Jump to content

Wikipedia:Village pump (technical)

From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by Sam at Megaputer (talk | contribs) at 18:27, 11 April 2021 (→Get a revid: re). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

 Policy Technical Proposals Idea lab WMF Miscellaneous 
Shortcuts
The technical section of the village pump is used to discuss technical issues about Wikipedia. Bug reports and feature requests should be made in Phabricator (see how to report a bug). Bugs with security implications should be reported differently (see how to report security bugs).

Newcomers to the technical village pump are encouraged to read these guidelines prior to posting here. If you want to report a JavaScript error, please follow this guideline. Questions about MediaWiki in general should be posted at the MediaWiki support desk. Discussions are automatically archived after remaining inactive for five days.


Proposal for finding longstanding bad articles

Thomas Ranch is about to be deleted, having existed in Wikipedia for over fifteen years without a single inline reference, and without ever having an external link to an independent source. Grub Smith, though a bit longer, has been in a similar state for an equally long time. I happened to come across these two not while searching for suspect articles but while doing general cleanup around the given name Thomas and the surname Smith. How many more articles are out there in this condition? I propose that the best way to find out, if technically feasible, would be to generate a list of articles that have never had an inline ref tag over the course of their existence, sorted by age, and push through them from the oldest on forward. If someone has the Wiki-fu to generate such a list, please have at it. BD2412 T 00:20, 5 April 2021 (UTC)[reply]

Never over the course of their existence is probably a tall order, but WP:RAQ is probably a better first stop. Izno (talk) 01:51, 5 April 2021 (UTC)[reply]
Yes, a tall order, but great efforts yield great rewards. Thanks for the pointer. BD2412 T 02:41, 5 April 2021 (UTC)[reply]
  • BD2412, I seem to remember that the database replicas only include metadata, not the actual page contents. To get the contents, you need to go through the API (which is much slower). Somebody needs to check me on that. Just as a proof-of-concept, I wrote a trivial little python script that iterates over all pages in mainspace and searches for <ref> anywhere in the text. It's processing about 10 pages/second. We've got about 6 million articles (from WP:STATS, which I assume is talking about mainspace when it says, "6,281,819 articles"). So, we could scan every mainspace article in about a week. One could envision doing that. I'm guessing the number of revisions is 2 orders of magnitude higher, so searching every revision of every article would likely be prohibitive. First guess, a couple of years. I'm not sure what value it would add anyway. -- RoySmith (talk) 02:49, 5 April 2021 (UTC)[reply]
    We could cut that down significantly by first removing from the search every article that has a ref tag right now (which should be a substantial majority), and of those that remain, only searching articles created before, say, 2007. The value would just be clearing out old garbage. BD2412 T 03:00, 5 April 2021 (UTC)[reply]
An database scan should work. I tried searching for an ref tag on a lot smaller wiki than enwiki and it seems to work. Queries would not work, because they are not logged in an sql table.--Snaevar (talk) 08:00, 5 April 2021 (UTC)[reply]
Snaevar, What query did you perform? Can you link to a Quarry page? -- RoySmith (talk) 14:02, 5 April 2021 (UTC)[reply]
They're likely talking about database dumps. Quarry won't work as you say – it doesn't have page texts. Pinging HaeB who I think can comment on the feasibility of processing the dumps for this task. – SD0001 (talk) 16:55, 5 April 2021 (UTC)[reply]
I did an database scan using AWB, on an small wiki and I simply searched the main namespace for pages who do not have <ref>. That kind of search is done offline. The dump file contained all pages with the most current version only (pages-meta-current in the filename). The dump file was downloaded from dumps.wikimedia.org.--Snaevar (talk) 19:33, 5 April 2021 (UTC)[reply]
As noted in the WP:RAQ discussion, inquiries can further be narrowed to articles categorized in the Category:Companies category tree, or articles categorized in Category:Living people. That's where these sorts of problems appear most likely to arise. BD2412 T 15:15, 5 April 2021 (UTC)[reply]
@BD2412: why not begin with something like Category:Articles lacking sources from December 2006? For companies specifically use something like Cleanup listing for WikiProject Business (currently does not exist for WikiProject Companies but you can request one). – Finnusertop (talk ⋅ contribs) 16:17, 5 April 2021 (UTC)[reply]
Neither of the articles noted above were so tagged, perhaps because they have external links (even though these link to sites unusable as sources). I am looking for things that have really slipped through the cracks. BD2412 T 16:20, 5 April 2021 (UTC)[reply]
Sure, BD2412, but if some type of action is needed on articles with these specific problems, it would make sense to start with the ones that have already been identified. But I get your point. I sometimes run across low-traffic articles with obvious problems that have slipped through the cracks as well and tag them. – Finnusertop (talk ⋅ contribs) 16:24, 5 April 2021 (UTC)[reply]

Does this seem plausible? -- RoySmith (talk) 02:27, 7 April 2021 (UTC)[reply]

BTW, https://github.com/roysmith/bad-articles -- RoySmith (talk) 02:31, 7 April 2021 (UTC)[reply]
Thanks. Somewhat remarkably, up until now none of those article were tagged as needing sources. I suspect that there are more out there, though. BD2412 T 04:04, 8 April 2021 (UTC)[reply]
@RoySmith scan 994,949 pages in just under 6 minutes using API or dumps? Either way, wow!
The latter sounds implausible though as there's a list of unreferenced BLPs at User:SDZeroBot/Unreferenced_BLPs - it's a lot more than 7. Or did you mean pages that already don't have an unsourced tag? – SD0001 (talk) 20:27, 8 April 2021 (UTC)[reply]
SD0001, Through the API. This is running inside a WMF data center, so I assume it's got a lot more bandwidth to the API servers than you or I would from a remote machine. Yeah, it seemed amazing, which is why I was looking for confirmation that it made sense. Clearly something isn't right, I'll dig more. -- RoySmith (talk) 20:55, 8 April 2021 (UTC)[reply]
Oh, my. Apparently I flunked Programming 101, as well as playing hookey during both semesters of Software Engineering 101 where they introduced the concept of testing one's code before publishing it, compounded by writing code past one's bedtime and further exacerbated by failing to perform even the most rudimentary sanity checks.
I've got a new version now. It should process articles at about 0.01% of the speed of the original, but with the offsetting advantage of actually doing something useful. It finds, for example Odalys Adams, which makes up for the lack of references by having 22 categories. It was on track to finish in about 24 hours, but I killed it to avoid beating up on the servers too much. I'm working on getting access to the dump files and I'll switch to using those once that's done. -- RoySmith (talk) 21:26, 8 April 2021 (UTC)[reply]
😂. Fetching a million pages in minutes did sound suspicious! As for dumps, I see they're there on toolforge at /mnt/nfs/dumps-labstore1006.wikimedia.org/enwiki/latest. – SD0001 (talk) 13:24, 9 April 2021 (UTC)[reply]

Has an enwiki search with -insource:"<ref" been mentioned? The results include XXX (a disambiguation page) and Hitler Youth (has heaps of references using {{sfn}} but no <ref> tags). To limit the search to a category, use incategory:"Living people" -insource:"<ref" Johnuniq (talk) 00:33, 9 April 2021 (UTC)[reply]

