Thanks, Mathglot (talk) 00:14, 15 August 2023 (UTC)
- This is a decent idea in theory, but it's unlikely Legoktm or anyone else watching this page will have the will to actually do it. Code contributions welcome. * Pppery * it has begun... 00:21, 15 August 2023 (UTC)
- I don't fully understand. Your problem[1] appears to have occurred more than one hour after the RfC listings were removed by Legobot.[2] What did you click to get there? ―Mandruss ☎ 00:40, 15 August 2023 (UTC)
- The problem *only* occurs after the Rfc header is removed. I clicked the Yapperbot notification on my talk page, here. The proposal is a solution to the problem of linking to a discussion that was formerly an Rfc anchored by its hex
rfcid
, but which no longer has an Rfc header, so the hex "id" which formed the basis for the inlink is gone, so when you click the Rfc link in the Yapperbot notification, it has no resolution, therefore it can only go to the top of the page. Mathglot (talk) 03:49, 15 August 2023 (UTC)- I see. I never encountered the problem because I never subscribed to those notifications. If you're the first editor in decades to see a problem that needs fixing, how bad can the problem be? As in anything like this, we should weigh the benefit against the cost of the complexity added to the ecosystem. The latter is too often overlooked, which is why the ecosystem is far too complex.And then there's the problem that the necessary support is simply not there, making the whole thing fairly moot. It's a known and apparently accepted issue with Legobot. Some things around here are just frozen because no qualified editor cares to take over when the original developer loses interest. ―Mandruss ☎ 04:09, 15 August 2023 (UTC)
- I much prefer the "glass half full" approach. And as far as "If you're the first editor in decades to see a problem that needs fixing, how bad can the problem be?", 1) you have no way of knowing that; perhaps everybody sees it; and 2) that is the universal Customer Service get-out-of-jail-free card, and implies that because the CS rep hasn't heard it before, therefore the customer problem either doesn't exist, or isn't serious enough to worry about. Every bug listed in Phab was first entered by *someone*; WP:VPT archives are full of bugs and WP:VPR full of proposals first raised by someone; sometimes they generate lively debate, sometimes they don't; no way to know, until it's out there. Finally, as far as "no qualified editor", I'm sticking with my glass half full. I totally get it why this would seem like 999th priority for someone who doesn't subscribe to Rfc notifications; everybody feels a different part of the elephant, and our experiences are all unique to a certain extent; that's your experience, and I respect that, and this is mine. I'd love to see this take shape, and if some volunteer editor finds it interesting and feels likewise, maybe it will, or maybe it won't; we'll see. Your comment about complexity is afaict the best reason not to enhance it, but of course, one might say that about any proposal. But this all feels very meta, and I'm eager to see if it goes anywhere. Meanwhile, I've been working on something concrete: namely, a template inspired by your helpful hatnote at that Rfc, that I'd like your feedback about. Stay tuned... Mathglot (talk) 05:21, 15 August 2023 (UTC)
- I'd like to mention three things regarding this suggestion. First, not all
{{rfc}}
tags are removed by Legobot - quite a number are removed by humans following the directions at WP:RFCEND, and if we add a new direction there, it may be overlooked by those who have used RFCEND sufficiently often to not need to read it through every time. Second, Legoktm has stated in the past that they are no longer interested in maintaining the RfC side of Legobot, and is hoping that a willing volunteer will take over that bot task with its associated unfulfilled requests for enhancement. Third, the anchor does not involve the letters rfcid - the actual anchor would be e.g.{{anchor|rfc_CC0FC1C}}
. --Redrose64 🌹 (talk) 08:07, 15 August 2023 (UTC)- Thank you for your comments. The manual-removal case offers a good seguë to the template I referred to above: it's at {{rfc demote}}. Thanks for the reminder about the rfc_ prefix; the template handles that now. I notice that the code base is in php, which might make it more amenable to being taken over by someone, as php is pretty easy going. Mathglot (talk) 08:53, 15 August 2023 (UTC)