Timeout at 200k but I skimmed some and most did not have inline ref on the page in question. Izno (talk) 00:47, 9 April 2021 (UTC)[reply]
Johnuniq, Can you actually search for "<ref", with the leading punctuation? I thought all punctuation was ignored by the search indexer. -- RoySmith (talk) 02:17, 9 April 2021 (UTC)[reply]
You are correct—from Help:Searching#insource:, "non-alphanumeric characters are ignored". I'm a newbie at searching and started with insource:/regexp/ (slashes instead of quotes) and (I believe) that does allow punctuation following regex rules. However the regexp timed out so I switched to quotes without much thought. At any rate, my point was that dab pages need to be ignored (which happens automatically if using Category:Living people) and that some found pages may have no ref tags yet still be good because they use one of the new-fangled referencing methods. Johnuniq (talk) 03:06, 9 April 2021 (UTC)[reply]

Fool-proofing WP:Sandbox notice

Is there anyway of fool-proofing the notice so it won’t ever get deleted again? Perhaps make it smaller? More hidden comments? Move it to the page’s edit notice? --Heymid (contribs) 22:18, 5 April 2021 (UTC)[reply]

Cyberbot I has a task that cleans the sandbox periodically, like this. I didn't dig in to its BRFA or logic to see why it sometimes takes a few hours between edits. cyberpower678 might be able to shed some light. – Jonesey95 (talk) 22:57, 5 April 2021 (UTC)[reply]
I know this. I haven't been living under a rock. So there’s no way of forcing the template to be there outside of putting it in the page’s wikitext along with everything else? A problem is that some users add the nobots template which disables bot editing of the page. And it’s not like the bot immediately restores the message once removed. --Heymid (contribs) 19:47, 6 April 2021 (UTC)[reply]
Cyberbot I's BRFA for the task in question says it isn't exclusion compliant, so {{nobots}} shouldn't be an issue. – Rummskartoffel (talk • contribs) 15:50, 8 April 2021 (UTC)[reply]

List of articles that need improvement

I've come across a cluster of Wiki articles that are loaded with unsourced statements. What is the best way to get them on a list of articles that need improvement that everyone can see? I've been adding the Ref Improve template to the top of such articles. Does that automatically put it on some "articles needing improvement" list?

In some cases, the entire article is unsourced. Some contain line after line of equations in math-major notation, with no supporting footnotes. The average reader has no way to ascertain whether that "stuff" is supported by experts in the field, generally accepted, or some anonymous Wikipedia editor's original work. I've been trying to find reliable sources, but it's difficult because that requires access to specialized textbooks that aren't readily available online or in my local library. The articles I'm talking about concern finance. For example:

Thanks, BuzzWeiser196 (talk) 10:54, 6 April 2021 (UTC)[reply]

@BuzzWeiser196: {{Refimprove|date=April 2021}} adds the hidden categories Category:Articles needing additional references from April 2021 and Category:All articles needing additional references. The latter has 400,000 articles. You can enable "Show hidden categories" at Special:Preferences#mw-prefsection-rendering. I recommend that for editors. PrimeHunter (talk) 11:08, 6 April 2021 (UTC)[reply]
Thank you. BuzzWeiser196 (talk) 11:12, 6 April 2021 (UTC)[reply]

Editing a talk page blanks the entire discussion

I just added a comment at the last thread Talk:macOS High Sierra, and the edit window blanked all the existing text. Consequently, this diff has removed a bunch of comments (when I didn't want it to!) and ended up with a reply to the wrong thread. I know talk page communication is difficult, but if I'm struggling, what hope have new users got? Ritchie333 (talk) (cont) 14:58, 6 April 2021 (UTC)[reply]

Same issue here, I was only able to get to edit here by doing it in mw:safemode. It doesn't appear to affect me while logged out (I popped into an incognito window and appeared to be able to edit fine). ‑‑ElHef (Meep?) 15:20, 6 April 2021 (UTC)[reply]
If I comment out all of my extensions in Special:MyPage/common.js, it starts working. I'll go back and uncomment each one in turn, and let you know which is the culprit. Ritchie333 (talk) (cont) 15:22, 6 April 2021 (UTC)[reply]
That's probably not what's going on with mine; my common.js is blank (has been for years). ‑‑ElHef (Meep?) 15:29, 6 April 2021 (UTC)[reply]
Well, if I comment out importScript('User:Gary/comments in local time.js');, the problem goes away. Ritchie333 (talk) (cont) 15:30, 6 April 2021 (UTC)[reply]

Apologies if this is the wrong place, but I don't know where else to ask
I regularly search for links in articles to User pages, as these are contrary to MOS:DRAFTNOLINK and are often vandalism - usually linking to unacceptable content - or people mistakenly signing their additions to articles
To find these I search for insource:User insource:/\[\[ *User *:/ sorted by edit date - as here
Today, User:polbot has auto-created 150-160 articles, each of which includes User:polbot and this; along with User:GreenC bot/Job 18 and User:Kvng/RTH are cluttering up the search results, making it difficult to find the links to "real" users pages. However, I don't know enough about the coding to omit those three "users" from the search results - can anyone help me out ? Thanks - Arjayay (talk) 15:14, 6 April 2021 (UTC)[reply]

@Arjayay This search works. Not sure why you had wikipedia namespace selected in there which was causing a timeout. – SD0001 (talk) 15:19, 6 April 2021 (UTC)[reply]
Thanks SD0001 - that's really helpful, and very prompt - I was trying to put the omissions at the end - Thanks again - Arjayay (talk) 15:24, 6 April 2021 (UTC)[reply]
@Arjayay: This omits pages with a user link inside comment tags. PrimeHunter (talk) 16:04, 6 April 2021 (UTC)[reply]
Wow PrimeHunter thanks a lot. I started off with 3250 search results this morning, SD0001 reduced that to 750, and you've got it down to 40 - My only regret is not asking sooner - thanks again to both of you - Arjayay (talk) 17:19, 6 April 2021 (UTC)[reply]
We also have Wikipedia:Database reports/Articles containing links to the user space but it's not live and it includes links made by templates, e.g. {{Proposed deletion/dated|...|nom = username}}, so it's less useful. It would be better if our search feature had a linksfrom: option (requested in phab:T253642). PrimeHunter (talk) 17:51, 6 April 2021 (UTC)[reply]

Transcluding mainspace pages

I've discovered that some pages like Doctor Who (film) are being transcluded in other mainspace pages. I'm not sure I agree with this practice. This is a job that I think should be done only with templates. Most editors, especially if you're using VisualEditor, probably don't realize that editing the table at Doctor Who (film)#Production effects over 340 other articles. With that in mind, I have semi-protected the page indefinitely as high-risk, which is something that happens automatically in the templatespace.

What's even worse I think is using parser functions like #section and #section-h. This kind of selective transclusion offers useful functionality but it is fragile. #section requires special markup on the target page, and #section-h will break all transclusions if an editor changes or removes the target section heading.

Is there any technical reason not to use templates for these use-cases? If not, should we propose a policy/guideline to disallow mainspace transclusion? The practice seems to be widespread (~3,500 articles have a transclusion count of 5 or more). It would take a lot of work to convert everything to use templates, but we could at least deprecate and/or formally discourage this practice.

Thoughts? — MusikAnimal talk 15:57, 6 April 2021 (UTC)[reply]

New plan for RADAR rollout

Since no one wants to be a lab rat, we are changing course for the rollout of my new tool. Registration is no longer be required for use.

  • Click here to use the tool. (link will begin working shortly) Sorry for the bad link. The actual link will be up next Tuesday.
  • Click here for instructional videos.

Be sure to save the links so you don't lose them. Sam at Megaputer (talk) 16:32, 6 April 2021 (UTC)[reply]

So what exactly is this tool? EGGIDICAE🥚 18:50, 6 April 2021 (UTC)[reply]
The tool does several things, but the main feature I am proud of it that it detects promotional articles. Or at least it will, once our IT guy get the link working (haha). So I've run a customized sentiment analysis on all articles about companies and schools to derive a promo score, which we can use to rank them from most to least promotional. I've also pulled in all of the promo tags, and monthly page views. Then the idea is that we apply various sorts and filters to discover various types of damaged articled. To give some examples, we can filter down to untagged articles and sort by promo score. The articles at the top of the list are promotional articles without tags, so we tag them. Or we can filter to articles that have POV tags, and then sort by monthly views. If we care about reducing damaged hits, this gives us a priority que for articles in need of cleanup. Sam at Megaputer (talk) 18:59, 6 April 2021 (UTC)[reply]
I don't understand why we need a tool hosted off Wiki which presumably will have access to private data for user accounts when we have perfectly good filters that already do this. EGGIDICAE🥚 19:05, 6 April 2021 (UTC)[reply]
Well, we can't sort by promo score. And I'm not sure what you mean by access to private data for user accounts. Once the link is working, no account setup will be required. Sam at Megaputer (talk) 19:09, 6 April 2021 (UTC)[reply]
Yes, we can. In the same way that ORES and filters already do. EGGIDICAE🥚 19:12, 6 April 2021 (UTC)[reply]
I'm not sure that ORES does exactly that. Could you show me what you mean? For example, could you get me a link to all articles on companies that have a notability tag, sorted from most to least promotional? Sam at Megaputer (talk) 19:20, 6 April 2021 (UTC)[reply]
MW:ORES. Can you please explain how exactly your tool works, what type of information it would "see" from using it? EGGIDICAE🥚 19:28, 6 April 2021 (UTC)[reply]
I make this video series to explain how it works from a user's prospective. If this can be achieved with ORES, It's certainly not obvious how to do it from that documentation. I am planning on opening up the tool for inspection after it has been out for two weeks and we have collected some data on efficacy. Sam at Megaputer (talk) 19:40, 6 April 2021 (UTC)[reply]
You still haven't explained what if anything it sees from a data standpoint, which is more important since you're asking people to test it out and promoting it all over the site without any meaningful discussion with the actual community. EGGIDICAE🥚 19:42, 6 April 2021 (UTC)[reply]
Are you asking me to explain how it creates the ranking? Sam at Megaputer (talk) 19:47, 6 April 2021 (UTC)[reply]
To be honest, this sounds to me like straight-up spam promoting your company's tool, and I'm half-minded to remove the spam you posted - the optics of this are very concerning for me. If it's genuinely not intended to be promotion and is intended to purely for the interests of furthering the project, why not open-source it and host it on wikitech:Toolforge? That way, concerns about leaking the private data of users of this tool can be mitigated through the requirement to follow Toolforge's terms of use as well. While I'm definitely interested in the anti-spam fight, just the way this seems to have been pushed from your end definitely feels like (to me, at least) a bait-and-switch, though I do assume that's not your intention.
You also mention you're collecting usage data from this, but I fail to see where you've disclosed your privacy policy detailing exactly what data you will be collecting and storing, and for how long. ​stwalkerster (talk)
Oh, that data collection. The data I am collecting is just an analysis of how these Wikipedia articles change over time. We aren't collecting any user data. And this tool was built with a proprietary software framework, so it can't be open-sourced. But it might be possible to create an open-source knock off if it is popular, and I would even be willing to help with that. I'm not sure what you mean by "bait and switch". Sam at Megaputer (talk) 19:56, 6 April 2021 (UTC)[reply]
The fact that you do not understand what I am asking about with regard to data is a problem. How do we know what data it is collecting from someone using the tool? How do we know this isn't just scraping IPs? What privacy assurances are there? EGGIDICAE🥚 19:59, 6 April 2021 (UTC)[reply]
It's the same as when you visit any website. If I had built this tool to scrape IPs, I would be doing that in a very inefficient way. You are probably wanting to know what Megaputer is trying to get out of this. That is a perfectly reasonable question. My company let me work on this because they are hoping to make the news. As for myself, I am a Wikipedian who wants to help along the project. Sam at Megaputer (talk) 20:02, 6 April 2021 (UTC)[reply]
No, I honestly don't care. I want to know what user data this tool is gathering from USERS ACCESSING IT considering you've promoted it all over the place. This isn't a goodness of your heart creation, you even said as much. My company let me work on this because they are hoping to make the news. EGGIDICAE🥚 20:07, 6 April 2021 (UTC)[reply]
Why would any company do all this work just to steal data from a dozen or so Wikipedians? That makes absolutely no sense. And of course this isn't being done out of the goodness of Megaputer's heart. Megaputer is a company, and companies don't have hearts. I started building this thing during my unpaid internship. Megaputer let me do it because I was unpaid, and they really didn't care what I did as long as I learned the software. Now I am releasing it. Quite frankly, I'd like to see it do some good. Sam at Megaputer (talk) 20:14, 6 April 2021 (UTC)[reply]
Again, you're totally missing the point, aside from you spamming on behalf of your company who wants to "get in the news", what a tool used on Wikipedia does with user data does matter whether it's intentionally nefarious or not. That's why Stwalkerster asked about the privacy policy. You're really toeing a line here and you need to stop. Further, just because you've disclosed as a paid editor, doesn't give you free reign to use Wikimedia as a marketing stunt. EGGIDICAE🥚 20:16, 6 April 2021 (UTC)[reply]
Here is a link to Megaputer's privacy policy. Sam at Megaputer (talk) 20:20, 6 April 2021 (UTC)[reply]
Thank you for the privacy policy link. I've had a look, and I'm disappointed in a number of the vague statements regarding potentially indefinite data retention and the depth of user-identifying information that you appear collect. I will not be using this tool. stwalkerster (talk) 20:43, 6 April 2021 (UTC)[reply]
I respect your decision. But please bear in mind that there are many Wikipedians who use Facebook, Twitter, and even Google, all of which are known to violate your privacy in ways that Megaputer never could, even if we wanted to. These Wikipedians will likely not have such an objection to Megaputer's privacy policy. I very much hope that a few of them will give my tool a try. Sam at Megaputer (talk) 20:58, 6 April 2021 (UTC)[reply]
The difference is that those are individual choices that have nothing to do with Wikipedia. A tool for Wikipedia is a different matter, especially one that is promoted heavily by someone with a vested interest in that company. I would recommend dropping this given you literally said that it was an attempt to get media coverage. EGGIDICAE🥚 21:13, 6 April 2021 (UTC)[reply]
@Praxidicae When you navigate to an external site by clicking on a link from Wikipedia, they'll be able to know that you've come from Wikipedia, but they can't possibly know what your user account is or any other information related to your account. They'll also get your IP address and user agent string – but that's true of every website you visit on the internet. If you don't agree to that just don't visit their site. No need to make a fuss about it as it spurs technical innovation and discourages future technical contributors from even trying to improve WP. Hosting on Toolforge is good practise but it is certainly not a requirement since it's always up to users whether they want to use the tool or not. – SD0001 (talk) 04:25, 7 April 2021 (UTC)[reply]
SD0001 You've completely missed my point. There is a difference between viewing an article, clicking a source. No one is worried about that data scraping. This is someone who was literally paid/hired to write about the company hosting this tool and is creating the tool and promoting it in hopes of getting media attention per their own admission in this thread. Further, it would be very easy to scrape data from accounts using a tool like this - say it pulls up 5 spammy articles and I go and AFD/tag/CSD them, it's fairly obvious who was using the tool at that point. Further, ORES already does this. This is nothing more than a marketing ploy. EGGIDICAE🥚 15:17, 7 April 2021 (UTC)[reply]
This is nothing more than a marketing ploy So much for WP:AGF. I'm not missing your point – all the comments you've made here indicate you were worried about user data collection. You even started this discussion by creating FUD (... presumably will have access to private data for user accounts – which is technically impossible by the way). Even if the sole purpose of creating the tool was to get user data – it doesn't sound like the type of data that can be sold to DMPs for money. ORES is maintained by an understaffed team at WMF. I'm quite sure it is possible to create more sophisticated ML algorithms. I do not know whether this is one of those. But by the way you're treating this person, you're ensuring that no one makes another attempt. – SD0001 (talk) 17:38, 7 April 2021 (UTC)[reply]
SD0001 My company let me work on this because they are hoping to make the news. is a direct quote from the OP, who is literally paid to write content about the company hosting this "tool". It is quite literally a marketing ploy. So perhaps, I don't know, read through the thread next time. EGGIDICAE🥚 21:32, 7 April 2021 (UTC)[reply]
As much as I don't want to engage with you, I feel the need to say that I do not appreciate being told that the product of my hard work is a "marketing ploy". You have claimed that my tool is redundant to ORES (it is not), that it steals user data (which doesn't even make sense), and even that I lacked community approval to build it. You are clearly just throwing out allegations and seeing what sticks. All this started with allegations of "spam", and while I could have handled the rollout better, your response to this has been far more disruptive than anything I have done. Sam at Megaputer (talk) 22:02, 7 April 2021 (UTC)[reply]
I never once said that it steals data, I said that it's possible given your lackluster response about the privacy policy. And if I'm disruptive, please feel free to take it to the appropriate noticeboard. My marketing ploy comment is completely reasonable given your response to someone asking why you built it was that your company wanted to be in the news. The disruptive one here is you. EGGIDICAE🥚 23:18, 7 April 2021 (UTC)[reply]