- I'd like to mention three things regarding this suggestion. First, not all
- I much prefer the "glass half full" approach. And as far as "If you're the first editor in decades to see a problem that needs fixing, how bad can the problem be?", 1) you have no way of knowing that; perhaps everybody sees it; and 2) that is the universal Customer Service get-out-of-jail-free card, and implies that because the CS rep hasn't heard it before, therefore the customer problem either doesn't exist, or isn't serious enough to worry about. Every bug listed in Phab was first entered by *someone*; WP:VPT archives are full of bugs and WP:VPR full of proposals first raised by someone; sometimes they generate lively debate, sometimes they don't; no way to know, until it's out there. Finally, as far as "no qualified editor", I'm sticking with my glass half full. I totally get it why this would seem like 999th priority for someone who doesn't subscribe to Rfc notifications; everybody feels a different part of the elephant, and our experiences are all unique to a certain extent; that's your experience, and I respect that, and this is mine. I'd love to see this take shape, and if some volunteer editor finds it interesting and feels likewise, maybe it will, or maybe it won't; we'll see. Your comment about complexity is afaict the best reason not to enhance it, but of course, one might say that about any proposal. But this all feels very meta, and I'm eager to see if it goes anywhere. Meanwhile, I've been working on something concrete: namely, a template inspired by your helpful hatnote at that Rfc, that I'd like your feedback about. Stay tuned... Mathglot (talk) 05:21, 15 August 2023 (UTC)
- I see. I never encountered the problem because I never subscribed to those notifications. If you're the first editor in decades to see a problem that needs fixing, how bad can the problem be? As in anything like this, we should weigh the benefit against the cost of the complexity added to the ecosystem. The latter is too often overlooked, which is why the ecosystem is far too complex.And then there's the problem that the necessary support is simply not there, making the whole thing fairly moot. It's a known and apparently accepted issue with Legobot. Some things around here are just frozen because no qualified editor cares to take over when the original developer loses interest. ―Mandruss ☎ 04:09, 15 August 2023 (UTC)
- The problem *only* occurs after the Rfc header is removed. I clicked the Yapperbot notification on my talk page, here. The proposal is a solution to the problem of linking to a discussion that was formerly an Rfc anchored by its hex
- I'd be happy to have this happen, or to have a different state for the original RFC template (like
{{rfc|rfc_240920|ended}}
). This might also make it easier to identify, in the future, discussions whose section headings claim they are RFCs but that were never in the system. WhatamIdoing (talk) 05:12, 19 August 2023 (UTC)- Unfortunately Legobot will choke on that; it searches Wikitext for the five-character case-insensitive string
{{rfc
and if found, assumes that it marks the start of an open RfC. Moreover, it also expects those characters to be followed by zero or more valid WP:RFCCAT parameters, then zero or one|rfcid=xxxxxxx
parameter, then the two characters}}
. There was an attempt a few years ago to add another parameter to the{{rfc}}
template, with unfortunate results. --Redrose64 🌹 (talk) 17:13, 19 August 2023 (UTC)
- Unfortunately Legobot will choke on that; it searches Wikitext for the five-character case-insensitive string
Bot error at MfD
[3] should be self-explanatory. * Pppery * it has begun... 04:42, 6 September 2023 (UTC)
Wikipedia:Requests for comment/Wikipedia policies and guidelines
I edited Wikipedia:Requests for comment/Wikipedia policies and guidelines inserting {{bots|deny=Legobot}}
because you keep messing up <small>...</small>
tags. Go ahead and remove the bots template, but please stop messing up the <small>...</small>
tags. (Also attn: WOSlinker) —Anomalocaris (talk) 07:13, 10 December 2023 (UTC)
- (talk page watcher) @Anomalocaris, WOSlinker, and Zinnober9: When editing WP:RFC/POLICY, did you not see the message
- This list is updated by Legobot; your edits will be overwritten if you edit this page.