I raised this on Sam's user page, but he took umbrage, which is undercutting my AGF. Is there any reason this user isn't yet blocked for all of the above as well as the non-answers & evasiveness re: his prior account? Courtesy @DGG, Praxidicae, Blablubbs, and RoySmith: StarM 18:23, 8 April 2021 (UTC)[reply]

Sorry but I don't agree with "helping the encyclopedia." He literally said it was because his company wanted to be in the news. Further, I'm super skeptical of an editor who is paid to write about their company reporting competitors for COI, afding articles and tagging other articles as paid/upe/coi. EGGIDICAE🥚 18:28, 8 April 2021 (UTC)[reply]
This is the seventh time you have claimed that I only built this thing as a marketing ploy, or to get in the news. [1][2][3][4][5][6][7] This is a totally unnecessary attack on my reputation. Sam at Megaputer (talk) 19:27, 8 April 2021 (UTC)[reply]
  • I was going to stay out of this, but since I was pinged, I'll comment. I would be happier if Sam disclosed his other account, but I've read WP:ALTACCN carefully, and I think a reasonable case can be made that forcing him to disclose the other account would be effectively forcing him to disclose his real-world identity, which we never force people to do. His employer is a small company; anybody familiar with the company would probably be able to figure out who Sam is based on the information he's already disclosed about his work there. He's disclosed that he's paid, disclosed that he's got another account, and disclosed additional details to a CU. Unless a CU wants to go further, I think it's time to drop this.
As for "We don't need this tool because it does the same thing as ORES", that's just plain silly. There's a huge range of AI algorithms used in classification problems. Even if you're using the same basic algorithms, training a model on a different data set is a significant difference. I'd be happier if they open-sourced the code, but people are free to use the wikipedia data for commercial purposes, with no requirement to disclose their source code. There's also the argument that in a domain like spam detection, disclosing every detail of your model just makes it easier for the bad guys to alter their behavior to avoid detection. So, even in a purely altruistic, "all knowledge should be free" universe, I can understand not disclosing everything.
And, yeah, I wish Sam would not be quite so effusive about how wonderful his stuff is. If this was in mainspace, it might well be WP:G11 material.
And finally, on the topic of potential privacy violations by Megaputer, to be honest, I think we're in WP:FRINGE territory here. WMF has an exceptionally conservative privacy policy. It is unreasonable to require anybody who uses WMF data have an equally conservative policy. As long as they're not running on WMF systems, what they log is between them and their users. -- RoySmith (talk) 19:11, 8 April 2021 (UTC)[reply]
I have to disagree with the sock part, Roy – the "Sam at Megaputer" account is being used in ways that a "regular", unpaid account is and participating in discussions internal to the project (participating in AfD's and this thread, tagging an article by a firm that works in a similar industry to Megaputer etc.); per WP:SOCKLEGIT, Although a privacy-based alternative account is not publicly connected to your main account, it should not be used in ways outlined in the inappropriate uses section of this page. WP:ILLEGIT is clear about the fact that editing projectspace with an undisclosed alternative account is prohibited, and it makes clear that evasion of scrutiny is a violation of WP:SOCK. Using a privacy sock to participate in projectspace discussions arguably is evasion of scrutiny because it splits contribution histories and prevents people from evaluating your edits as a whole. Blablubbs|talk 12:56, 9 April 2021 (UTC)[reply]
  • I can confirm that the alternate account has been disclosed to me as a checkuser. I have checked just now the editing history. It has not been used to overlap in any manner the editing or interests of the current account. Whatever the merits of the project being discussed here, there is in my opinion no violation of sock policy. DGG ( talk ) 14:09, 9 April 2021 (UTC)[reply]
    DGG, the assertion being that it's okay to participate in discussions internal to the project with two undisclosed accounts as long as there's no overlap? If so, I think that that would need to be clarified in WP:SOCK. Blablubbs|talk 13:00, 10 April 2021 (UTC)[reply]
If you continue to have concerns, please do not discuss them in an open forum, but take them to arb com, to avoid the possibility of outing. DGG ( talk ) 23:33, 10 April 2021 (UTC)[reply]
DGG, I'm not seeing the outing concerns here - nobody's tossing out possibilities for who the "other" account is, and Sam has been quite clear about their relationship with Megaputer. They are just discussing whether this situation appears to be in violation of PROJSOCK/SOCKLEGIT, which is entirely reasonable as long as nobody's speculating about their other identity. GeneralNotability (talk) 23:59, 10 April 2021 (UTC)[reply]
It seem obvious to me that further discussion on this line will lead to revealing the identity of the other account. I I do not consider this suitable for discussion on a page devoted to purely technical matters. have listed this on the checkuser list for my colleagues awareness and advice. DGG ( talk ) 04:38, 11 April 2021 (UTC)[reply]
DGG, that is not "obvious" to me at all - again, nobody is even speculating about the identity of the other account. Perhaps it's obvious to you since you've been officially notified of the user's other account. While the discussion is currently more "policy" than "technical", I believe it is still reasonably on-topic (and moving the policy discussion onto VPP would unnecessarily split the discussion). GeneralNotability (talk) 14:18, 11 April 2021 (UTC)[reply]

Bot Partially Non-functioning

I think that I have a specific issue and a general issue to report. The specific issue is that a bot, User:MDanielsBot, has stopped doing one of its tasks. The task is Task 6, which is to clerk the Dispute Resolution Noticeboard by maintaining a table that summarizes the status of disputes. It has stopped maintaining that table. The bot seems to be doing Task 4 properly, which is to dispose of stale reports at the vandalism noticeboard. The instructions say that if the bot is malfunctioning, administrators can press a button to block it, or non-admins can report it at WP:ANI. Presumably a report at WP:ANI will result in the bot being blocked. Any bot should be blocked if it is doing something wrong. In this case the bot isn't doing anything wrong. It isn't doing something right that it should do. The general issue is what should be done if a bot stops doing one of its tasks, and the bot maintainer is on a long wikibreak. It appears that User:Mdaniels5757 has posted a notice saying that they will be back in a few months. In the meantime, their bot is doing its most important task, and isn't doing another task.