- The proper thing to do is to fix the problem at its source, which in this case must be in Wikipedia talk:Biographies of living persons, because that is where Legobot is copying text from. See WP:RFC#Statement should be neutral and brief: this says
Legobot will copy the markup of your statement (from the end of the
; earlier on, note [1] says{{rfc}}
tag through the first timestamp) to the list of active RfCs, if it is sufficiently brief; a long statement will fail to be copied. For technical reasons, statements may not contain ... complex formatting, although these may be added after the initial statement (i.e., after the first timestamp).The "statement" is the part that is located between the
{{rfc}}
tag (exclusive) and the first valid timestamp (inclusive), and which is copied by bot to various pages. - This means that not only was Legobot working as designed, it is a bad idea to follow the
{{rfc}}
tag with notes that are not part of the RfC statement itself, expecially if a template like{{block indent}}
and a HTML tag like<small>
are both still open when the next timestamp occurs. Such notes must always be after the timestamp that closes the statement. The problem was introduced by Novo Tape (talk · contribs) with this edit, and the fix is very simple. --Redrose64 🌹 (talk) 10:52, 10 December 2023 (UTC)- Thank you for the explanation. I will admit that while I'm well versed in many other syntax errors, I am not well versed in RFCs, so was unaware of this limitation with their notation. I was not aware of the offending statement being in the wrong sequential location, so thought instead that Legobot was adding a line break within the rfc statement (line breaks surprisingly break a lot of tags). While I did see the "Legobot will overwrite" message, I would have thought that the nobot tag would have had the higher power, and only added it/readded it so that we could have a little timeout from the bot reinstating the issue while we fully figured out what was going on. This issue reminds me again that I really hate transcluded errors since they act all fine and innocent on the original page, but cry bloody hell on other pages. Sorry for the error in my understanding, and thank you for enlightening us on this issue and for the simple solution. Zinnober9 (talk) 18:03, 10 December 2023 (UTC)
- @Zinnober9: Where it says "your edits will be overwritten if you edit this page", it really means it. The whole page is built by Legobot, based upon a set of data tables held off-wiki plus the current text of any page that transcludes
{{rfc}}
. Essentially, Legobot runs once an hour and builds fresh copies of each and every RfC listing page, which it saves, regardless of whatever happens to be on the page at the time. It therefore ignores{{nobots}}
and any other content that it didn't add itself on its previous run. Whilst this may seem to be a violation of WP:OWN, it's quite normal for bot-built reports, of which there are many. --Redrose64 🌹 (talk) 23:04, 10 December 2023 (UTC)
- @Zinnober9: Where it says "your edits will be overwritten if you edit this page", it really means it. The whole page is built by Legobot, based upon a set of data tables held off-wiki plus the current text of any page that transcludes
- Thank you for the explanation. I will admit that while I'm well versed in many other syntax errors, I am not well versed in RFCs, so was unaware of this limitation with their notation. I was not aware of the offending statement being in the wrong sequential location, so thought instead that Legobot was adding a line break within the rfc statement (line breaks surprisingly break a lot of tags). While I did see the "Legobot will overwrite" message, I would have thought that the nobot tag would have had the higher power, and only added it/readded it so that we could have a little timeout from the bot reinstating the issue while we fully figured out what was going on. This issue reminds me again that I really hate transcluded errors since they act all fine and innocent on the original page, but cry bloody hell on other pages. Sorry for the error in my understanding, and thank you for enlightening us on this issue and for the simple solution. Zinnober9 (talk) 18:03, 10 December 2023 (UTC)
Bot wars
@Legoktm: Your bot's been reverting itself twice daily since January 2023, and before that daily since December 2021, on Talk:Crusades/Archive index (history). Similarly, it's been reverting itself daily since November 2021 and more sporadically before that since March 2017 on User talk:Safety Cap/Archives/Destination (history). —Cryptic 17:57, 26 October 2023 (UTC)