What should be done with this bot?

What should be done with bots that are partly non-functioning and do not have a current bot administrator? Robert McClenon (talk) 18:16, 6 April 2021 (UTC)[reply]

This is a typical problem with bots. There's nothing that can be done about it. Try emailing Mdaniels; if he sees it, and has time, he might fix it. If not, WP:BOTREQ to find someone else to create a similar bot (presumably this may help make it quicker for someone to do so). ProcrastinatingReader (talk) 18:20, 6 April 2021 (UTC)[reply]
User:ProcrastinatingReader - Is the link the Python code that performs the task? If so, that would mean that it would be minimal work for a Python coder to use the existing code, and that would be why User:Firefly is able to make such an offer. I will send the email and provide an update in between 24 and 72 hours. Robert McClenon (talk) 22:27, 6 April 2021 (UTC)[reply]
It certainly appears to be - I've not tested it, but at the very least it would cut down the work required as the basic algorithm is there to see. ƒirefly ( t · c ) 06:23, 7 April 2021 (UTC)[reply]
Just noting that I’d be happy to have FireflyBot take this on if it’s deemed necessary (BRFA would be required of course). Ping me if MDaniels doesn’t get back to you! ƒirefly ( t · c ) 20:55, 6 April 2021 (UTC)[reply]
User:Firefly - I have not received a response from User:Mdaniels5757, and it has been more than 72 hours since I sent it. They are either not answering email or not responding to Wikipedia email. That is all right, since we are all volunteers, and they said that they were on break. So it would be appreciated if your bot could take on this extra task. Thank you. Robert McClenon (talk) 03:52, 10 April 2021 (UTC)[reply]
@Robert McClenon: Not a problem, I'll have a look. ƒirefly ( t · c ) 11:11, 10 April 2021 (UTC)[reply]

Where is this totally incorrect media viewer caption coming from?

Resolved
 – Fixed via editing. — xaosflux Talk 09:43, 7 April 2021 (UTC)[reply]

Hi. On John C. Calhoun, click on his photo in the infobox to open the viewer. The caption is totally wrong:

Oil on canvas painting of John C. Calhoun, perhaps in his fifties, black robe, full head of graying hair

Where is this coming from? It is obviously false so I tried to correct it, but it is not on the Commons page for the photo or anywhere on the article it displays on. Thanks, DemonDays64 (talk) 02:34, 7 April 2021 (UTC)[reply]

@DemonDays64: it is in the page text, the 8th line of wikitext as the "|alt=" attribute for that image. — xaosflux Talk 02:36, 7 April 2021 (UTC)[reply]
@Xaosflux: oh lol somehow i forgot to check the source. thanks for the super quick answer DemonDays64 (talk) 02:37, 7 April 2021 (UTC)[reply]
The image was changed 2 April. PrimeHunter (talk) 09:42, 7 April 2021 (UTC)[reply]

Read-only time April 28 at 05:00 UTC

Because of maintenance on their primary database master, there will be a read-only period on English Wikipedia starting around 05:00 UTC on April 28. This means you can read, but not edit, the wiki. The window is 30 minutes long, but it will most likely be significantly shorter. See phab:T279505. /Johan (WMF) (talk) 11:36, 7 April 2021 (UTC)[reply]

Clean up watchlist of bots

Tracked in Phabricator
Task T258098

Hi, A significant portion of the watchlist table in this wiki is just watchlist of bots. It's because "watch pages and files I create" is turned on by default and bots that create pages (like user talk pages) have them added to theirs. It distorts Special:UnwatchedPages and I doubt any user checks their bot's watchlist so I'm inclined to delete all of them here but let me know if you object to this. Also, if your bot created lots of pages here and you're okay with the rows being deleted, please let me know here and I'll delete it sooner. I already deleted mine. See phab:T258098 for more info. Ladsgroupoverleg 13:54, 7 April 2021 (UTC)[reply]

@Ladsgroup: can we get a list of the bots with the most entries? We can open a discussion at WP:BOTN, call out the bot operators, and do a silence-is-consent period. — xaosflux Talk 14:01, 7 April 2021 (UTC)[reply]
I'd also be supportive of having a dev manually turn off that setting (one time run) for any of those same identified bots that are still creating pages. — xaosflux Talk 14:02, 7 April 2021 (UTC)[reply]
The auto-watch preference has been ignored by all bots since January 7, following phab:T258108 — MusikAnimal talk 16:17, 7 April 2021 (UTC)[reply]

The top bots are: 

+-----------------------+----------+
| user_name             | count(*) |
+-----------------------+----------+
| ClueBot NG            |  3664794 |
| SmackBot              |  1048014 |
| SineBot               |   595420 |
| COIBot                |   464596 |
| HostBot               |   437022 |
| RjwilmsiBot           |   398358 |
| XLinkBot              |   372374 |
| Polbot                |   332758 |
| AvicBot               |   313928 |
| ClueBot               |   292196 |
| RussBot               |   285662 |
| LaraBot               |   233820 |
| ClueBot III           |   225346 |
| Citation bot          |   220294 |

The total number of rows in watchlist is 287M rows so the first user alone is responsible for 1.3% of all rows. Ladsgroupoverleg 14:19, 7 April 2021 (UTC)[reply]

Oh lord, cluebot what have you done. I think it would be worthwhile to also purge the watchlists for these bots if not most/all bots. —moonythedwarf (Braden N.) 14:39, 7 April 2021 (UTC)[reply]
@Ladsgroup: A proposal to clear these has been opened at Wikipedia:Bots/Noticeboard#Clearing_bot_watchlists. We normally give operators a week or two to respond to bot clean ups of this sort. — xaosflux Talk 14:52, 7 April 2021 (UTC)[reply]
phab:T258108 was resolved in early January and was announced in Tech News, so these bots have already stopped auto-watching new pages. If they do need to watch pages, they have to explicitly set the 'watch' option when editing with the API. — MusikAnimal talk 16:07, 7 April 2021 (UTC)[reply]

The logo looks blurry in Timeless skin

Tracked in Phabricator
Task T279645

as title--John123521 (Talk-Contib.) 14:28, 7 April 2021 (UTC)[reply]

It displays https://en.wikipedia.org/static/images/mobile/copyright/wikipedia.png which is 100px but in Timeless it's shown at 135px, at least for me. That does look bad. PrimeHunter (talk) 15:45, 7 April 2021 (UTC)[reply]
Can confirm here.--Vulphere 15:43, 9 April 2021 (UTC)[reply]
phab:T279645 opened, replicated on eswiki and dewiki as well. — xaosflux Talk 10:59, 8 April 2021 (UTC)[reply]

Category for the Philippines

The universal template for sports events by month by country works fine for most countries (and is very useful!) but for the Philippines displays the category “2015 in Filipino sport” instead of “2015 in Philippine sport” (which has a temporary fix). Can this be fixed please? See Category:January 2015 sports events in the Philippines. Hugo999 (talk) 21:41, 7 April 2021 (UTC)[reply]

Hugo999 The problem is that the nationality isn't the same as the adjective for something pertaining to the country. Looks like a {{country2adjective}} template should be created to deal with these cases. I think Le Deluge may be intressted in dealing with this. --Trialpears (talk) 22:00, 7 April 2021 (UTC)[reply]
Well it should be possible; cf Dutch sport & Events in the Netherlands! Hugo999 (talk) 09:48, 9 April 2021 (UTC)[reply]
An adaptation of the (bicontinental) Russian (or Turkish) sports categories works fine for the Philippines (and does not require editing for specific months/years!); but other editors have to be aware that they cannot use the usual universal category for the Phillipines; see Category:January 2015 sports events in the Philippines. Hugo999 (talk) 11:12, 11 April 2021 (UTC)[reply]

Update and improvement regarding Template:Singapore legislation

With effect from 1st March 2021, chapter numbers are no longer used when citing Singaporean enactments. According to the Interpretation Act 1965, as amended by the Statute Law Reform Act 2021:

8.—(1) It is sufficient for all purposes to refer to a written law—

(a) in the case of an Act—

...

(ii) where the Act, as enacted or revised, provides that it may be cited by a short title—by that short title; and

...

— section 8(1)(a)(ii), Interpretation Act 1965

Examples of using the new citation method can be seen at new bills like this and this.

I therefore propose to make updates on {{Singapore legislation}} to bring conformity with the new citation method.

Also, I would be grateful if the updated Singapore legislation template can direct references to individual sections, like that in {{Cite Hong Kong ordinance}}. This helps standardization of various citations.

As I have no experience on dealing with these complex templates, and are clueless about how to ensure that enactments are redirected to the correct external pages, I hope anyone who are familiar with templates can give a helping hand on these problems, many thanks.廣九直通車 (talk) 03:24, 8 April 2021 (UTC)[reply]

@Sgconlaw and Jacklee:Also notifying relevant users who are experienced in Singaporean law to give suggestions.廣九直通車 (talk) 03:24, 8 April 2021 (UTC)[reply]
OK, I can look into it. Whether linking to specific sections, etc., is possible depends on whether the website actually has links to them. — SGconlaw (talk) 04:53, 9 April 2021 (UTC)[reply]
Perhaps it will be better to create another new template for more specific section linking? Anyways, I also find that adding "#pr*-" (where "*" is a number) at the end of the principal URL at Singapore Statues Online directs you to the specific section.廣九直通車 (talk) 09:17, 9 April 2021 (UTC)[reply]

Wikipedia app's SuggestEdit ruining short descriptions

Tracked in Phabricator
Task T279702

Not sure if this is the right place to discuss this, but the SuggestEdit-add 1.0 feature on the android app is encouraging new users to change short descriptions in a way that always violates WP:SDFORMAT, and are half the time complete nonsense. Firstly, it suggests users uncapitalize short descriptions, which leads to new editors (very reasonably!) trusting the app and incorrectly uncapitalizing short descriptions en masse: [8] [9] [10] [11] [12].

When the program does decide to change the actual content in short descriptions, it very often adds complete nonsense. Here's a really lovely short description [13] it suggested (and successfully encouraged a new user to add!) to natural science: branch of science about the natural world and how it relates to statistics, prediction, low-entropy thinking, extra sensory perception (ESP), Prophecising, Apocalyptic Revelations, God, The Trinity, David, Kyle, Alpha, Omega, K, Cabal, cosmos, Wikis. Great stuff.

It's plausible that there's some team I could take this issue up with and some marginal improvement can be made to the program in a months time, but really, I can't see what benefit this possibly serves to the project, it is effectively always disruptive. ‑‑Volteer1 (talk) 17:34, 8 April 2021 (UTC)[reply]

@Volteer1: you may want to report this as a bug using this form - sounds like the problem is that this feature is making inappropriate suggestions. — xaosflux Talk 17:38, 8 April 2021 (UTC)[reply]
Bug report submitted, T279702. – Jonesey95 (talk) 18:00, 8 April 2021 (UTC)[reply]

Problems accessing WMF sites

Hi - an editor I know and trust has contacted me by e-mail to tell me that she is unable to access WMF sites today - no commons, no EnWiki, no EsWiki. Not that she can't edit, but she can't connect at all, her browser is giving her "The server at en.wikipedia.org is taking too long to respond." She's checked with friends and family in the same locale with same ISP, they can get access, but despite multiple reboots, trying different browsers, even trying her husband's computer, she's not getting access. Other websites all work fine, just WMF sites won't load for some reason. Any thoughts about what to try? If it matters, she's in Mexico. Thanks in advance for any suggestions, which I will have to forward by e-mail (since she can't see this page!). GirthSummit (blether) 18:15, 8 April 2021 (UTC)[reply]

I wonder what en.wikipedia.org resolves to for her? SQLQuery me! 18:19, 8 April 2021 (UTC)[reply]
Tell her to try using Google Public DNS (8.8.8.8 and 8.8.4.4), this is almost certainly an issue on her end. ‑‑Volteer1 (talk) 18:22, 8 April 2021 (UTC)[reply]
Thanks both. I'll ask her... GirthSummit (blether) 18:31, 8 April 2021 (UTC)[reply]
Oh, she just told me that she's been able to get onto https://wikipedialibrary.wmflabs.org/users/my_library/, but that's the only WMF related site she's been able to get to load so far. In case that gives you a clue. GirthSummit (blether) 18:33, 8 April 2021 (UTC)[reply]
SQL - she sent me a screenshot of what happens when she types 'en.wikipedia.org' into her address bar. The address bar changes to https://en.wikipedia.org/wiki/Main_Page, but she's just seeing the failure screen: The connection has timed out. The server at en.wikipeida.org took too long to respond, etc. Any more thoughts?
Volteer1 - I'm afraid that might be excessively technical for her. She's retired, a first-rate writer, researcher and overall contributor, but her technical skills are not too slick, she's daunted at the thought of attempting that, having had one look at what comes up when she Googles 'Google Public DNS' (I confess, I would be too). GirthSummit (blether) 19:11, 8 April 2021 (UTC)[reply]
Girth Summit, Try https://wikitech.wikimedia.org/. That's another wiki that's in a WMF datacenter, but outside the WP:SUL domain, as is (I believe), https://wikipedialibrary.wmflabs.org. If you can get to both of those, but none of enwiki. eswiki, commons, wikidata, etc, then this smells like an auth-related issue.
Today is WP:THURSDAY. A new version of the software was rolled out to enwiki today. There were some recent changes that dropped support for some very old browsers. I don't think that was in today's release, but it's something to consider if she's running something very old.
I would suggest looking in the javascript console and/or network tabs of the developer's tools, but from what you say, I'm guessing that's beyond her technical reach. -- RoySmith (talk) 22:31, 8 April 2021 (UTC)[reply]
RoySmith, thanks very much - I'll forward that. GirthSummit (blether) 22:42, 8 April 2021 (UTC)[reply]
Girth Summit, Hmmm, interesting, it turns out https://wikipedialibrary.wmflabs.org isn't running WikiMedia at all. It's running something completely different. Which makes me suspect even more that this is a WP:THURSDAY issue. -- RoySmith (talk) 23:23, 8 April 2021 (UTC)[reply]

Tracked in Phabricator
Task T279693

All the edit links are now showing an arrow indicating an external link. I do not know if this is a bug or what, but the class=plainlinks should stop it, and yet it is not. It looks dreadful. Can this be fixed? Aasim (talk) 01:18, 9 April 2021 (UTC)[reply]

BTW this is happening for other plainlinks, including links in AFD templates, links on shortcuts templates, even on some interface messages. Aasim (talk) 01:20, 9 April 2021 (UTC)[reply]
Awesome Aasim, Can you give a specific example? A screenshot would help too. -- RoySmith (talk) 01:38, 9 April 2021 (UTC)[reply]
See . This appears to be a Timeless issue. Aasim (talk) 03:37, 9 April 2021 (UTC)[reply]
Yes, there is already a Phab task for it. Izno (talk) 03:58, 9 April 2021 (UTC)[reply]
Can you link it in the discussion? Aasim (talk) 04:49, 9 April 2021 (UTC)[reply]
So, hotfix by adding plainlinks to MediaWiki:Timeless.css, or wait for an fix in mediawiki core? Plainlinks is not present in Timeless at all, currently.--Snaevar (talk) 06:09, 9 April 2021 (UTC)[reply]
@Awesome Aasim: just want to verify that this is not actually all the edit links on a page - for example on this random page the "edit links" on: sections, the literal labeled section "edit links" in the languages box, and the "edit this page" control at the top of the page are fine. You are only seeing this on full url's that contain things such as ?action=edit or ?redirect=no correct? — xaosflux Talk 11:01, 9 April 2021 (UTC)[reply]

Bug report (New lines in API textextracts)

Tracked in Phabricator
Task T279789

The API seems to be interpreting line breaks in the wikitext as new lines even though they don't render that way to the reader. For example, in the article Hytracc Consulting the API call sends the sentence:"The company is \nheadquartered in Stavanger, Norway, and also has offices in \nAberdeen, Calgary, Groningen, Houston, Trondheim, and \nKuala Lumpur." Can someone link me to the right place to report this? Sam at Megaputer (talk) 14:16, 9 April 2021 (UTC)[reply]

The API is accurately reflecting the source wikitext and doesn't know anything about how HTML is rendered by a browser. This is not a bug. Izno (talk) 14:24, 9 April 2021 (UTC)[reply]
If the API is reflecting the source wikitext, then why do double linebreaks get reported as single line breaks (see the text after Kuala Lumpur.\n)? Sam at Megaputer (talk) 14:35, 9 April 2021 (UTC)[reply]
What you've linked to is the output of the mw:Extension:TextExtracts API, which uses certain heuristics to produce a page extract and so unlike mw:API:Revisions it will not necessarily reflect the source wikitext. If you think it's a bug, report it here, though high changes are that it's normal behaviour. The wikitext formatting of that page is weird anyway. – SD0001 (talk) 14:57, 9 April 2021 (UTC)[reply]
Thanks. I'm going to report this since it's effecting my data. We'll see if I get a response. Sam at Megaputer (talk) 15:06, 9 April 2021 (UTC)[reply]
Sam at Megaputer, Sam, this is working as designed. See WP:LINEBREAK#Causing line breaks. Sadly, I don't know of any authoritative documentation for wiki syntax and semantics. The language is poorly designed and full of historical oddities.
If you're interested in parsing wikitext, the official answer is to use Parsoid. Parsoid is big, complicated, and confusing to use, but it is the final word on parsing wikitext. If you're using Python, mwparserfromhell is a reasonable alternative; it's much easier to use, but diverges from Parsoid in some minor details. If you're trying to parse wikitext with regexes, you're doing it wrong. -- RoySmith (talk) 16:00, 9 April 2021 (UTC)[reply]
Thanks for the suggestions. I see from WP:LINEBREAK#Causing line breaks that "\n" in the wikitext is expected to register as " " in HTML, while "\n\n" registers as a line break in HTML. But the TextExtract API maps both values to "\n", which makes it impossible to tell if "\n" should be interpreted as a line break or not. If I don't get a response to my report, it might just be easier for me just to consider this as acceptable noise that to try to re-parse the wikitext. That's always an option, though. Sam at Megaputer (talk) 17:54, 9 April 2021 (UTC)[reply]
Sam at Megaputer, I have never used TextExtracts myself, but from reading the docs, it looks like this was intended as a way to grab just the text without any of the markup, as you might want if you were building a search index. If you want to get any of the markup semantics (i.e. "there's a paragraph break between these two words"), I suspect Parsoid is your friend. -- RoySmith (talk) 20:32, 9 April 2021 (UTC)[reply]

Text size changes in references on mobile device?

insert a caption here

This is one of those things that's been annoying me forever. On my Android phone viewing the desktop site with Chrome (Vector skin), references appear in either normal or reduced size text (see screenshot).

This seems to be related to whether I've visited a link or not, but it's not that simple. For example, if I tap on the first reference (Connor), the first thing that happens is the text shrinks to the small size. If I tap again, it takes me to the UPI page. If I go back to the wikipedia page and reload, it's back to being the normal size. -- RoySmith (talk) 17:20, 9 April 2021 (UTC)[reply]

@RoySmith: It sounds like Wikipedia:Village pump (technical)/Archive 188#What does it mean when a table column font has an unwanted enbiggenment not commanded by wikitext? No way to stop it was posted, apart from using the mobile site. PrimeHunter (talk) 17:37, 9 April 2021 (UTC)[reply]
PrimeHunter, Hmm, yeah, that does sound like exactly what I'm seeing, right down to "often result in clicking an unwanted link which has now squirmed to under your finger". Just for fun, I put my desktop Chrome into mobile emulation mode, and couldn't reproduce it, not that I really expected it would. -- RoySmith (talk) 21:20, 10 April 2021 (UTC)[reply]
@Sagittarian Milky Way: who started that thread. -- RoySmith (talk) 21:22, 10 April 2021 (UTC)[reply]
You can set long press threshold to 0.3 seconds or so. The whole holding down a link then letting go then re-aiming then clicking without risk procedure doesn't have to take that long. Sagittarian Milky Way (talk) 22:06, 10 April 2021 (UTC)[reply]

coding error?

See Wikipedia:Administrators' noticeboard/Incidents#lowercase sigmabot III malfunction?. Part of an archived thread is visible in the edit window, but does not display live (bot does not appear to be the issue). My code-fu is apparently not sufficient to puzzle it out. Beeblebrox (talk) 18:42, 9 April 2021 (UTC)[reply]

Fixed. A closing ref had wrong syntax. PrimeHunter (talk) 19:24, 9 April 2021 (UTC)[reply]

google books URLs need a bot

URLs q.v.:

https://books.google.com/books?id=sDeE9r_6HdsC&pg=PT234&lpg=PT234&dq=Jiyu+hiwar&source=bl&ots=uqCPAz1gx6&sig=2ZDD2RkxILyXnIqm0o42i5PfndU&hl=en&sa=X&ved=0CBwQ6AEwAGoVChMI3qqt6NvhyAIVxWQsCh07igs-#v=onepage&q=Jiyu%20hiwar&f=false

can be cleaned of contributor metadata:

https://books.google.com/books?id=sDeE9r_6HdsC&pg=PT234&lpg=PT234&dq=Jiyu+hiwar&hl=en
.... 0mtwb9gd5wx (talk) 02:45, 10 April 2021 (UTC)[reply]
From my experience (from some time ago), page numbers are useless. So the bare minimum is:
https://books.google.com/books?id=sDeE9r_6HdsC&dq=Jiyu+hiwar MarMi wiki (talk) 15:36, 10 April 2021 (UTC)[reply]

User:Citation bot does some GB URL cleanup. But I agree we need a dedicated bot for Google Books because there are many issues that can only be determined/fixed with header and page checks. There are millions of Google Books links on enwiki, and many of them no longer work as originally intended: hard and soft 404, link has a page # but redirects to "About this book", book ID usurped and points to a different book, etc.. more at Wikipedia:Google Books and Wikipedia. -- GreenC 03:04, 10 April 2021 (UTC)[reply]

IABot

This - https://iabot.toolforge.org/index.php?page=runbotqueue - used to be the link to IABot that I've used for years to archive links. For the past week or so there is an error message 503 Service Unavailable. Any idea what's going on? Firefox Browser, latest version. Thanks. Twofingered Typist (talk) 12:40, 10 April 2021 (UTC)[reply]

Tracked in Phabricator
Task T279341
It's down at the moment owing to a database issue. See phab:T279341. ƒirefly ( t · c ) 12:56, 10 April 2021 (UTC)[reply]
Thanks! Twofingered Typist (talk) 14:39, 10 April 2021 (UTC)[reply]

TOC font size in modern skin

Has there been a change made to the modern skin css to reduce the TOC font size? I might be imagining it but the font size seems a lot smaller than it was. Nthep (talk) 14:34, 10 April 2021 (UTC)[reply]

I have 100% certainty that this is the same TOC issue in Timeless as I added to phab:T279693. Izno (talk) 15:27, 10 April 2021 (UTC)[reply]
Glad it's not my eyesight or memory! Nthep (talk) 16:38, 10 April 2021 (UTC)[reply]

Picture title mis-spelled

There is a picture of the opening of the Stockton and Darlington Railway (File:Opening of the Stockton and Darlington Railway.jpg). There is also a cropped version (File:Opening of Stocking and Darlington Railway (crop).jpg) whose title is obviously mis-spelled (Stocking for Stockton). The latter is only linked once in en.wikipedia (from Stockton and Darlington Railway), but it is also linked from the es.-, ka.-, no.-, and zh.wikipedias.

Can I move this page to a correct spelling? Will this leave a redirect in place, and will that redirect work for the other wikipedias? I have never moved an article, and this is a file: rather than a mainspace article, so I don't want to make a mess through ignorance.

Also, how can I put a link to the file into this talk that shows the title (informative) rather than pulling in the picture (pretty, but not so germane)?--Verbarson (talk) 20:43, 10 April 2021 (UTC)[reply]

1. You don't have the requisite permissions to move the file, but you can request someone who does move to file by visiting the file's page on Commons (c:File:Opening of Stocking and Darlington Railway (crop).jpg) and clicking the "Move" link (which may be in a "More" menu). I've now done that for you. Once a Commons file mover processes my request, the move will leave a redirect, which will work on other Wikipedias. Nevertheless, it's customary for them to use a script to automatically replace all usages on all wikis, although I'm not sure why.
2. You prefix the link with a leading colon ([[:File:Example.png]] produces File:Example.png) * Pppery * it has begun... 20:55, 10 April 2021 (UTC)[reply]
Thank you. I'll keep an eye open for the rename to take effect.--Verbarson (talk) 21:38, 10 April 2021 (UTC)[reply]
As a Commons filemover, I've done the requested move. Graham87 03:15, 11 April 2021 (UTC)[reply]

Table of contents limit doesn't appear to work on QAnon page

A couple of times over the last few months, I've tried to set a Table of Contents limit (Template:TOC limit) on the QAnon page. There are currently 54 sections listed in the articles' ToC. Limiting it top-level headers only would bring the number down to a much more reasonable 15, making the page more readable for all users. However, each time I try to put the limit in place, it breaks something and the ToC either disappears, or the lead does. Does anyone know what's going wrong? Thanks. Ganesha811 (talk) 21:50, 10 April 2021 (UTC)[reply]

It looks like you inserted {{TOC limit|1}}, which does nothing. TOC level 1 is the page title. Level 2 are the main headings on the page, so {{TOC limit|2}} limits the TOC to a single level. A bit confusing, I know. – Jonesey95 (talk) 22:29, 10 April 2021 (UTC)[reply]
Jonesey95, thanks for the help! Makes sense. Ganesha811 (talk) 23:59, 10 April 2021 (UTC)[reply]
Level 1 headings have a single "=" but we don't use that in articles and rarely in other places. It's used for date headings in some pages, e.g. Wikipedia:Help desk. PrimeHunter (talk) 11:05, 11 April 2021 (UTC)[reply]

I've just noticed that Reflinks are still down. Would we be able to get this fixed please because I find it so important for content creation. I've asked on Meta but they haven't been much help. The C of E God Save the Queen! (talk) 08:05, 11 April 2021 (UTC)[reply]

I don't see how. The user seems to have disappeared. —TheDJ (talk • contribs) 09:22, 11 April 2021 (UTC)[reply]
It's giving the code errors, would WikiMedia be able to copy it over and fix it? @TheDJ: I ask because it allows you to fix all the errors at once rather than Citer which does it only one at a time. The C of E God Save the Queen! (talk) 10:34, 11 April 2021 (UTC)[reply]
Alternatively you could try using reFill which, presently, is working and will do multiple cites. - Derek R Bullamore (talk) 14:30, 11 April 2021 (UTC)[reply]
FYI The C of E reflinks was down for more than a year - more or less. Then it started again a couple months ago which was a big help but now it has been gone for a couple weeks so I wouldn't expect it back anytime soon. Its formatting of PDFs and marking of dead links are big pluses over refill but we just have to soldier on with the tools that are working. Thanks for your work dealing with bare urls. MarnetteD|Talk 18:19, 11 April 2021 (UTC)[reply]

Johnny Iguana

This article, which I recently created, has rendered the article's title in italics. I believe it is because I have included an album infobox within the article, as per the guidance given here. Is there a way of getting the article's title back to normal type, without the disruption of deleting the album infobox altogether ? Thank you. - Derek R Bullamore (talk) 13:02, 11 April 2021 (UTC)[reply]

Derek R Bullamore, you're completely right about the cause, the solution is adding italic_title=no to {{Infobox album}} --Trialpears (talk) 13:11, 11 April 2021 (UTC)[reply]
Thanks for that, and the correcting edit. Like a flash of lightning ! - Derek R Bullamore (talk) 13:13, 11 April 2021 (UTC)[reply]

Get a revid

Anyone know how to grab the latest revid for a given page from an API? Sam at Megaputer (talk) 18:15, 11 April 2021 (UTC)[reply]

Special:ApiSandbox#action=query&format=json&prop=revisions&titles=Wikipedia:Sandbox&rvprop=ids&rvlimit=1 works. * Pppery * it has begun... 18:17, 11 April 2021 (UTC)[reply]
Or Special:ApiSandbox#action=query&format=json&prop=info&titles=Wikipedia. Nardog (talk) 18:20, 11 April 2021 (UTC)[reply]
Thanks. Both of those work well. Sam at Megaputer (talk) 18:27, 11 April 2021 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(technical)&oldid=1017250487"